Fluid Viscous Dampers General Guidelines For Engineers Including A Brief History Taylor Dampers

Download as pdf or txt
Download as pdf or txt
You are on page 1of 288

Fluid Viscous Dampers

General Guidelines for Engineers


Including a Brief History

@Seismicisolation
@Seismicisolation
l
)
I

I/

I
'

'

l ��1 I I
I /
f
\"

I J,

I �1 I/ I' '' I
I J , I I
I I
\
�' -1---+-+--+..---------+--------+tll ·� /
@Seismicisolation #
@Seismicisolation l f � {. J \/
11 111 l/N ! ' J�
I

j
I/

!NU I / / � n A I/ J

1 1111 1
Table of Contents
Contributors .......................................................................................................................................................... 3
Foreword ............................................................................................................................................................... 5
1.0 —Introduction ................................................................................................................................................ 7
2.0 — Fluid Damping Devices – A Century of History.................................................................................. 9
3.0 — Design Description of Dampers ......................................................................................................... 15
4.0 — Damper Output Characteristics .......................................................................................................... 17
5.0 — Generalized Effects of Adding Fluid Dampers to a Structure ...................................................... 21
6.0 —Damper Mounting Arrangements ........................................................................................................ 27
7.0 —Design and Analysis of Building Structures Using Fluid Viscous Dampers ............................... 33
Part-1: Creating an ETABS Analysis Model with FVDs for NLRHA .................................. 34
Part-2: New Construction – Special Moment Frames with FVDs .................................... 66
Part-3: Seismic Retrofit – Moment Frames with FVDs ....................................................... 81
8 .0 —Fluid Damper Performance vs. Other Technologies ....................................................................... 95
9.0 —The Typical Process for Incorporating Dampers into Buildings ................................................. 105
10.0 —Photographs ......................................................................................................................................... 113
11.0 —Mounting Hardware ........................................................................................................................... 133
12.0 —Available Sizes and Dimensions ..................................................................................................... 141
13.0 —Case Studies ........................................................................................................................................ 145
14 .0 —Taylor Devices’ Literature................................................................................................................. 149
15.0 —Sample Technical Manual ................................................................................................................. 171
Appendix A – Case Study Papers ................................................................................................................. 181

@Seismicisolation
@Seismicisolation
1
@Seismicisolation
@Seismicisolation
2
Contributors
The staff at Taylor Devices is very Mark Berquist
proud to have collaborated with Taylor Devices, Inc.
experts in the field of damping to
produce this manual. We thank Richard DePasquale
everyone for their inspiration and Taylor Devices, Inc.
contributions.
Sean Frye
Taylor Devices, Inc.

Amir Gilani
Miyamoto International

Alan Klembczyk
Taylor Devices, Inc.

David Lee
Taylor Devices, Inc.

Aaron Malatesta
Taylor Devices, Inc.

John Metzger
Taylor Devices, Inc.

Robert Schneider
Taylor Devices, Inc.

Chris Smith
EQC Global

Douglas Taylor
Taylor Devices, Inc.

Shanshan Wang
University of California, Berkeley

Craig Winters
Taylor Devices, Inc.

© 2019 Taylor Devices, Inc.


All rights reserved throughout the world.
Reproduction and/or creation of derivative works strictly prohibited.

@Seismicisolation
@Seismicisolation
3
@Seismicisolation
@Seismicisolation
4
Foreword
Over the last 30 years or so, I have had the unique opportunity and
honor to work at Taylor Devices, Inc. Over that period, we have
seen substantial evolution within the Company, its products, its
customers, and its culture.

In the early to mid-1990’s, we had the opportunity to adapt some


of our new and old designs for use in structures to provide for
better earthquake resistance. Based on the results of laboratory
tests, analytical studies, and field performance during wind and
seismic events, this technology has spoken for itself and has shown
remarkable improvement in structural control. Indeed, fluid
dampers reduce deflection, stress, and acceleration at the same
time without increasing cost substantially or even at all.

It can be quite astonishing to see the dramatic effect that a


relatively small amount of damping can have on the performance
of a structure during dynamic events. However, it is also surprising
to me to see the use of inferior products and inferior technologies being used for improving
performance when well-proven designs already exist. When considering the use of a potentially
life-saving component, I like to remind people that what is acceptable for use in the US military and
aerospace sector ought to be appropriate for saving lives. It is important to note that fluid dampers
allow a structure to return to its initial condition after a large event. Demolishing a structure or
replacing energy absorbing elements after an event at a great cost is simply not necessary.

As structural engineers have now progressed out of the infancy phase of incorporating fluid
dampers into structures for seismic control, I feel that we have a responsibility to share the benefits
with as many concerned influences as possible. As managers, engineers, or academia, we have a
responsibility to move mankind forward with the knowledge we have obtained. However, we
should always do so with the highest standards of honesty and integrity while also holding each
other to the highest standards, regardless of the final design of a structure.

For this reason and many others, I am asking all our readers to share this knowledge while
recognizing that we do indeed share a high level of responsibility to make the earth a better place
by helping to save human lives in the event of natural phenomena for which we have little control.

This Damper Manual attempts to accommodate our common goals while recognizing that we all
have important choices to make. Taylor Devices is always available to help make the right choices.
We recognize the merits of our products. We hold our products and ourselves to the highest quality
standards.

Additionally, this Manual is not intended to represent fluid dampers as simple commodities and
therefore encourage the use of inferior processes or products. Although the technology has proven
itself, the process and products shall always be held to the highest standard. Taylor Devices
recognizes that important responsibility.

Alan Klembczyk

President, Taylor Devices, Inc.

@Seismicisolation
@Seismicisolation
5
@Seismicisolation
@Seismicisolation
6
1 Introduction

The end of the Cold War in 1990 heralded a restructuring period for the American military and
defense industry. One of the outcomes of this new era was that political and economic change
allowed previously restricted technologies to become available to the general public. This conversion
of defense technology is typified by highly advanced products and services that suddenly appeared in
the marketplace, seemingly out of nowhere. Perhaps the best known of these is the now ubiquitous
Internet, which in reality came from 1970's defense technology intended for use by government
agencies in the event of nuclear war.

In the civil engineering field, high capacity fluid dampers have transitioned from defense related
structures to commercial applications on buildings and bridges subjected to seismic and/or wind
storm inputs. Because fluid damping technology was proven thoroughly reliable and robust through
decades of Cold War usage, implementation on commercial structures has taken place very quickly.

Indeed, over the last 30 years, utilizing various types of added-damping devices in structures has
emerged as a useful, reliable and predictable tool in significantly improving the resiliency of structures
to a dynamic input. Much research and testing have been performed that verifies the benefits of
incorporating added-damping devices in structures. Linear and non-linear fluid viscous dampers
continue to demonstrate excellent performance in reducing deflection, acceleration response, inter-
story drift and stress. Damping device designs that have been well proven through decades of use
are available in configurations that provide forces that depend on input velocity, deflection, or a
combination of both.

Although various building codes have emerged throughout the world that address methods and
response requirements of structures when utilizing damping devices, these codes do not provide a
general comparison in improved resiliency that is realized through their use.

The concept of damping within a structural system can have different meanings to the various
engineering disciplines. To the civil engineer, damping may mean only a reference note on a seismic
or wind spectral plot, “5% damped spectra” being the most common notation. To the structural
engineer, damping means changes in overall stress within a structure subject to shock and vibration,
with frequent arguments whether a structure will have “2%, 3%, 4%, but not more than 5%” structural
damping. On the other hand, mechanical engineers do not necessarily view damping as a benevolent
feature, since machines, by definition, are supposed to transmit forces and motions efficiently,
without energy losses. Thus, the need for damping in a machine often signifies that an engineering
design error has been made.

In the classical mechanical engineering text “Vibration Theory and Applications,” William Thomson
[1] avoids a single, direct definition of damping by offering the following descriptions: “Vibrating
systems are all more or less subject to damping because energy is dissipated by friction and other

@Seismicisolation
@Seismicisolation
7
resistances. Since no energy is supplied in free vibration, the motion in free vibration will diminish
with time, and is said to be damped.

It follows from these descriptions that a damper is an element which can be added to a system to
provide forces which are resistive to motion, thus providing a means of energy dissipation. Assuming
that this working definition will suffice for general use, the next area of interest is to generally
describe the functional output of a damper. As with the definition of damping, the functional output
of a damper is somewhat controversial, since different output equations exist within the context of
the various engineering disciplines.

Alternatively, damping can be defined as that attribute of a dynamic system that results in a decrease
in the amplitude of oscillation. This results in the removal of some amount of energy in that system.
In keeping with the law of conservation of energy, this energy is actually transformed into another
form. Consequently, the term “damper” can be defined as that mechanism or internal property that
provides this transfer of energy. Typically, damping converts mechanical energy into heat. This
heat is then dissipated to the surroundings through any of the 3 modes of heat transfer defined as
conduction, convection and radiation.

Fluid viscous dampers operate by providing a resisting force only when moving. They do not add
stiffness to a structure, and they do not carry any static load.

Like automobiles driven on a bumpy road, buildings in seismic regions are a dynamic problem. Who
would ever buy or manufacture a car without shock absorbers? The dynamic laws of physics are the
same for each.

It is with great pleasure that Taylor Devices offers this damper manual as a guide for engineers with
various levels of experience in order to take advantage of this technology that has been successfully
transitioned from previous applications to now improve the dynamic performance of structures and
to help save lives throughout the world.

REFERENCES
1. Thomson, William, 1965, Vibration Theory and Applications, Prentice-Hall, Englewood Cliffs, New
Jersey.

@Seismicisolation
@Seismicisolation
8
2 Fluid Damping Devices
A Century of History

It is axiomatic that during times of war, new technology develops extremely quickly, since the fates
of nations may well depend upon which antagonist can mass-produce improved weapons more
quickly. In the case of fluid dampers, the evolution of large bore artillery and naval guns in the late
1800's provided the need for the product, and the various major governments were only too eager to
provide the development funding.

The Guns of War, 1897-1918 — Necessity Fosters Invention


The evolution of large dampers began with the advent of large breech loaded cannons in the 1860's.
Prior to this, large guns were muzzle loaded in a very time-consuming manner. Gaining easy access
to the gun’s muzzle end for loading was simple; the weapon was merely allowed to move backwards
anywhere from one to twenty feet after firing. Motion was retarded by means of a spade-like device
literally digging into the earth on land-based weapons. Shipboard guns used friction slides or inclined
surfaces to arrest their firing motion, often aided with block and tackle mechanisms. After loading,
the gun crew would push the gun back into its “battery,” or ready to fire position. The advent of breech
loading allowed for much more rapid (and safer) loading of the weapon, and a desirable higher rate
of fire. Unfortunately, the high firing rate required that the gun crew work much faster repositioning
the gun, quickly exhausting the crew.

Several unsuccessful concepts of arresting gun recoil were attempted, involving both coil springs and
rubber blocks. Meanwhile, the inventors of that time were investigating the new field of hydraulic
components, and by the late 1860's, experiments were taking place using hydraulic dampers to
arrest gun recoil. It is reported by Hogg [1] that the British Army was the first to use hydraulic recoil
dampers on gun carriages in 1862. The first mass-produced hydraulic recoil damper was used on the
75 mm French field gun, Model M1897. This weapon was hailed as a true technological marvel and
is considered to be the first modern artillery piece. The carriage of the weapon included a slide to
support the gun itself, and a 48-inch stroke fluid damper combined with a light spring to attenuate
recoil energy and return the gun to battery. The French M1897 went on to serve in both World War
I and World War II. Many variations of the weapon exist since many countries “borrowed” the design
after capturing one or more examples during World War I. One of the more unusual uses for the low
recoil French M1897 was by the U.S. Army Air Corps during World War II. The Air Corps needed
a ground attack aircraft with as much firepower as possible. The solution to the problem involved
mounting a complete M1897 with recoil dampers into the nose of the U.S. Model B-25 “Mitchell”
Bomber, firing forward. The modified aircraft proved successful, and the use of the hydraulic dampers
eliminated damage to the aircraft.

By the end of World War I, tens of thousands of fluid dampers were being used on field artillery
pieces, naval guns, coastal guns and railway guns. Some dampers of this period were even of the semi-

@Seismicisolation
@Seismicisolation
9
active type, where changing the gun elevation angle would change the resultant damping force. This
was accomplished by using a gear train between gun carriage and the damper. The gear train would
rotate an adjustment rod or screw protruding from the damper cylinder. As the gun was elevated,
the damper would become “stiffer”, and use less displacement. This feature allowed the gun carriage
to be reduced in size and weight, since at high elevation angles, the carriage no longer needed to
maintain clearance to the ground for the entire recoil stroke.

Toward the end of World War I, another advantage of fluid dampers was discovered. This was that
reduced recoil allowed weapons to easily fire larger projectiles, with larger propellant charges to
obtain greater range. Indeed, from March to July of 1918, the City of Paris was attacked by the
German Army with a weapon of “super gun” proportions. Details did not become available until the
war ended, and then only after intense efforts by the allies. The weapon was named the Paris Gun,
and included a 130-foot long barrel, which fired a 210 mm diameter shell at a range up to 85 miles.
The gun itself, with fluid dampers, weighed over 140 tons, not including the weight of the tremendous
carriage that carried the weapon. Three of the Paris Guns were built, but all were withdrawn from
service as the allied armies approached their locations. Mysteriously, none were recovered by the
allied forces after the war ended.

The Automotive Damper - Optimization Through Evolution


The 1920's and 1930's were a period when the automobile became a dominant feature of American
culture. Since the automobile was a relatively new product with a large potential market, automotive
manufacturers were forced, by competitive pressures, to produce a product that would be appealing
to the consumer. One of the most appealing traits that an automobile could possess was a smooth
ride over all possible road surfaces; this proved to be a true challenge for automotive engineers of
this period.

The earliest auto suspensions were simply carried over from horse-drawn wagons. The suspension
consisted of multiple leaf elliptical or semi-elliptical springs. Damping was limited to the inter-
leaf friction which occurred as the spring leaves ground over one another as the spring deflected.
Damping would obviously have a high variance from day-to-day, depending on whether the spring
was dry, wet, rusty, dirty, or recently cleaned and oiled.

This day-to-day damping change proved unacceptable to the consumer, and external friction pad or
rubber dampers were added to the suspension. These provided a small but noticeable improvement
over using the spring itself as a damper, plus it was possible to make the damper adjustable for wear.
The “ideal” damping material was usually pure asbestos washers or pads, compressed between two
iron plates. One plate was fixed to the car frame by a bolt, the other was attached to an actuating arm.
A large draw bolt went through the center of the damper assembly, and tightening or loosening of the
bolt served to adjust the damping force.

The high maintenance and marginal improvement obtained with friction and rubber dampers caused
automotive parts suppliers to look for improved damping systems, and fluid dampers quickly entered
the scene. The biggest problem with adapting the fluid damper for automotive use proved to be poor
quality seals. The guns of World War I usually needed a major overhaul every 500 rounds or so, due
to barrel wear, and this was an opportune time to change damper seals, which usually were leaking
badly after 500 cycles. Considering that the seals of the day consisted of cut lengths of hemp rope
forced into a pocket with a hammer, this was no surprise! “Improved” seals of the 1920's consisted of
a stack of round leather washers forced into position with a packing nut. These were an improvement
over hemp strands, but still could not provide the cyclic life necessary for automotive use.

@Seismicisolation
@Seismicisolation
10
In 1925, Ralph Peo of the Houdaille Company in Buffalo, New York, invented a solution to the seal
problem. Instead of improving the seal, he redesigned the damper to use a rotating piston rod and
vane assembly, thus replacing long travel, sliding seal motion with a short 60-120 degree rotary
travel. The Houdaille rotary damper was actuated by crank arms attached to the moving components
of the suspension. The short rotary travel of the seal allowed for roughly 10,000 miles of road travel
before seal replacement became necessary. Within a short period, most automobiles were using
the Houdaille rotary damper. Figure 2.1 is one of the original patent sheets depicting Peo’s 1925
invention.

In 1949, the Delco Division of General Motors finally designed a sliding seal damper that had an
adequate life for automotive use, thus ending the rotary damper era. Present-day automotive shock
absorbers have an internal construction that is very similar to the gun recoil buffers of World War I,
except that modern seals provide substantially greater life.

Figure 2.1
Patent Sheet – R. Peo’s Rotary Shock Absorber

The Cold War - Dampers Go Underground


History texts will eventually include great amounts of information about the Cold War period, which
lasted from the end of World War II to approximately 1990. Early on in the Cold War, both the United
States and The Soviet Union began developing intercontinental ballistic missiles (ICBM), equipped
with nuclear warheads. Although still debated, most defense analysts state that the U.S. strategic
war doctrine was such that our missiles would not be launched until enemy warheads had actually
detonated on or above U.S. soil. Adherence to this doctrine assumed that the enemy’s initial targets
would be U.S. missile launchers, striking as many of these as possible in a first strike. In order for
the U.S. to launch a counterstrike under these conditions, our missiles needed to be designed and/
or based in such a way that they could survive a nuclear attack without damage. Initially, land based
missiles were simply placed underground in heavily reinforced launch silos, usually accompanied by
underground launch facility buildings. However, as missile guidance systems evolved, the accuracy
of enemy missiles was improved, and the need for shock isolation devices became apparent. Early
missile isolators consisted of simple coil springs with fluid dampers. In some cases, the spring-
damper units were used to isolate the missiles themselves and various critical items inside the launch
complex. In other cases, entire structures were base isolated in vertical and horizontal planes.

During the 1960's, it became impossible to provide large enough mechanical springs to provide the

@Seismicisolation
@Seismicisolation
11
optimal isolation, so fluid dampers were converted to liquid-spring dampers, an extremely powerful
yet compact isolation component. In a liquid spring-damper, the operating fluid is compressed and
orificed simultaneously. By selecting special fluids with high compressibility, it was possible to
produce both high spring and damping forces in an extremely small package. Without becoming
too specific (for security considerations), some of the liquid spring-dampers of the late 1980's could
simultaneously provide spring forces of 50 tons and damping forces of 150 tons from a package of
only seven inches in diameter! Operating fluid pressures of up to 50,000 psi were relatively common.
In comparison, a high-powered hunting rifle has peak firing pressures in the 40,000 psi range. Some
of these products for large land based missiles had more than four feet of displacement, with output
forces up to 500 tons.

The successful use of high capacity fluid dampers and liquid spring-dampers on land based missile
facilities led to additional applications on shipboard and submarine missiles and related equipment
items. By the end of the Cold War, a typical U.S. Naval warship would have more than 1,000 fluid
damping devices installed on its missiles and primary electronics systems. These devices range from
1 ton to 50 tons of output force.

During the 1990's, the end of the Cold War combined with the political and economic climate caused
a dramatic downsizing of U.S. defense capabilities. At the same time, security restrictions on the sale
and commercial use of Cold War era technology had been greatly relaxed.

After the Cold War - Transition of Defense Technology to the


Private Sector
Defense firms found very few new opportunities in their traditional markets when the Cold War ended.
Some firms grew smaller, or maintained sales levels by oftentimes painful mergers or consolidations.
Relatively few firms were able to transition their technology to the commercial marketplace. In 1987,
Taylor Devices, Inc., began to look for commercial outlets for its defense products.

Taylor Devices’ defense expertise involved the design and manufacture of large, fluid damping devices
for protection of missiles, electronics systems, and large structures against the effects of weapons
explosion. The company’s staff elected to pursue commercial applications related to seismic and high
wind protection of structures. The damper style selected dated from the 1970's, and was developed
on a sole-source basis by the firm for use on the U.S. Air Force’s MX Ballistic Missile, and the U.S.
Navy’s Tomahawk Cruise Missile. On the latter program, the company has produced more than
29,000 fluid damping devices for use on the shipboard launched Tomahawk.

Early on, it was decided to pursue joint research on fluid damped building and bridge structures with
the National Center for Earthquake Engineering Research (NCEER). NCEER was conveniently located
on the campus of the State University of New York at Buffalo, just a short distance from Taylor Devices’
facilities. The research involved taking existing military production fluid damping devices, and simply
installing them onto scaled models of civil engineering structures, as supplemental components. The
structures were then subjected to seismic transient testing on the University’s large seismic shake
table. All tests proved excellent, with dramatic reductions of stress and deflection occurring with
added fluid damping in the 15-40% of critical range.

In general, it was found that adding 20% critical damping to a structure will triple its earthquake
resistance, without increasing stress or deflection. Numerous reports were published by NCEER and
the University, documenting the improvements obtainable with fluid dampers. The U.S. Department
of Defense proved very cooperative in allowing Taylor Devices to disclose the origins and applicable
design concepts for the damping devices used in the research.

@Seismicisolation
@Seismicisolation
12
For example, steel building structures were tested with fluid dampers being currently produced for
the B-2 Stealth Bomber. Concrete building structures were tested using Tomahawk missile dampers.
Bridge structures were tested with dampers from the CIA’s famed Glomar Explorer Research Vessel.
Other bridge structures were fitted with spring-damper units from submarine based torpedoes.

It became evident that there were no barriers towards commercial implementation of Taylor’s
damping products, and by 1993, an order was received for 186 dampers to be used on all five
buildings of the new Arrowhead Regional Medical Center in Colton, California. Specifications for
these dampers are provided in Figure 2.9, and a photo of a completed damper follows in Figure 2.10.

More than 600 additional building and bridge projects followed the Arrowhead Medical Center order
over the subsequent 20 years. The transition of fluid dampers from military to civilian has proven to
be the quintessential example of literally “turning swords into plowshares.”

Damper Specifications
San Bernardino County Medical Center
Displacement = 48 in.

Maximum Damping Force = 320,000lb

Maximum Operating Velocity = 60 in./sec.

Power Dissipation = 2,170,000 watts

Length = 14.5 ft. extended

Diameter = 14 in.

Weight = 3,000 lb.

Quantity Required = 186

Table 2.1
San Bernardino County Medical Center Damper Specification

@Seismicisolation
@Seismicisolation
13
Figure 2.1
Photograph of Completed Damper

REFERENCES
1. Hogg, I.V., 1971, The Guns 1914-1918, Ballantine Books Inc., New York, New York.

@Seismicisolation
@Seismicisolation
14
3 Design Description of Dampers

Basic Design Description


The essential design elements of a fluid damper are relatively few. However, the detailing of these
elements varies greatly and can, in some cases, become both difficult and complex. Figure 3.1 depicts
a typical fluid damper and its parts.

Figure 3.1
Typical Fluid Damper & Parts

The Damper shown in Figure 3.1 is shown in its mid-stroke position. The main pressure chamber is
referred to as the Cylinder (not labeled). It is completely full of Fluid including the volumes on both
sides of the Piston Head. The Piston Rod is attached to the Piston Head. On the left end of the Piston
Rod is a Clevis for attachment to the structure. As the damper reciprocates during a dynamic event,
this Clevis, the Piston Rod and the Piston Head move as one component. All the other parts remain
stationary.

As the Piston Head moves, the Fluid on either side of it is forced through orifices in the Piston Head.
On the left side of the Cylinder is a Cap and Seal to encapsulate the Fluid against static and dynamic
pressure. On the right side of the Cylinder is another Cap and Seal. As one end of the Piston Rod
moves into the Cylinder, the other moves out, thus maintaining conservation of volume of the Fluid
without a build-up of static pressure. Both Clevises are typically outfitted with spherical bearings to
allow some level of mis-alignment with the surrounding structure.

Force Rating
Taylor Devices Dampers are available in force ratings up to 1800 KIP (8000 kN). The design of the
structural components of these dampers provides a safety margin to yield of approximately 2:1. All
components of the dampers, including the pressure vessel, have been analyzed for strength using

@Seismicisolation
@Seismicisolation
15
modern-day finite element analysis. Actual damper stress reports are available upon request based
on actual project parameters.

Typical seismic dampers require a velocity exponent of less than 1.0 (see equation 4.1.) This means
that although the safety margin is 2:1 based on force, the actual safety margin based on velocity is
much higher.

Power Rating
Where applicable, damper designs are evaluated for power transmitting capabilities. Oftentimes this
becomes a design driver for wind damper applications whereby substantial power is absorbed by the
damper for extended periods of time as would be experienced during wind storms. The short-term
power capabilities of dampers (during earthquakes) is typically many times higher.

Taylor Devices uses a proprietary power analysis software tool that accurately predicts internal and
external surface temperature versus time with given power input scenarios. For each application,
the power is evaluated to be sure that the damper would be fully capable of absorbing short-term
and long-term (continuous) input without experiencing any degradation in function due to increased
temperatures.

Fire Rating
Taylor Devices dampers have been evaluated for time-based fire rating to be sure that they would be
operable during temporary exposure to fire. These dampers are robust hydraulic devices containing
thick-walled alloy steel cylinders. The components that provide the necessary damping function are
not sensitive to short-term exposure to high heat. Only features such as elastomeric bellows and
paint may be susceptible to damage from short-term exposure.

Fire ratings have been established by using Heisler chart information for temperature-time history of
cylinders exposed to fire at approximately 1900°F (1020°C). A conservative approach was used that
neglected the insulative effects of the working fluid of the damper and therefore assumed that the
damper was a solid steel bar. The failure mode of a damper during exposure to fire would be melting
of the seals, thereby releasing the non-toxic, non-flammable fluid to the surrounding area. No risk of
explosion exists since the melting of the seals would release any high-pressure build-up prior to there
being any high stress in the pressure vessel.

The results of the fire rating analysis of Taylor Devices dampers provides a typical fire rating well in
excess of ½ hour for all models, and even greater than 1 hour for medium to large standard dampers.
This means that even being exposed to direct flame at 1900°F, Taylor Devices dampers will operate
properly for at least ½ hour. It is anticipated that if any area of a structure was exposed to fire for this
period of time that the structure would sustain catastrophic damage. Therefore, damage to dampers
due to exposure to fire should not be regarded as a primary concern.

@Seismicisolation
@Seismicisolation
16
4 Damper Output Characteristics
and Unique Benefits

Fluid dampers operate by providing a resisting force only when moving. They do not add stiffness to
a structure, and they do not carry any static load. However, stiffness can be added to a damper upon
request. A fluid damper consists of a piston moving back and forth through a viscous fluid thereby
generating high pressure. This piston has custom designed orifices that produce an optimized
relationship that produces this pressure (force) that varies with velocity. The greater the velocity,
the greater the resisting force that is produced. This relationship is typically characterized by the
following equation:

Damping Force = Damping Constant (C) x Velocity α α (4.1)

where α is referred to as the damping exponent. This exponent can typically be set to anywhere
between 0.2 and 2.0 depending on the specific application. In most cases, α is an exponent having a
specified value in the range of 0.3 to 1.0. Values of α, which have proven to be most popular, are in the
range of 0.3 to 0.5 for present-day building designs with seismic inputs. Bridge applications in U.S.
Seismic Zones 3 and 4 use similar damping exponent values. Wind damping applications presently are
most popular with exponents in the range of 0.5 to 1.0, with the lower values being used in structures
driven by both wind and seismic inputs. Fluid dampers for use in tuned mass dampers use exponents
as high as 2.0.

Because the fluid damper only produces a resisting force only while moving and does not provide
a restoring (spring) force, energy is absorbed into the damper’s fluid and converted to heat. This
absorbed energy is simply the summation of the damping force multiplied by the deflection. Because
dampers can be designed to generate greater than 10,000 psi of pressure, the force and therefore the
absorbed energy can be relatively high.

It is this absorbed energy that significantly reduces the necessity of the structural portions of the
building to absorb that energy. Although this seems like a simple concept, the benefits are often
not fully realized. This is because of the fact that it is not only important how fluid dampers absorb
energy, but also when they absorb energy.

Imagine a structure moving due to a transient input. A significant response of that structure will
be along its dominant natural frequency as a sine wave. As the structure moves through its initial
position, the deflection stress at this moment in time is zero. It is also at this moment in time that
the structure is moving with greatest velocity and therefore the damper is reacting with its greatest
force. Conversely, as the structure reaches its peak deflection and stress farthest away from its initial
position, the velocity reduces to zero and therefore the damper is reacting with zero force at that
moment in time. Because of this, utilizing fluid dampers actually reduces the amount of stiffness
(steel or concrete) that a structure must have to increase its ability to withstand earthquakes. This is
an efficient means to improve structural performance in terms of both cost and weight.

@Seismicisolation
@Seismicisolation
17
The benefit of fluid dampers to be out of phase with the structural deflection stress is not the case
with elements that increase stiffness or elements that are not velocity sensitive such as friction
dampers or buckling restrained braces (BRB’s).

Returning to Equation (4.1) above, if a comparison is made with sinusoidal input to a fluid damper of
the energy absorbed for each cycle to the damping exponent, it is demonstrated that a fluid damper
with a lower damping exponent absorbs more energy per cycle than one with a higher exponent. This
is illustrated in Figure 4.1 below.

Referring to the red curve in Figure 4.1 where the damping exponent is 0.4, it is demonstrated that
the energy under the force versus displacement curve (hysteresis) is higher than the blue curve
where the damping exponent is 1.0 (i.e. linear damping). However, if a damping exponent of less than
0.4 is plotted, a point of diminishing returns on this effect is realized. Note the very small difference
between the 0.4 line in red and the 0.3 line in black. Additionally, as stated above, since the damping
force is out of phase with the deflection stresses, it is important that this effect is not compromised
by an exponent that is too low. Therefore, a damping exponent of approximately 0.3 oftentimes
provides an optimal combination of maintaining a high amount of energy absorbed per cycle and at
the same time minimizing the stress of adjacent structural members.

Figure 4.1
Comparison of Energy Absorbed with Varying Damping Exponents with Sinusoidal Input

@Seismicisolation
@Seismicisolation
18
Damper output characteristics for a given application will be determined by analysis as outlined in
Chapter 7 of this manual. As stated above, damping exponent values are available between 0.20
and 2.0. However, available damping coefficients are virtually unlimited. That is, fluid dampers can
be manufactured with very high and very low damping coefficients. Usually, if a desired damping
coefficient is extremely high, the stiffness of the surrounding structure becomes a limiting factor
since the relative motion at the damper ends becomes very small if the surrounding stiffness is too
low to transmit the motion into the damper. If a desired damping coefficient is extremely small, the
amount of energy absorbed by the damper is also small, and therefore the benefit provided by the
damper is limited.

Chapter 12 of this manual provides series of charts with available damper sizes and their dimensional
information. Dampers are available with any value of C and α as discussed above. However, a small
selection of damper equations is listed in Table 4.1 below to choose from within the force limitations
of each damper series.

Rated
Suggested C Values in KIP - Sec / Inch where F = C x (V)^0.30
Force (KIP)
55 33.9 27.6 24.4 22.4

110 67.9 55.1 48.8 44.8

165 102.0 83.0 73.0 67.0

220 136.0 110.0 98.0 90.0

330 300.0 165.0 146.0 134.0

440 270.0 220.0 195.0 180.0

715 440.0 360.0 317.0 290.0

900 555.0 450.0 400.0 365.0

1450 895.0 727.0 644.0 590.0

1800 1110.0 900.0 800.0 733.0

Table 4.1
Suggested C Values for Taylor Devices’ Standard Dampers

Note that this table provides damping equations with a damping exponent of 0.3, for reasons stated
above. These equations result in a damper force equaling the rated force at velocities of 5, 10, 15, and
20 in/sec velocity.

@Seismicisolation
@Seismicisolation
19
@Seismicisolation
@Seismicisolation
20
5 Generalized Effects of Adding
Fluid Dampers to a Structure

Damping is one of many different methods that have been proposed for allowing a structure to
achieve optimal performance when it is subjected to seismic, wind storm or other types of transient
shock and vibration disturbances. Conventional approach would dictate that the structure must
passively attenuate or dissipate the effects of transient inputs through a combination of strength,
flexibility, deformability and energy absorption. The level of damping in a conventional structure is
very low, and hence the amount of energy dissipated during transient disturbances is also very low.
During strong motions, such as earthquakes, conventional structures usually deform well beyond
their elastic limits, and remain intact only due to their ability to inelastically deform. Therefore, most
of the energy dissipated is absorbed by the structure itself through localized damage.

The concept of added-on dampers within a structure assumes that some of the energy input to the
structure from a transient will be absorbed, not by the structure itself, but rather by supplemental
damping elements. An idealized supplemental damper would be of a form such that the force being
produced by the damper is of such a magnitude and occurs at such a time that the damper forces do
not increase overall stress in the structure. Properly implemented, an ideal damper should be able to
simultaneously reduce both stress and deflection in the structure.

Figure 5.1 depicts earthquake spectra capacity and demand curves for a sample building with 20%,
30% and 40% damped demand curves. This figure is reproduced from FEMA 274 [2] and assumes
linear or viscous damping elements are used.

The effects of added supplemental damping in a structure subjected to earthquake transients is


depicted in the test results provided in Figures 5.2 and 5.3. The tested structure was a single story,
steel building frame, using steel moment frame connections. Figure 5.3 shows the response of the
test structure under a scaled input of 33% of the 1940 El Centro earthquake. Note that a small
hysteresis loop is apparent in Figure 5.2, revealing that the test structure was at the onset of yield.
Structural damping in the frame was in the 2% range. In comparison, Figure 5.3 is the same structure
with 20% added damping, obtained by the addition of two small linear fluid dampers installed as
diagonal brace elements. The large energy dissipation of added damping is readily apparent in the
football shaped damping curve superimposed over the structural spring rate curve. Note also that
the input in Figure 5.3 is the full 100% El Centro earthquake, yet base shear and deflection of the
frame are virtually unchanged from the undamped case of Figure 5.2. Thus, in this case, the addition
of 20% added linear damping to the structure increased its earthquake resistance by a factor of 3,
compared to that of the same structure without added damping. Most importantly, this threefold
performance improvement was obtained without increasing the stress or deflection in the structure.
In fact, it is this tremendous performance improvement that has caused much of the interest in fluid
dampers for structural engineering use. To paraphrase the body builders saying, this is a case where
dampers provide a big gain, without any pain!

@Seismicisolation
@Seismicisolation
21
Figure 5.1
Spectral capacity and demand curves for
rehabilitated one-story building

Figure 5.2
One-story structure, no dampers,
El Centro 33.3%

Figure 5.3
One-story structure, two
dampers, El Centro 100%

@Seismicisolation
@Seismicisolation
22
The test results from Figures 5.2 and 5.3 used the 1940 El Centro earthquake transient as a test
input. When these results were first obtained, they included tests showing similar performance gains
with other notable earthquakes for which transient records were available. Nevertheless, questions
have arisen in the ensuing years as to whether fluid dampers would be functional with other inputs,
including actual earthquakes such as the 1994 Northridge, California and 1995 Kobe, Japan events,
plus hypothetical inputs such as Aa big, purely impulsive quake or a slow rolling sine wave quake. In
addition, potential customers with wind storm inputs wanted to know if seismic dampers worked in
wind, and Government customers wanted to know if damage from terrorist attacks against buildings
would be reduced by dampers. The actual question being raised was simply: Fluid dampers appear to
be a useful engineering component. Are they truly useful for all types of shock and vibration inputs?
The answer is a definite yes, and it is relatively easy to demonstrate this by considering generalized
qualities of a transient pulse.

The first and most important parameter of a transient is the peak translational velocity. The peak
velocity is of primary importance because this determines the peak amount of energy that must be
managed by the structural system. This velocity can be achieved by either a small acceleration over a
long time period, or by a large acceleration over a short period. Thus, the maximum acceleration rate
of the pulse is the second most important parameter of a transient, since the structure and the fluid
dampers must be designed to accommodate the acceleration without being damaged by impulsive
loadings. Figure 5.4 provides tabular data for maximum velocities and accelerations for catastrophic
inputs. The least important parameters of the transient are those related to the actual shape of the
various portions of the pulse. This is simply because no two discrete transients can be expected to be
identical, these events being chaotic by their very nature. If one considers how a damped structure
behaves under transients having a given maximum translational velocity and maximum acceleration
then, in reality, only two simple extreme cases need to be considered.

Case One: The structure is excited by a step function, with acceleration equal to the maximum
acceleration expected, for a time duration such that maximum translational velocity is obtained.

Case Two: The structure is excited by a forced sine wave at the frequency of the structure’s first
resonant mode, with input amplitude increased until the maximum specified acceleration or velocity
is achieved.

An example of structural response to the first case, the impulsive input, is provided in Figure 5.5,
for both the undamped and fluid damped condition. The response in this case assumed infinite
acceleration, with velocity stepping from zero to maximum value instantaneously, and an elastic
structure. It is readily apparent that the fluid damped structure experiences substantially less force
and deflection than the undamped structure, even though each structure is storing or absorbing
equal amounts of impulse energy.

An example of the second case is provided in Figure 5.6, and depicts the magnification factor on
input amplitude for a system subjected to forced harmonic excitations with linear fluid damping. The
condition of resonance is obtained at a frequency ratio of 1.0, and shows the tremendous benefits of
fluid damping. The equation for magnification at resonance is:

1
magnification factor = 2g

where ζ = the damping ratio cccr

@Seismicisolation
@Seismicisolation
23
Tabular Data for Maximum Velocities and Accelerations
Peak Acceleration Peak Velocity

Northridge Earthquake .9 G 51 In/Sec.

Kobe Earthquake .8 G 35 In/Sec.

Ship, Moored Mine 25. G+ 90 In/Sec.+

Missile Silo, Nuclear Air Burst 80. G+ 450 In/Sec.+

Submarine, Nuclear Depth Charge 600. G+ 500 In/Sec.+

Figure 5.4
Catastrophic Transients

Figure 5.5
Response to impulsive inputs

Of particular note is that for a typical building with 2% damping, the magnification factor at resonance
is 25 to 1. This number reduces to a much more manageable value of only 2 to 1 at 25% damping. It
is of value to the engineer to note that virtually no structure is built with the safety factor of 25 to 1
necessary to accommodate the 2% damped resonant response. In comparison, most structures have
sufficient safety factors to accept the 2 to 1 magnification for the 25% damped structure subjected
to forced resonance.

From these examples, it is relatively easy to understand that fluid damping will always improve the
response of a structure, under any expected transient.

Three Generic Types of Dampers and How Each of Them Affects a Structure:

Fluid dampers have the unique ability to simultaneously reduce both stress and deflection within a
structure subjected to a transient. This is because a fluid damper varies its force only with velocity,
which provides a response that is inherently out-of-phase with stresses due to flexing of the
structure. Other dampers can normally be classified as either hysteretic, where a fixed damping force
is generated under any deflection, or as visco-elastic, where a damper behaves as a complex spring
and damper combination. In the latter case, force may be a displacement and velocity dependent
parameter. Figure 5.7 provides representative outputs from sine wave excitation of these three
damper types. Inclusive in these non-fluid damper types are yielding elements, friction devices,
plastic hinges, friction slides, bonded rubber, molded rubber, and shaped rubber. None of these
other devices have an out-of-phase response to structural flexural stresses. This is simply because

@Seismicisolation
@Seismicisolation
24
Figure 5.6
Magnification factor for
forced harmonic excitation

the outputs of these devices are dependent upon parameters other than, or in addition to, velocity.
Hence, all of these other types of dampers will decrease deflection in a structure at the same time
they are increasing column stress. The out-of-phase response that is unique to fluid dampers can be
easily understood by considering a building shaking laterally back and forth during a seismic event or
a windstorm.

Column stress is at a peak when the building has flexed a maximum amount from its normal position.
This is also the point at which the flexed columns reverse direction to move back in the opposite
direction. If we add a fluid damper to the building, damping force will reduce to zero at this point
of maximum deflection. This is because the damper stroking velocity goes to zero as the columns
reverse direction. As the building flexes back in the opposite direction, maximum damper force occurs
at maximum velocity, which occurs when the column flexes through its normal, upright position. This
is also the point where column stresses are at a minimum. It is this out-of-phase response that is the
most desirable design aspect of fluid viscous damping.

@Seismicisolation
@Seismicisolation
25
Figure 5.7 Output of the Three Generic Damper Types

@Seismicisolation
@Seismicisolation
26
6 Damper Mounting Arrangements

Using Distributed Damping throughout a structure is a direct method of applying damping to a


structure or mechanism, generally placing the dampers at multiple levels of the structure. The main
benefit of distributed damping is that the damping system captures and absorbs the energy at its
source throughout the structure, instead of applying damping at a secondary system, elsewhere, or
at one location in the structure to later collect and attempt to dampen motions as they are amplified
throughout the structure. Of primary benefit with distributed damping is that the (direct) damping
can absorb energy in any and all frequencies of input vibration, instead of being confined to a tuned
or “dialed-in” frequency. This is highly beneficial with wind or earthquake motion input which do not
necessarily shake the structure at only its fundamental period of motion.

Many methods exist to implement distributed damping in a structure, the typical underlying concept
is to connect the moving masses (floor levels) with dampers so that as they move or deflect relative
to one-another in a shearing-type motion, the dampers capture this motion and resist in both tension
and compression directions with an opposing force. This concept works well in typical moment frame,
shear wall, or braced frame office or residential type buildings, and can be applied to short, medium
and tall structures. It is noted that all of these systems are passive, meaning that no external power
is needed to make the dampers function. The dampers simply react at any time they are deflected.
Some of the implementation methods include the following:

Chevron frames are depicted in Figure 6.1. In this configuration, the dampers are placed horizontally,
and connected to a frame (chevron) that is intended to be near-rigid with the floor it is connected to.
The advantage with this direct damping orientation is that the horizontal flexibility of the structure
injects this full movement directly into the horizontal orientation of the damper. However, a small
amount of motion can be lost due to the constraints of the attainable stiffness of an economical
chevron frame.

@Seismicisolation
@Seismicisolation
27
Figure 6.1
Dampers in Chevron Braced Frames

Figure 6.2
Dampers in Diagonal Braced Frames

Dampers in Diagonal bracing schemes are depicted in Figure 6.2. In this orientation, the horizontal
movement of the structure only allows an angular component of the full deflection to go into the
damper, but thence takes this motion directly to the next floor level, straight through a strong tension/
compression member. Often this diagonal bracing scheme is considered the most basic, or simplistic
method to apply distributed damping in a structure.

Toggle frames as shown in Figure 6.3 can be used as a mechanism to amplify deflections into the
damper in otherwise stiff, or tiny deflection situations, creating a more efficient damping system.
Toggle Frames utilize a bent-brace mechanism theory to capture deflections in one plane and
translate the deflections into another plane and therefore provide very efficient damping, but these
toggle frames do require an intricately designed and manufactured custom mechanism/system, in
order to perform properly. These toggle frames are not only a patented system (Refer to US Patents
# 5870863 and 5934028), but they also need to be very carefully detailed by an authorized designer/
manufacturer.

@Seismicisolation
@Seismicisolation
28
Figure 6.3
Dampers in Toggle Braced frames

So-called Mega Braces shown in Figure 6.4 can be used to capture deflections over multiple floor
levels and collect the larger motion from these levels and pass that motion through the energy
absorbing damper connected to major structural nodes. This concept is similar to diagonal bracing,
but over a much longer span.

Figure 6.4
Dampers in Mega-Brace Frames

@Seismicisolation
@Seismicisolation
29
Additionally, an outrigger solution to apply damping to taller, more slender building systems can be
used where it is determined that the gross motion of the structure does not fall into the traditional
shearing-type movement pattern, but exhibits more of an overall tension/compression on the
opposing outer columns of the building. Often outrigger damping can be accomplished by creating a
rigid level near the top of a building that moves with the core and connecting dampers between the
rigid level and the outer columns of the building. This useful system is shown in Figures 6.5 and 6.6.

Figure 6.5
Dampers in Outrigger Systems (a)

Figure 6.6
Dampers in Outrigger Systems (b)

@Seismicisolation
@Seismicisolation
30
Dampers can also be distributed in base isolation systems, as depicted in Figure 6.7, where the damper
is used to augment the vertically supporting isolators, and most often provide viscous (velocity
dependent) damping to significantly improve the performance and usefulness of the isolators.

Figure 6.7
Dampers in Base-Isolation Systems

The Open Space Damper System shown in Figure 6.8 is somewhat similar to the toggle system in
that the system uses motion amplifiers and mechanisms, but in this system, the idea is to push the
damping elements to the perimeter of the frame such that the space inside the frame is opened-up for
windows or doorways, hence the name “open space” damping system. In this system, there is a (near)
vertical pusher bar connecting a cam-rotating mechanism that in-turn is connected to the beam and
to a damper, such that as the top of the frame moves laterally, the damper deflects and dissipates
energy. Open Space Dampers provide very efficient damping, but do require a carefully oriented and
analyzed system, to function properly. These Open Space Damper Frames are a patented system
under US Patent # 9,580,924, and they need to be very carefully detailed by an authorized designer/
manufacturer.

Figure 6.8
Dampers in Open Space Frames

@Seismicisolation
@Seismicisolation
31
The Negative Stiffness concept is a unique new passive damping concept in structural engineering
that takes a still building or frame, and creates a “softening” effect for that frame. The system uses
a series of gapping elements with a preloaded spring and dampers to develop the reactions needed.
The technique here is to allow the structure to have its initial stiffness for wind or other constant
lateral inputs, but for dynamic inputs, the negative, or call it “anti-stiffness“ system actually gives
the structure a little push, using a passive spring system, then restricts that motion with damping.
The result allows damping to more-effectively be used in a stiff frame system, without cutting, or
softening the frame, to let damping do its job. The Negative Stiffness system is shown in Figure 6.9
below.

Figure 6.9
Dampers in Negative Stiffness Frames

Numerous other techniques can be applied to implement distributed damping by using similar
principles and/or different orientations, or structural systems and mechanisms to capture structural
movement and inject that movement into the distributed dampers. Additional techniques and
concepts are currently undergoing research and development, and therefore are not elaborated
upon in this document.

@Seismicisolation
@Seismicisolation
32
7
Design and Analysis of
Building Structures with
Fluid Viscous Dampers

SCOPE
Fluid Viscous Dampers (FVDs) serve to protect new and existing structures during hazardous
seismic events and provide an economical solution for resilient design. Adoption of FVDs for
applications in building and bridge structures has become prevalent in countries like Japan and
Taiwan, however FVDs are under-utilized in many other high seismic regions across the world. One
of the factors contributing to this under-utilization is the lack of familiarity of many structural
engineers with damping devices and the relative ease in which they can be designed and
implemented in commercial structural engineering software.

This document aims to educate structural engineers on the use of FVDs and assist in promulgating
the application of FVDs for seismic protection of buildings and bridges. The discussions in this guide
are intended to be informative and are envisioned to demonstrate general Modeling and design
processes of building structures with FVDs. The reader is referred to industry documented
literature, such as the SEAOC: IBC 2012 Structural/Seismic Design Manual - Volume 5 and FEMA P-
1051: 2015 NEHRP Recommended Seismic Provisions - Design Examples for a more detailed
description on the design application of supplemental damping for building structures.

@Seismicisolation
@Seismicisolation
33
PART 1: CREATING THE ANALYTICAL MODEL W/ FVDs FOR NLRHA

OVERVIEW
Implementation of FVDs in the commercial program ETABS (CSI 2017) is demonstrated by providing
a step-by-step procedure for generating the analytical building model. The selected building model
is a generic example and is not representative of any particular structure.

The following is noted:


• The guide demonstrates implementation of FVDs using ETABS; the same approach is applicable
to other CSI Inc. software such as SAP2000 and other commercial software programs.
• The guide demonstrates modeling and design with building structures; however, a similar
approach can be used for bridges, and other types of structures.
• The guide focuses on modeling of FVDs for Nonlinear Response History Analysis (NLRHA),
although dampers can be modeled within ETABS for Response Spectrum Analysis (RSA) as well.

CREATING THE ETABS MODEL

Summary of Building Properties


The key properties of the example building are summarized in Table 7.1.1. This is a rectangular
reinforced concrete moment frame structure.

Property Value
Bays in the X direction 7 @ 24 ft
Bays in the Y direction 3 @ 24 ft
Stories 3
Story Height 12 ft
Column 24x24 in square
Beam 14x24 in rectangular
Slab 8 in NWC
Concrete Compressive Strength 4 ksi
Grade of Steel Reinforcement 60 ksi
SDS = 1.50 g

Site Seismicity SD1 = 0.6 g


Site class = D
Risk category = II

Table 7.1.1
Key Properties of the Example Building

@Seismicisolation
@Seismicisolation
34
Creating a New Building Model

It is assumed the user is familiar with the basics of generating an ETABS model of a building. However,
brief discussion on the topic is presented here.
• Under the tab file, open new and initialize model (Figure 7.1.1). In this figure, the user selects
the units, steel and concrete design code and databases used to select the structural members
from the pull-down tabs.
• Select grid only option and specify the number of bays in X and Y direction, bay width in each
direction, number of stories and story height (Figure 7.1.2); press OK and in plan view, the
program will display the grids (Figure 7.1.3).

Figure 7.1.1
Model Initialization

Figure 7.1.2
Selection of Templates Using Grids Only

@Seismicisolation
@Seismicisolation
35
Figure 7.1.3
Grid Pattern in Plan-View

Defining Material Properties


• Click the Define > Material Properties command to access and define the material form
(Figure 7.1.4)

Figure 7.1.4
Defining Material Properties

@Seismicisolation
@Seismicisolation
36
Defining Frame (Beam-Column) and Shell (Slab) Objects

• Use Define>Section Properties>Frame to open the dialog box (Figure 7.1.5)


• Click on Frame sections box, and then add new property /section type concrete rectangular,
then enter values as required for section properties (Figure 7.1.6). Under Property Modifiers
click on Modify/Show Modifiers, to modify section properties if required. In the same menu,
click on Modify/Show Rebar to assign reinforcement data (Figure 7.1.7)
• Use Define >Section Properties>Slab Sections, to define slab objects for the floor (Figure 7.1.8)

Figure 7.1.5
Define Selection Menu

Figure 7.1.6
Defining and Modifying Column Sectional Properties

@Seismicisolation
@Seismicisolation
37
Figure 7.1.7
Defining and Modifying Column Sectional Properties

Figure 7.1.8
Defining and Modifying Slab Section Properties

@Seismicisolation
@Seismicisolation
38
Draw Frame and Slab Objects

• Use the Draw > Draw Beam/Column/Brace Objects command or click one of the five buttons
shown in this topic to draw frame objects. When the menu command is used, a menu of five
subcommands displays (Figure 7.1.9). Once a property is selected, the user can then graphically
draw the frame element by clicking on the correct geometrical locations
• Use the Draw > Draw Floor/Wall Objects command or click one of the five buttons shown in
this topic to draw floor objects. When the menu command is used, a menu of five subcommands
displays (Figure 7.1.10). Once a property is selected, the user can then graphically draw the
floor/ wall element by clicking on the correct geometrical locations.

Figure 7.1.9
Drawing of Frame Objects

Figure 7.1.10
Drawing of Floor Objects

@Seismicisolation
@Seismicisolation
39
Display Model
Once all beams, columns, and floors are drawn, assign property fixity to the base of the columns
and the model can be displayed in 3-D (Figure 7.1.11) in plan (Figure 7.1.12), or elevation (Figure
7.1.13 and Figure 7.1.14). Once the floor plan or elevation videos have been defined for one of the
stories or gridlines, the floor plans can be replicated (Z direction) to other stories and the elevations
can be replicated (X and Y directions) along the building gridlines.

At this step, a three-dimensional model of the building has been completed. The model will next be
updated by addition of FVDs.

Figure 7.1.11
3D Model After Replicating Plans

@Seismicisolation
@Seismicisolation
40
Figure 7.1.12
Plans for Beam and Column Objects

Figure 7.1.13
Transverse (Y) Elevation

Figure 7.1.14
Longitudinal (X) Elevation

@Seismicisolation
@Seismicisolation
41
DESIGN AND MODELING OF FLUID VISCOUS DAMPERS

Fluid Viscous Dampers


FVDs are velocity dependent devices that are used in buildings to dissipate seismic energy through
viscous damping. In general, the higher the viscous damping in a building, the lower the responses
(forces, displacements, stresses, and drift ratios). In typical buildings, an inherent damping ratio (ζ)
of 2% to 5% of critical is expected. FVDs can be distributed along the height of the building to
provide a substantially higher amount of viscous damping. The following is noted:
• FVDs are frequency independent devices without a stiffness component. So unlike tuned devices,
no tuning of FVDs to any particular frequency is required.
• FVDs are velocity dependent devices and unlike some other devices, there is no significant shift
in the building period (T) when added to structures, and the output forces generated by FVDs are
primarily out-of-phase with the maximum strains on the building structure.
• FVDs are classified as passive devices. So, no external power source is required to activate them.
Dampers are activated when there is relative motion between the two ends of the unit and seismic
energy is converted to heat and safely dissipated into the atmosphere.
• FVDs do not require regular maintenance and for typical applications, there is no degradation of
the performance with use or typical temperature variations.
• FVDs need not be placed in every level of a building and are often placed in strategic locations to
maximum efficiency of the dampers, or to meet other project constraints.

Mathematical Modeling of Fluid Viscous Dampers


The behavior of a fluid viscous damper is idealized as a pure dashpot as shown in the constitutive
equation below:
Damping Force (F) = Damping Constant (C) x Velocity (V) α (Eq. 7.1.1)

Equation 7.1.1 provides the relationship between the damper output force and velocity, where C
and α (alpha) are the damping constant and velocity exponent, respectively. An alpha of 1.0
represents linear dampers, whereas values other than 1.0 indicate nonlinear dampers. See the figure
below for different behavior of an FVD with constant C value and varied alpha. Specifications for
alpha typically range from 0.3 to 1.0; in general, the lower the exponent the more efficient the
viscous damping for seismic energy dissipation.

@Seismicisolation
@Seismicisolation
42
Figure 7.1.15
Damper Force-Velocity Relation

Candidate Building Types


FVDs are activated by the relative motion at the two ends of the damper and work best to mitigate
dynamic response of more flexible buildings like steel or concrete special moment frames; base-
isolation systems are also significantly improved with additional damping.
FVDs are ideal candidates to address building irregularities. For buildings, with a soft story,
placement of dampers at only the bottom level significantly reduce the soft story response, without
increasing structure period, and therefore total base shear. For buildings with plan irregularities,
placement of dampers opposite the more rigid side of the building can be used to reduce the
torsional amplification of the structure during motion. FVDs are efficient in resolving building
separation issues as well.
Dampers have been used in stiffer buildings, but oversized dampers or motion amplification
configurations would be required to amplify the small relative motion at the two ends of the
dampers. For this type of installation, “no play” connections are also required to ensure that the
dampers are fully engaged.

Damping Configuration
Like any lateral force resisting system, it is important to ensure that dampers are placed in a
configuration that does not introduce asymmetry to the structure. The most efficient placement of
dampers would be equivalently about the building’s center of mass to control any torsional motion
of the building; consequently, the placement of dampers will be best along the perimeter of typical
structures.

@Seismicisolation
@Seismicisolation
43
In most applications at least two dampers are placed in each direction and on each side of building’s
center of mass for system redundancy. For buildings with a larger footprint, more than two dampers
will often be used in order limit the damper force output. Limiting force output from the dampers
can lead to more economical structural member sizes and connections. The exact placement of
FVDs, like bracing elements, are subject to architectural constraints
As building height increases so does the fundamental period; correspondingly inter-story floor
velocities are less, and damping efficiency increases. For low-rise buildings, FVDs are typically
placed at more floor levels than needed for mid and high-rise buildings. FVDs are usually not
required at every floor and are often terminated before the top levels or alternated at different floor
levels.

FVDs can be arranged in many different configurations and some of the most common
configurations include diagonal, double-diagonal, Chevron, and the inverted Chevron. These
configurations are demonstrated in Chapter 6 of this guide. For tall buildings, damped outriggers
and mega dampers spanning many floors can be extremely efficient. In this chapter the Chevron,
diagonal, and double diagonal configurations will be discussed.

Preliminary Damper Sizing


NLRHA is not required to develop conceptual damping system designs. There are several methods
for estimating dynamic response of a structure with supplemental damping using linear or response
spectrum analysis (RSA) procedures. ASCE 7-16 has documented procedures in chapter 18 Section
18.7 Alternative Procedures based on the modal strain energy method. The section’s source
document MCEER Technical Report 00-0010 outlines the procedure for implementation of the
method thoroughly.
Using this method only a modal analysis and assumed damping configuration is required to
determine a conceptual design for desired performance objectives. In general, an estimate of total
equivalent damping for each mode is determined and directly applied to the spectral response of
each mode shape for RSA.

Total equivalent damping of each mode shape can be


approximated using equation 7.1.2 below, where Wj is the work
done by each viscous device in one complete cycle of loading and
Wk is the kinetic energy at each floor level:
∑ 𝑊𝑗
𝜻𝑉𝑖𝑠𝑐𝑜𝑢𝑠 = 4𝜋 ∑ 𝑊𝑘
(𝐸𝑞𝑛. 7.1.2)
When considering linear dampers, the following equation can be
applied, where each variable is as shown in Figure 7.1.16:
2
𝑇 ∑ 𝐶𝑗 𝜙𝑟𝑗 𝑐𝑜𝑠 2 Ѳ𝑗
𝜻𝑣𝑖𝑠𝑐𝑜𝑢𝑠 = (𝐸𝑞𝑛. 7.1.3)
4𝜋 ∑ 𝑚𝑘 𝜙𝑘2

Figure 7.1.16
Idealize 2DOF System with Assumed Damping Configuration and Modal Shape Variables

@Seismicisolation
@Seismicisolation
44
If FVD elements are modeled in ETABS, linear damping coefficients can be provided and ETABS will
use the modal strain energy method to apply total equivalent viscous damping for each mode based
on damper configuration and properties during RSA. These design and analysis procedures are
outside the scope of this document as NLRHA is required for buildings in regions of high seismicity
and provides more consistent and accurate results. However, these procedures are helpful for
preliminary sizing in all projects.

FVD Properties in ETABS


FVDs are velocity-dependent and do not exhibit any stiffness in parallel with damping output. The
analytical model need not include any elastic stiffness from the dampers for dynamic analysis.
Simplified design and analysis procedures have been developed that may utilize “equivalent
stiffness” for dampers, but this stiffness does not reflect the actual behavior of the devices.
The definition of FVDs in ETABS follows the constitutive relationship defined in Equation 7.1.1, plus
a series spring constant, K, as shown in figure 7.1.17 below. The damping properties are based on
the Maxwell model of viscoelasticity having an exponential viscous damper in series with a linear
spring.

k = the series spring constant


c = the damping coefficient
α = the damping exponent
d = the deformation across the spring
k
𝑑̇ = the deformation rate across the damper
c

Total Link Displacement, d = dk + dc (Eqn. 7.1.4)

Total Link Force Output, f = k dk = cḋαc (Eqn. 7.1.5)

Figure 7.1.17
Exponential Damper Link

The series spring constant reflects the elastic flexibility of the damping devices’ fluid column and
connecting mechanisms, Ks, and the extender brace, Kextender, often used to connect the damping
device from one story to another as shown in figure 7.1.18 below.

Figure 7.1.18
Exponential Damper Link Model versus Damper and Extender Brace Installation

@Seismicisolation
@Seismicisolation
45
Theoretically both stiffness components would act in series as shown in the equation below.
1 1 1
= + (𝐸𝑞𝑛. 7.1.6)
𝐾 𝐾𝑠 𝐾𝑒𝑥𝑡𝑒𝑛𝑑𝑒𝑟
In nearly all applications the designer can assume infinite rigidity of the damping device, because
Taylor Devices manufactures each device to meet the designer’s exact specifications for the
constitutive relationship between force and velocity and therefore, K need only be represented as
the axial flexibility of the extender brace.

𝐾 = 𝐾𝑒𝑥𝑡𝑒𝑛𝑑𝑒𝑟 (𝐸𝑞𝑛. 7.1.7)


Extender braces should be sufficiently stiff as to drive the FVD piston rod without significant loss
of motion due to their own axial lengthening or shortening. The relationship between a linear
damper force output and the damper and extender brace properties is derived as show in the
equation below.

1 1 −1
𝑓 = 𝑑[ + ] (𝐸𝑞𝑛. 7.1.8)
𝜔𝐶 𝐾
Figure 7.1.19 below shows two plots for a damper of given damping coefficient and total link
displacement. On the horizontal axis the extender brace stiffness is varied. The vertical axis on the
left is for a plot of the damper force output and the vertical axis on the right is for the total loss in
damping, or total reduction in area of the damper hysteresis loop.

Figure 7.1.19
Damper Link Example with Varied Extender Brace Stiffness

@Seismicisolation
@Seismicisolation
46
The relationship between series spring stiffness and damping efficiency is nonlinear. From this plot
designers can choose an optimal extender brace stiffness for design efficiency.
The following table has Taylor Devices recommended properties for the series spring constant, K,
to be used with different FVD sizes and strokes. When damping is required at smaller strokes, then
K should be increased, and more axial stiffness is required from the extender braces.

TAYLOR DEVICES RATED FORCE STROKE FVD Series Spring, K


MODEL NUMBER (kip) (in) (kip/in)
17120 55 ±3 625
17130 110 ±4 940
17140 165 ±4 1,400
17150 220 ±4 1,875
17160 330 ±4 2,800
17170 440 ±5 3,000
17180 715 ±5 4,800
17190 900 ±5 6,000
17200 1450 ±5 9,750
17210 1800 ±5 12,000

Table 7.1.1:
Taylor Devices FVD Model Recommended Stiffness for Extender Braces

Defining FVDs in ETABS


In ETABS, use the following steps to define the FVD properties:
Click on Define >Link >Link Properties, Link Property Name (Say Damp), Link Type, on drop down
menu select Damper-Exponent Type (Figure 7.1.20)

For Damper-Exponential (Fluid Viscous Dampers), select directional properties U1, and check box
for Non-Linear and click on Modify/Show for U1. (Figure 7.1.21)

Enter the values for Series Spring Stiffness, Damping, and Damping Exponent on Nonlinear
Properties data form (Figure 7.1.22).

For this example, the following properties are used to define the force-velocity relation of the FVD:

1. damping constant, C = 100 kips – (sec/in) α

2. velocity exponent, α = 0.5

3. Series Spring Constant, K = 2000 k/in

@Seismicisolation
@Seismicisolation
47
The linear damper properties can be provided for use with response spectrum analysis, although
ETABS will not consider nonlinear exponents and damper forces will not be provided as output.

The values of mass and weight account for damper mass and can be left at zero or a small value can
be specified to assist in analysis convergence. Typical damper weights would range from several
hundred pounds to several thousand pounds, depending on the damper size needed/used.

Figure 7.1.20
Selecting of Damper Type

Figure 7.1.21
Defining Damper Properties

@Seismicisolation
@Seismicisolation
48
Figure 7.1.22
Defining Damper Nonlinear Properties

Adding Dampers to the Model


• Use the Draw > Draw Links command select draw links (Figure 7.1.23).
• Under submenu, select the damper property that was defined in the previous section (Figure
7.1.24). Note, if the incorrect link type is selected, then change the property under
Assign>Links>Link Properties

Diagonal Dampers
Once the correct FVD property is selected, the user can then graphically draw the link element by
clicking on the correct geometrical locations for the first and second ends of the damper. Ensure
that the link ends are physically connected to nodes in the model.

Figure 7.1.23 Drawing of Links

Figure 7.1.24 Select Damper Properties

@Seismicisolation
@Seismicisolation
49
The damper locations and properties can be displayed by viewing elevation (Figure 7.1.25) or 3D
(Figure 7.1.26) views. Diagonal dampers are shown here. X-dampers are drawn similarly.

Figure 7.1.25
Damper Location in Longitudinal (X) Elevation

Figure 7.1.26
Dampers in 3D Model

Drawing Dampers in Chevron (or Double Diagonal) Configuration


In this section, a Chevron Damper between grid BC and FG is added. The approach is similar to the
diagonal dampers. However, first the midpoint of beams needs to be determined.
• In the snap menu, click on select ends and mid-point (Figure 7.1.27)

The user can then graphically draw the link element by clicking on the correct geometrical locations
for the first and second ends of the damper. One of the damper ends will be located at the middle
of the beam (Figure 7.1.28).

@Seismicisolation
@Seismicisolation
50
Figure 7.1.27
Turn Midpoint and End-Snap On

Figure 7.1.28
Dividing the Frame at Midpoint

Figure 7.1.29
Damper Location Along Longitudinal Elevation

Drawing Horizontal Dampers and Chevron Braces


In this section, horizontal dampers between grid BC and FG are added. The main advantage of this
configuration is that since the dampers are horizontal, they are most effective in resisting the
horizontal component of seismic forces. As a result, smaller dampers can be used. The main
disadvantages of this configurations are:

• Since one end of dampers must move with the lower story, stiff bracing needs to be provided.
• Since the damper is attached eccentrically to the beam-column joint, additional moment is
introduced to the beam or column.
• Detailed drawing for the damper connecting elements must be provided to allow for the free
movement of dampers, and out of plane motion of dampers needs to be restrained.
• The construction cost for the attachment members is higher because of the more complex
configuration.

@Seismicisolation
@Seismicisolation
51
The approach is similar to the diagonal dampers. However, first the attachment point of beams must
first be determined.
• In the edit menu, select replicate and type the distance for the new attachment node (Figure
7.1.30). Next using the same process, replicate the node in the vertical direction. This provides
the one end of the two horizontal dampers in one bay. Replicate these nodes by the length of
damper to obtain the second end of damper (Figure 7.1.31).
• In the property definition menu use material steel and define both the Chevron brace size
(Typically square HSS or WF section) and also the section properties for the stiff connections (See
Defining Material Properties for defining material properties and Defining frame (beam column)
and shell (slab) objects for defining frame sections).
• Draw the frame elements and horizontal dampers (Figure 7.1.31). Refer to Draw frame and slab
objects for a description of drawing frame and Drawing Dampers in chevron Configuration for
drawing damper objects.
• Repeat the same procedure by drawing or replicating dampers in the other bays.

Figure 7. 1.30
Replicating a Node

Figure 7. 1.31
Dividing the Frame at Midpoint

@Seismicisolation
@Seismicisolation
52
The damper locations and properties can be displayed (Figure 7.1.32).

Figure 7.1.32
Damper Location Along Longitudinal Elevation

At this stage the analytical model of the building structure with FVDs is complete. The next step is
to define seismic loading for the structure.

@Seismicisolation
@Seismicisolation
53
SEISMIC LOADING
Building Codes, ASCE 7 and ASCE 41, require NLRHA for design and assessment of building
structures with fluid viscous dampers in high seismic zones. Performing NLRHA using direct-
integration methods can be time-consuming, however there is a much faster NLRHA method that
can accurately predict seismic behavior of buildings structures with fluid viscous dampers.
Fast-nonlinear analysis (FNA) is a modal analysis method that uses modal Ritz vectors to accurately
predict the behavior of a structure under dynamic seismic loading with small amounts of material
nonlinearity and nonlinear FVDs through link objects. This analysis technique is permitted in building
codes and underscores that once dampers are added to a structural system, the demand on the
members is significantly reduced; it is anticipated in most scenarios, the structure will remain
essentially elastic. FNA is used in this design guide and is part of the standard of practice in many
design offices. For more information on FNA refer to the CSI Analysis Reference Manual.
In order to accurately capture the combination of vertical and seismic loads, a preload of the vertical
loads using the ramp function shall be applied before seismic loading. See CSI Analysis Reference
Manual for details on using the ramp function to apply vertical loads appropriately using FNA. FNA
for seismic loading requires definition of several input variables including building mass, modal case,
and seismic acceleration histories.

Mass Source
The user shall specify the total seismic mass to be considered in the NLRHA. The seismic mass
includes the self-weight of the structural elements (beams, columns, braces, slabs, walls, etc.),
superimposed dead load and a portion of live load (if applicable).

Click on Define >Mass Source, add or modify mass sources data if required by clicking on submenu
(Figure 7.1.33). The user also has the option of selecting whether only lateral mass is considered or
if vertical contribution of mass is also included in analysis (for cases when there is a gravity preload
preceding the lateral analysis).

Figure 7.1.33 Defining Mass Source

@Seismicisolation
@Seismicisolation
54
Modal Cases
Click on Define >Modal Cases, to open the modal dialogue box. Click on Modify/Show Case to
open the modal case data (Figure 7.1.34). Next:
• Under subcase type select Ritz
• Check P-Delta as needed
• Under Loads applied, select Ux and Uy (and Uz if necessary) and select all links.
Link elements (in this case FVDs) must be activated to contribute to the dynamic response. This
is achieved in the modal case definition

For the number of modes, select a reasonable number to capture desired mass participation. Usually,
two degrees of freedom (axial force and axial deformation) are required for each damper.
Additionally, sufficient number of modes needs to be selected to capture the response of the
building. In most cases, in addition to the number of modes associated with dampers, selecting 30-
60 modes will be sufficient to capture the building response.

Figure 7.1.34 Modal Case Data

@Seismicisolation
@Seismicisolation
55
Response History Function
Although response spectrum analysis is not used directly in this design guide, the response spectrum
function will be used as part of development of seismic acceleration histories.
• Define>Function>Response Spectrum to open menu (Figure 7.1.35) select ASCE 7-10 and specify
a name.
• Modify/Show Function open sub-menu and input site-specific parameters (Figure 7.1.36).

Figure 7.1.35
Response Spectrum Definition

Figure 7.1.36
Site-Specific Response Spectrum Function

@Seismicisolation
@Seismicisolation
56
Acceleration Histories
On most projects the ground motion suites are selected and scaled to match site-specific seismic
characteristics by either a geotechnical engineer or seismology specialist for the structural
engineer’s final analysis and design report. There are several methods for developing these ground
motion suites and the reader is referred to the ASCE building code specifications and commentary
for more detailed discussion on this subject. For the initial phases of a project a couple ground
motions can be selected and scaled using the spectral matching tool in ETABS.

Input of Code-Compliant Acceleration History


Ground motion suites can be exported from the PEER Ground Motion Database as text files and
directly imported to ETABS.

Define>Function>Time History to open the dialogue box and select From File from the drop-down
menu (Figure 7.1.37).

Select Add New Function click on Browse to locate the file and load it and then complete the format
of the file in the remaining boxes. The option to view file allows the file to be opened to check and
the acceleration vs time is displaced graphically (Figure 7.1.38).

Figure 7.1.37
Defining Time History Functions

@Seismicisolation
@Seismicisolation
57
Figure 7.1.38
Importing a Record from Analysis

Input of a General Acceleration Record


Once a ground motion has been imported, site-specific ground motions can be developed using the
site-specific response spectrum function and the matching application:
• Input response spectrum function as described earlier
• Input a selected acceleration history as described in the above section. During ETABS setup, a
number of recorded acceleration histories are also installed in the folder Time History Functions
(Figure 7.1.39).
• For this example, the two components of the Newhall records were imported.
• In the Define Time History Function menu, specify Matched to Response Spectrum in the pull-
down menu (Figure 7.1.40) and click Add New Function
• The time history match menu opens. In this example, the input target spectrum and ground motion
are used to obtain the matched record. Matching is done in the time domain and default properties
are used. Note the units for both response spectrum (g in this example, and seed acceleration
(cm/s2 in this example). The program engine performs the matching and generates the matched
record (Figure 7.1.41). Note that the matched records have the units of g.

@Seismicisolation
@Seismicisolation
58
Figure 7.1.39
Program Included Acceleration Records

Figure 7.1.40
Selection of Matching Menu

@Seismicisolation
@Seismicisolation
59
Figure 7.1.41
Program Matched Record

This approach was used to match both components of Newhall record for this example. Next the
FNA load cases are generated.

Load Cases
Define>Load Case to open the load case menu. Select add new load case and the load case menu
(Figure 7.1.38) Note that the load case type is Time History and FNA selected from pull down
menus. The two matched records are used as input functions and scaled (g=386 in/sec2). Analysis is
conducted for 20 sec and a time step of 0.01 is used for analysis. The inherent damping is assumed
to equal 5% of critical as it is done for concrete frame structure. The choice of inherent damping
does not affect the response significantly, because much larger effective damping is introduced in
this particular analysis by the viscous dampers. In general, an assumption of 5% inherent damping
is probably higher than reality, and the base model should probably assume 1% to 3% inherent
damping.

@Seismicisolation
@Seismicisolation
60
Figure 7.1.42
Load Case Definition for Seismic Loading

At the conclusion of this step, the definition of the model has been completed and user can start
analysis.

Analysis
ASCE Building codes require design and analysis based on the use of seven sets of two-component
ground motion suites; structural response are based on the average value of the seven ground
motions. In this guide, a single two component motion suite will be used for illustration.

Set Load Cases to Run


Analysis>Load Cases to run, ensure that the FNA cases are with dampers and the Ritz modal case
are selected (Figure 7.1.43) and then run analysis. The program then runs analysis and stops at the
conclusion of all the load cases.

@Seismicisolation
@Seismicisolation
61
Figure 7.1.43 Analysis Cases

Analysis Results
Results of seismic NLRHA including story drift ratios, member demands, total viscous damping, etc.
can be reviewed to verify that a structure meets desired performance objectives. If the results are
acceptable, then no additional iterations of damper properties are required, and damper results can
be assessed for specification. If the results do not meet performance objectives, then damping
properties may need to be modified. There are several sophisticated algorithms to optimize damper
properties. The simplest change is to increase or decrease the damping constant and leave the
velocity exponent unchanged. A larger damping constant should be applied if building response is
not acceptable, and a smaller damping constant if damper forces are deemed too large. For typical
applications, it is economical to group the dampers in a few groups.

Damper Response
Display>Show Tables to check analysis, check results, check link results (Figure 7.1.44) Link results
can be exported to an Excel file.

Damper displacements are summarized in Table 7.1.2. Damper forces are summarized in Table 7.1.3.
Designers will likely decide to use more than one size of damper for any given building structure.
For example, one size damper can be used on the bottom two floors and a smaller size on the upper
floor. Alternatively, different damper sizes can be used in the two directions. In this example, for
simplicity, only one damper size will be specified.

From analysis results, maximum damper displacement is +/-2.4 in, thus a total stroke of
approximately 5 inches total. Force is approximately 450 kips.

@Seismicisolation
@Seismicisolation
62
Figure 7.1.44
Selecting Link Results

Table: Link Deformations


Link Load Case/Combo U1

K3 THS-1 Max 0.3


K3 THS-1 Min -0.5
K6 THS-1 Max 0.5
K6 THS-1 Min -0.3
K9 THS-1 Max 0.3
K9 THS-1 Min -0.5
K12 THS-1 Max 0.5
K12 THS-1 Min -0.3
K17 THS-1 Max 0.4
K17 THS-1 Min -0.4
K18 THS-1 Max 0.4
K18 THS-1 Min -0.4
K23 THS-1 Max 0.4
K23 THS-1 Min -0.5

Table 7.1.2
Damper Displacements. Inches

@Seismicisolation
@Seismicisolation
63
Table: Link Forces
Link Load Case/Combo Location P
K3 THS-1 Max I-End 206.645

K3 THS-1 Max J-End 206.645

K3 THS-1 Min I-End -214.237

K3 THS-1 Min J-End -214.237

K6 THS-1 Max I-End 214.491

K6 THS-1 Max J-End 214.491

K6 THS-1 Min I-End -206.871

K6 THS-1 Min J-End -206.871

K9 THS-1 Max I-End 206.888

K9 THS-1 Max J-End 206.888

K9 THS-1 Min I-End -214.396

K9 THS-1 Min J-End -214.396

K12 THS-1 Max I-End 214.303

K12 THS-1 Max J-End 214.303

K12 THS-1 Min I-End -206.595

K12 THS-1 Min J-End -206.595

K17 THS-1 Max I-End 192.495

K17 THS-1 Max J-End 192.495

K17 THS-1 Min I-End -203.899

K17 THS-1 Min J-End -203.899

K18 THS-1 Max I-End 210.408

K18 THS-1 Max J-End 210.408

K18 THS-1 Min I-End -198.181

K18 THS-1 Min J-End -198.181

K23 THS-1 Max I-End 195.179

K23 THS-1 Max J-End 195.179

K23 THS-1 Min I-End -206.689

K23 THS-1 Min J-End -206.689

K24 THS-1 Max I-End 207.664

K24 THS-1 Max J-End 207.664

K24 THS-1 Min I-End -195.592

K24 THS-1 Min J-End -195.592

Table 7.1.3
Damper Forces, kips

@Seismicisolation
@Seismicisolation
64
Cumulative Energy Plot
Display>Cumulative Energy Components to see the figure 7.1.45 below and make sure to select the
appropriate Load Case at the bottom left corner.
Total cumulative energy is displayed in this plot. From this plot one can determine total viscous
damping ratio by scaling the ratio of energy dissipated by the global damping versus viscous
damping. In this case global damping is specified at 5, and as viscous damping is approx. 4 times
larger, the total viscous damping is approximately 4 x 5% = 20% damping.

Figure 7.1.45
Cumulative Energy Plot

@Seismicisolation
@Seismicisolation
65
Part 2: New Construction - Special Moment Frames with FVDs
The combination of a primary structural system and a supplemental damping system is an attractive
and efficient solution for seismic protection of buildings in regions of high seismicity. Pairing fluid
viscous dampers (FVDs) with either steel or reinforced concrete special moment resisting frames
(SMF) results in a highly damped, low-frequency building that limits seismic demand on structural
and nonstructural components. FVDs can be incorporated into seismic design to produce large
equivalent viscous damping; reducing the demand on the special moment frames significantly.

The general approach is to design the SMF members for the minimum strength requirements of the
building code only, meeting all the relevant requirements of ASCE 7-16 except the limitations for
the story drift ratios (SDRs). FVDs are then used to provide reduce the SDRs to comply with
deformation requirements.

An additional design check is required to assure that the structural system is satisfactory to carry
the demands from the dampers; the designer should check force demands on connecting structural
elements and the foundation system. However, since the force in the FVDs is primarily out-of-phase
with the maximum dynamic displacements, the demand on the primary structural system and the
foundation are generally not increased, and the initial design for the ASCE 7-16 strength
requirements is sufficient.

One of the main advantages of using supplemental fluid viscous damping with special moment
frames is the reduction in the steel or concrete tonnage. Since the design of a SMF is generally
governed by the SDR, larger steel or concrete sizes than required to resist building code level design
forces would be required to meet this requirement. When using FVDs to control SDR, smaller
member sizes can be used.

A secondary advantage is that unlike bracing elements, FVDs need not be placed directly in line with
the primary structural system or at each floor level, so long as the diaphragm, collectors, etc. can
carry the load between both systems. This provides greater flexibility to meet architectural
objectives that other lateral systems will not permit.

Evaluation procedures defined by FEMA P-58 – Seismic Performance Assessment of Buildings make
evaluating seismic performance of building structures explicit and can be used to show building
owners and clients the benefits of improved seismic performance. Building structures designed with
fluid viscous dampers will experience a significantly lower level of member nonlinearity (damage)
compared to a code design building without dampers. A decrease in nonlinearity of the primary
structural system greatly reduces the chance of any residual deformations post-seismic activity. The
drifts, accelerations, and the demand on the structural system and other components are all reduced
such that total loss after an earthquake will be significantly less.

@Seismicisolation
@Seismicisolation
66
Overview
The design of a new steel SMF with FVDs is presented. The building considered in this report
is intended to represent realistic construction with seismic risk category II occupancy.

The main components of this section are the following:


• Describe the building and site seismic characteristics.
• Size members and complete design per strength requirements of ASCE 7-16, compute SDRs,
tabulate steel tonnage for SMF members.
• Determine the global equivalent viscous damping required to satisfy the code SDR
requirements. Use the results to estimate the damping constant for analysis.
• Develop a set of matched ground motions for analysis.
• Perform nominal and bounded analysis.
• Perform design check of primary structural elements.
• Provide damper specifications.
• Other design considerations.
• Testing and QC.

Building and Site Seismic Characteristics

Description of the Building


The building is square in plan measuring 150 ft on side consisting of five 30-ft long bays. Typical
stories are 13 ft tall. The gravity system consists of 4-in thick concrete slab supported by steel
gravity beams and columns. The lateral force resisting system (LFRS) comprises three bays of steel
SMF placed on the perimeter. Figure 7.2.1 presents line drawings of plan and elevation views of the
structures. The SMF bays are shown as thick lines in the figure.

The additional dead and live load on the floor and roof are 40/80 and 20/20 psf, respectively. The
building seismic mass is approximately 10,000 kips.

Site Seismic Characteristics


US Geological Survey (USGS) web tools can be used to determine the mapped spectral accelerations
for 0.2 sec, SS, and 1.0 sec, S1, based on the longitude and latitude of the site. The seismic demand
criteria were based on a typical location in Los Angeles, California, with mapped short-period (SS)
and 1-second (S1) spectral accelerations of 1.5g and 0.6g, respectively. The structure was classified
as Risk Category II (I = 1.0) and located on Site Class D, resulting in site coefficients of Fa and Fv of
1.0 and 1.5, respectively. Thus, the risk-based maximum considered earthquake (MCER) short- and
1-second spectral accelerations (SMS and SM1) equaled 1.5g and 0.9g, respectively. The design
earthquake (DE) short- and 1-second spectral accelerations (SDS and SD1) were equal to 1.0g and 0.6g,
respectively. The DE values were based on 2/3 MCER values. This value placed the structures in
Seismic Design Category (SDC) D, according to the ASCE/SEI 7 definition, for both short- and 1-
second spectral intensities.

The spectral acceleration (Sa) as a function of period (T) can be obtained for all period ranges of
interest. The design spectrum is shown in Figure 7.2.2.

@Seismicisolation
@Seismicisolation
67
Plan

Elevation

Figure 7.2.1 Building


Geometry

@Seismicisolation
@Seismicisolation
68
Design of Special Moment Frames
The building model was generated as discussed in Part 1. Next the seismic loading was defined using
the program’s ASCE 7 automatic seismic loading based on the equivalent lateral load procedures.
The seismic load cases were defined using ASCE 7-16 procedure for steel SMF buildings. The load
combinations of ASCE 7-16 were used for the design process. The steel members were assigned
auto select sizes and the programs design module was set to iterate and determine the member
sizes. It is noted that the design met the strength requirements of the code including the provisions
for strong column-weak beam. It is further assumed that since this is a new building, continuity
plates are provided, and if necessary, doubler plates will be added to the columns’ webs. However,
to keep the design economical, a limited number of beam and column sizes were used, as is the
practice in the field. For example, same side beams were used at a floor for a given SMF.
Furthermore, columns were sized to reduce the requirements for doubler plates, again consistent
with field practice. Figure 7.2.3 presents the demand to capacity ratios (DCRs) for the building. The
gravity joists are not shown in the model. There is a choice of explicitly including the joists in the
model or adding bracing points for the beam to account for these members.

Figure 7.2.2
Design Response Spectrum for the Example Building

Figure 7.2.4 presents the distribution of SDR along the building height. The SDR was computed
based on the unreduced (inelastic demand) per ASCE 7-16 requirements. The plots are shown for
two percentages of damping of critical. The first curve corresponds to 5% equivalent damping that
is the basis for the code design. As seen in the figure, the SDR for 5% of critical damping ratio
exceeds the building code threshold of 2% for SDR at nearly all the floors and SDR is larger at the
middle floors.

To comply with the building code provisions, FVDs will be added to the building to reduce SDR.
When a global equivalent viscous damping ratio is 20%, SDR is reduced to 2%, which provides
compliance with the code limit of 2.0% for SDR for DE event. As a starting point, an approximate
damping ratio of 20% (linear damping for the initial trial) will be targeted for the FVDs. This is
somewhat smaller than the required FVD damping because of the following reasons:
• Discrete damping is less effective than global damping. Global equivalent damping is an idealized

@Seismicisolation
@Seismicisolation
69
Figure 7.2.3
Special Moment Frame Design Check

Figure 7.2.4
Computed SDR

@Seismicisolation
@Seismicisolation
70
case and is used to obtain an initial estimate for the required size of FVDs. It is not intended to be
used in design or analysis, as it assumes an idealized orthogonal damping matrix and uniform
damping for the building. This overestimates the performance of discretely distributed FVDs that
produce a non-orthogonal damping matrix and complex mode shapes.
• Bounding analysis is required by ASCE 7-16 which reduces the effectiveness of damping.
• The inherent building damping is likely less than the assumed 5% values.
• Research has shown that adding dampers for a total damping ratio of approximately 20%
improves performance of buildings when subject to large earthquakes.

Damping Configuration and Properties

Damping Configuration
The final size of dampers will be determined by nonlinear response history analysis (as discussed
later). For analysis, an initial damper size is required. Figure 7.2.5 presents the proposed damper
configuration. Note the following:
• The dampers placed symmetrically and along the perimeter are most effective.
• There are two dampers in each direction on each side of the building’s center of mass, satisfying
the redundancy requirements defined by ASCE 7-16.

Figure 7.2.5
Damping Configuration

@Seismicisolation
@Seismicisolation
71
• Dampers are placed on the gravity bays as not to interfere with the SMF bays.
• The dampers are placed at all floors. Given that this is a five-story building, and because the SDR
are similar at different stories, this setup seems intuitive. During analysis, dampers at upper stories
could potentially be eliminated. This is the usual practice in design, however, was not considered
in this report.
• The dampers are placed diagonally. This configuration is effective since the damper axis is at only
23 degrees from horizontal. The diagonal configuration also simplifies the design of connection
elements for dampers. In practice, architectural constraints might necessitate moving of dampers
to different bays. The effect of such relocation of dampers on the performance is minimal as long
as the damper symmetry is maintained and the redundancy requirements are met, given that the
building has rigid floor diaphragms.
• Dampers are placed on the same bay at all levels. Some engineers prefer staggering the dampers
along different bays along the building height to limit the force imparted to the columns. For the
stagger arrangement, collectors need to be provided to transfer the damper force from the floor
above to the floor below. In this report, the single bay approach is used.
• See Section 6.0 of this manual for other damper placement options

Preliminary Damper Size


As discussed in Part 1, many methods and procedures have been developed for preliminary sizing
and configuration of dampers for buildings. This example uses the stiffness proportional damping, a
method derived from the modal strain energy method to obtain the preliminary size of dampers.
Stiffness proportional damping follows the following constitutive relation:

𝐾𝑇𝜁
𝐶𝑡𝑜𝑡𝑎𝑙 = (𝐸𝑞. 7.2.1)
𝜋
Where ζ is the viscous damping ratio, T is the building period, and K is the story stiffness. The
calculation for the initial damping constant in Table 7.2.1 follow:
• K is the story stiffness obtained from static analysis of the building
• Ctotal is the total required damping at a given floor (based on a linear damping exponent)

For this design example the following information is provided, plus Table 7.2.1 below:

• Desired Damping Ratio, ζ = 22% Note: For new construction of low-rise buildings, an
additional 20 25% damping of critical is considered the most efficient for design.
• Building period, T, of 2.1 sec
• There are 4 dampers at each floor and in each direction, and the damper angle, Ѳ, is at 23.5-
degree angle with respect to horizontal.

Note: The damping constant for each damper along its axis can be computed by dividing by the
cosine of the damper angle.

Floor K Ctotal Cdamper


Level (k/in) (k-sec/in) (k-sec/in)
5 422 56 17 Table 7.2.1
Calculation of
4 391 52 16 Preliminary Damping
3 414 55 16 Coefficient
2 458 61 18
1 800 107 32

@Seismicisolation
@Seismicisolation
72
Note from the table that the damping coefficient, C, is similar for the four upper levels, but larger
for the first level since that level is stiffer due to the fixity at the column base; story drifts are also
lower at that level and therefore one damping constant, C = 18, will be specified. In practice, it might
be desirable to use more than one size damper or C value along the building height or in the two
directions, if significantly different reduction in SDR is needed. However, even in such cases, it is a
good practice to limit the number of damper specifications to economize the design. The value of C
= 18 k-sec/in is for linear dampers. In this example we want to use a velocity exponent, α = 0.5, in
order to limit the damper force, but also maintain the same energy dissipation.
A simple method can be used to equate a linear damper with a nonlinear damper. Nonlinear damping
is amplitude dependent so we can recognize our desired SDR of 2%. This means the displacement
of the damper, d, is equal to 2% x story height x cosine(Ѳ).
For this case, the damper displacement, d, is 3.1 in. Then the energy dissipated in one cycle of
motion, Wj, can be compared between both a nonlinear and linear damper as shown in Figure 7.2.6
below. The energy dissipated in one cycle of motion, Wj, is calculated as the area within force-
displacement loop. Reference MCEER Technical Report 00-0010 for calculation of Wj for both
nonlinear and linear dampers.

Figure 7.2.6
Force-Displacement Loop for One Cycle of Motion

For this example, the initial damping constant value of 50 with a velocity exponent, α =0.5, is used.
As a check, once can calculate the maximum damper force for each damper. Since the building
period is 2.1 sec, then the damper velocity can be computed:
2𝜋
𝑉= 𝑑 = 9.3 𝑖𝑛/𝑠𝑒𝑐 (𝐸𝑞. 7.2.3)
𝑇
Now for a linear damper (α =1), the damper force is then:

𝐹 = 𝐶𝑉 = 168 𝑘𝑖𝑝𝑠 (𝐸𝑞. 7.2.4)


Whereas for a nonlinear damper (α=0.5), the damper force is then:
𝐹 = 𝐶𝑉 𝛼 = 152 𝑘𝑖𝑝𝑠 (𝐸𝑞. 7.2.4)
Even though, both dampers produce the same amount of energy dissipation, the damper force is
approximately 10% smaller for the nonlinear damper, compared to the linear damper, showing the
higher efficiency of nonlinear (low exponent) dampers.

@Seismicisolation
@Seismicisolation
73
Bounding Procedures
Seismic protective devices (isolators and dampers) are propriety products manufactured by a select
group of vendors. Similar to all manufacturing, certain variation in properties from nominal can be
expected. ASCE 7-16 recognizes this and specifies:

A maximum and minimum analysis and design property shall be established for each modeling parameter
as necessary for the selected method of analysis. Maximum velocity coefficients, stiffness, strength, and energy
dissipation shall be considered together as the maximum analysis and design case, and minimum velocity
coefficients, strength, stiffness, and energy dissipation shall be considered together as the minimum analysis and design
case.

The standard requires two steps:

1.Determination of property modification (λ) factor (PMF).

2.Performing bounding analysis.

The PMF can vary significantly from product to product and from one manufacturer to another. The
standard recognizes the good quality control in manufacturing of FVD by Taylor Devices Inc. (See
Figure C18.2-1 of ASCE 7-16, the lowest permitted value of 15% for the PMF can be used. The
PMF applies to the damping constant only and the velocity exponent is not changed. Thus, the
following two cases are considered:
• Lower bound analysis C=85%C nominal or C=43. This case governs for the SDR as the lower
damping ratio is used.
• Upper bound analysis C=120% (1/85%) C nominal or C=60.This case results in larger damper
forces and governs the design of all components and members that transmit the damper force to
ground.

Seismic Loading
For design and analysis of new structures with FVDs, engineers can use either the nonlinear
response history analysis (NLRHA) procedure or alterative procedures based on the modal strain
energy method. The use of modal strain energy methods are subject to certain limitation. For
example, dampers are required at all levels, there is a cap on the effective damping ratio, and
application is limited to lower seismic regions. Methods other than NLRHA are developed to provide
approximate analysis and the standard recognizes this by referring to them as alternate procedures
and emphasizes the use of NLRHA procedure.

The NLRHA procedure requires that the dampers be modeled as nonlinear elements to capture their
velocity dependence and hysteretic behavior. However, the primary structural system can be
modeled as linear elements because the use of FVDs reduces the demand on the buildings, and thus
limits the extent of nonlinear response. This approach is used in this report.

Development of Acceleration Records


For this example, 7 pairs of strong motion suites were chosen from the PEER Ground Motion
Database. The records were selected from strong motion data that can be expected at the site. The
records were then spectrally matched to the target spectrum of Figure 7.2.2 using the ETABS
application shown in Part 1. Figure 7.2.7 shows the response spectra for the 14 acceleration records
tightly matched the target response spectrum.

@Seismicisolation
@Seismicisolation
74
Matched Records DE Level

Figure 7.2.7
Spectrally Matched Ground Motion Records

Analysis Results
Two models, with upper bound and lower bound damping coefficients, were analyzed for the seven
two-component spectrally matched ground motion suites. A load combination was defined
averaging the ground motion results for each model, including the effect of vertical loads. The lower
bound model was used to check design of the primary structural system, compute the SDR, and
check damper stroke requirements; the upper bound model was used to check the damper force
demands.
Lower Bound Analysis
Figure 7.2.8 presents the SDR computed from response history analysis for the building with
dampers and from the original model. It is noted that the dampers are quite effective in reducing
SDR (reduction of approximately 40%) and for the building with dampers, the maximum SDR is
approximately 1.6%, meeting the ASCE 7-16 requirements for SDR.

Figure 7.2.10 presents the demand to capacity ratios (DCR) for the load combination of the average
of the NLRHA cases. For clarity, data from only the SMF elements are shown. Note the following:
• Most members remain elastic.
• For a code design building an R factor of 8 is used. This implies large energy dissipation by the
ductile steel (concrete) members. This energy dissipation is obtained by the yielding of ductile
members. Thus, a code-compliant building is expected to perform well but sustain significant
damage in the design earthquake. By contrast, the building with dampers experiences very limited
nonlinear behavior, and therefore much less damage (or possibly no damage).

@Seismicisolation
@Seismicisolation
75
Figure 7.2.8
Reduction in SDR After the Addition of Dampers

Upper Bound Analysis


Figure 7.2.9 presents the force-displacement relation for one of the dampers on the second floor.
For this acceleration record, the maximum damper force is approximately 200 kips. As seen in the
figure, the dampers dissipate a large amount of seismic energy (cumulative area in the loops).

Figure 7.2.9
Force-Displacement Plot - Second Floor Damper

@Seismicisolation
@Seismicisolation
76
Typical Frame

Figure 7.2.10
Design Check

@Seismicisolation
@Seismicisolation
77
DCR values are smaller for this analysis but the damper forces are larger. As such, components such
as diaphragms, collectors, and foundation elements need to be checked for the damper force
computed from the upper bound analysis. Table 7.2.2 presents the damper responses.

Floor Damper Force Damper Displacement


Level (kips) (in)
ROOF 123 1
L4 145 1
L3 166 2
L2 179 2
L1 157 1
Table 7.2.2
Calculated Damper Responses

Fluid Viscous Damper Specification Requirements


One size of damper can be specified for this building since all the dampers have the same properties
and similar force and stroke demands. When specifying damper properties with Taylor Devices, Inc.
the drawings and specifications should indicate the following:
• Number of dampers, n = 40
• Nominal Damping constant, C = 50 kips-sec/inα
• Velocity Exponent, α = 0.5
• 130% MCE Damper Stroke = +/-4 in
Note: The maximum computed displacement at the design earthquake is approximately 2 in. The
anticipated displacement at MCER would be 1.5xDBE, approximately +/-3 in. It is also desirable to
include a factor of safety of 1.3 to avoid the damper bottoming out. This does not add much to the
damper cost and results in a 4 in stroke.
• (130% Velocity-Rated) MCE Level Force = 270 kips
Note: The maximum damper force is approximately 180 kips from the design earthquake. MCE level
analysis is required for damper specification per ASCE; it is also required that MCE level velocities
be scaled by 130% to specify damper forces. This is done simply by back calculating MCE velocity
from the specified damper properties. Typically, damper force specification requirements are no
more than 50% larger than DBE level force demands.

Figure 7.2.11 is Taylor Devices, Inc standard damper sizes. Material requirements for the dampers
are controlled by force and stroke and thus the cost of fluid viscous dampers is dependent on these
parameters. The cost is minimally affected by specification of the damping constant and the velocity
exponent. Also, note that Taylor Devices typically adds a factor of safety of approximately 2.0 to
the damper capacity shown in the figure 7.2.11.

@Seismicisolation
@Seismicisolation
78
Figure 7.2.11
Fluid Viscous Damper Sizes (Taylor Devices)

Design Considerations
Special moment frame design and the damper specifications were adequate and did not require
iteration in this example. Typically, one or two cycles of iterations can be necessary before final
design is complete.
Amongst other details, the designer will need to consider the following details for final design:
• Check damper connections, columns, diaphragms, collectors, and foundations connected to
damping devices to resist damper force specifications elastically.
• Design of the extender brace. The extender brace serves to attach the damper unit to the
structure. It must be strong enough to withstand the damper force without buckling and
adequately stiff as discussed in Part 1. The actual extender brace stiffness is computed using
the size and length of the brace.

Steel Tonnage
Table 7.2.3 lists the column and beam sizes for two models: one designed without dampers meeting
both the strength and drift requirements of the building code, and one meeting only the code
strength requirements and utilizing dampers to control drift. The reduction in steel tonnage
compensates for the additional cost of the dampers.

@Seismicisolation
@Seismicisolation
79
LFRS member sizes Code Damper
L1-L3 W24x229 W24x146
Columns
L4-Roof W24x176 W24x131
L1-L3 W24x94 W24x76
Beams
L4-Roof W24x76 W24x62
Beams 78 63
Steel Columns 109 73
Tonnage total 187 136
Table 7.2.3
Steel Members for Moment Frames

Testing and Quality Control


ASCE 7-16 requires prototype testing of new devices. These tests are extensive and time consuming
and can therefore add unnecessary cost/time to a project. However, since in almost all cases, a
damper similar to a unit previously tested by Taylor Devices is specified, these tests can be
eliminated.

ASCE 7-16 also requires QC production testing of dampers. Taylor Devices tests 100% of their units
prior to shipment to the job and these tests can be witnessed by the design engineer, if desired.
Refer to Taylor Devices sample specifications for typical damper production testing.

@Seismicisolation
@Seismicisolation
80
Part 3: Seismic Retrofit – Moment Frame with FVDs

Steel special moment frame (SMF) buildings constructed prior to the 1994 Northridge Earthquake
typically used welded unreinforced flange (WUF) beam-to-column connections. The Northridge and
Kobe earthquakes revealed that this then-popular construction was subject to premature brittle
failure. FEMA 351 (FEMA 2000) presents methodology for the retrofit of such buildings. The
application of fluid viscous dampers is one of the most highly recommended seismic retrofit options
for efficiency and construction flexibility. In most cases a foundation retrofit is not required,
significantly shortening the construction schedule and reducing cost.

The FEMA document writes:

The intent of seismic upgrades employing supplemental energy dissipation devices, also called dampers,
is to reduce the amount of deformation induced in the structure during its response to ground shaking.
In this respect, it is similar to upgrades accomplished through global structural stiffening. However,
rather than introducing stiffening to a structure, this upgrade technique reduces deformation through
the dissipation of energy within a series of devices that are introduced into the structure as part of the
upgrade.

Fluid viscous dampers (FVDs) provide an ideal solution for the seismic retrofitting of SMF buildings.
They serve to reduce the story drift ratios (SDRs) and thus the demand on the vulnerable
connections by increasing the damping in the system.
• The SMF buildings are flexible by nature and thus FVD will become easily activated by
motion of the building.
• FVDs do not add stiffness to the system and thus the period of the building will remain
unchanged. For more flexible building, this retrofit approach will maintain the building period
away from the constant amplitude plateau and thus limit the seismic forces. By contrast, when
retrofit strategies, which increase lateral stiffness, are implemented, the building period will shift
towards the plateau and thus result in an increase in the seismic demand.
• The force in FVDs is velocity dependent and thus primarily out-of-phase with the inertial forces
of the building. Thus, the increase in demand on the existing members and foundations is
generally less than conventional seismic retrofitting.
• FVD can be sized to further limit the force in the dampers and building.
• The addition of supplementary damping reduces the peak floor accelerations and thus protects
the vulnerable acceleration-sensitive nonstructural components and building content.
• Since the seismic energy is dissipated by the FVDs, the structural damage is reduced or
eliminated. This is in contrast to well-engineered ductile buildings for which energy dissipation
occurs by the nonlinear behavior in the ductile elements. To attain such ductility, the engineer
accepts the inevitable associated controlled damage.

@Seismicisolation
@Seismicisolation
81
OVERVIEW
In this Part, the seismic retrofit of an existing steel SMF with FVDs is presented. The building
considered in this report, is intended to represent a realistic construction with seismic risk
category II occupancy, commercial or residential occupancy.

The main components of this section are the following:


• Describe the building characteristics and site
• Select target performance per ASCE 41-17
• Determine a desired damping ratio using ASCE 41-17 table and size dampers
• Perform nominal and bounded analysis
• Compute demands and capacities for structural members and connections and compare with the
m-factors of ASCE 41-17
• Provide damper specifications
• Other Design Considerations
• Testing and QC

BUILDING AND SITE SEISMIC CHARACTERISTICS


The five-story building is square in plan measuring 150 ft on each side consisting of five 30-ft long
bays. Typical stories are 13 ft tall. The gravity system consists of 4-in thick concrete slab supported
by steel gravity beams and columns. The lateral force resisting system (LFRS) comprises three bays
of steel SMF placed on the perimeter. Figure 7.3.1 presents line drawings of plan and elevation
views of the structures. The SMF bays are shown as thick lines in the figure. The additional dead
and live load on the floor and roof are 40/80 and 20/20 psf, respectively. The building seismic mass
is approximately 10,000 kips including contribution from nonstructural components and gravity
members. The building was constructed in early 1980s and thus WUF beam- to-column connections
were used. Material nonlinearity was not included in the analysis because the members were
anticipated to essentially remain elastic (DCR of 1.5 or less). Therefore, linear modeling of structural
members and connections were permitted by ASCE 41-17. Furthermore, the connections did not
have continuity plates or doubler plates. The design of the building was based on the strength and
drift provisions of the building code. As such, the existing building had a maximum SDR of 2%. This
was code-compliant, although, the vulnerable connections could not resist the rotations associated
with that level of drift. The building is retrofitted using the provisions of ASCE 41-17. The seismic
retrofit had the following objectives:
• Meet the basic performance objective of ASCE 41-17 for life safety at BSE-1N (approximately
475-year return period) earthquakes.
• The building to remain operational during the retrofit.
• The seismic retrofit to allow for the existing architectural features and occupancy constraints.
• Retrofit to be cost effective and also meet certain life-cycle goals.
• The building is a candidate for rating program and the goal is to obtain a gold or platinum rating.

To meet these requirements, FVDs were used to reduce the SDR to approximately 1% at BSE 1N.
At such a low level of SDR, the structural members and connections are expected to remain
(essentially) elastic. The selected damper retrofit allowed continued occupancy and was integrated
to meet the architectural constraints.

@Seismicisolation
@Seismicisolation
82
Plan

Elevation

Figure 7.3.1
Building Geometry

@Seismicisolation
@Seismicisolation
83
Site Seismic Characteristics
US Geological Survey (USGS) web tools can be used to determine the mapped spectral accelerations
for 0.2 sec, SS, and 1.0 sec, S1, based on the longitude and latitude of the site. The seismic demand
criteria were based on a typical location in Los Angeles, California, with mapped short-period (SS)
and 1-second (S1) spectral accelerations of 1.5g and 0.6g, respectively. The structure was classified
as Risk Category II and located on Site Class D, resulting in site coefficients of Fa and Fv of 1.0 and
1.5, respectively. Thus, BSE-2N (approximately 2475 year) short- and 1-second spectral
accelerations equaled 1.5g and 0.9g, respectively. The BSE-1N (approximately 475 year) short- and
1-second spectral accelerations were equal to 1.0g and 0.6g, respectively. The spectral acceleration
(Sa) as a function of period (T) can be obtained for all period ranges of interest. The design spectrum
is shown in Figure 7.3.2.

Figure 7.3.2
BSE 1N Response Spectrum for the Example Building

Conditions of Existing Structure


The building met both the strength and drift requirements of the code. Since the design of SMF
buildings is usually, governed by drift requirement, the member sizes were selected based on this
drift requirement. Figure 7.3.3 presents the distribution of SDR along the building height. The SDR
was computed based on the unreduced demand. As seen in the figure, the SDR is approximately 2%
for this building.

Figure 7.3.3
Computed SDR

@Seismicisolation
@Seismicisolation
84
DAMPING CONFIGURATION & PROPERTIES

The final size of dampers will be determined by nonlinear response history analysis as discussed
later. For analysis, an initial damper size is required. Figure 7.3.4 presents the proposed damper
configuration. Note the following:
• The dampers are placed along the perimeter. This is the most effective and allows the owner to
showcase that this is a state-of-the-art retrofit.
• There are at least 2 damper bays in each direction on each side of the building center of mass.
Thus, the redundancy requirements of ASCE 41-17 are satisfied.
• Dampers are placed in the Chevron (double diagonal) configuration because it was required to
keep the middle portions of the bays open. Only dampers at the bottom three levels are used in
the first pass. During analysis, dampers at upper stories could potentially be added. However,
dampers will not be placed at the top story due to architectural constraints.

Figure 7.3.4
Placement of Dampers

@Seismicisolation
@Seismicisolation
85
Preliminary Damping Estimate
When supplementary damping is added to structures, an estimate in the reduction in dynamic
response can be determined using the numerical damping coefficient, B1, defined by ASCE 41-17
section 2.4.1.7.1. Below is the Damping Coefficient, B1, for varying levels of damping. A given
dynamic response can be divided by the damping coefficient to approximate reduced response.

Effective damping, g Damping coefficient (B)


2% 0.8

5% 1.0

10% 1.2

20% 1.5

30% 1.8

40% 2.1

50% 2.4

Table 7.3.1
Numerical Coefficients for Damping Ratios

Since the objective of the seismic retrofit is to reduce maximum SDR from 2% to 1%, a numerical
coefficient of approximately 2 is desired. This value corresponds to a damping ratio of 30%-40%.
FVDs were initially sized to provide damping ratio of approximately 35% of critical
• Since there are only three levels of dampers, the same dampers were used at all levels.
• A velocity exponent (a) of 0.4 was used to limit the damper force at large earthquakes (high
damper velocities).
• The nominal damping coefficient, C, used for each damper is 70 kip-sec/in.

SEISMIC LOADING
For retrofit of existing structures that use energy dissipation devices, engineers can use either the
nonlinear response history analysis (NLRHA) procedure or alterative procedures based on the modal
strain energy method. The use of methods other than NLRHA are subject to certain limitations. The
NLRHA procedure requires that the dampers be modeled as nonlinear elements to capture their
velocity dependence and hysteretic behavior. However, the primary structural system can be
modeled as linear elements because the use of FVDs reduces the demand on the buildings, and thus
limits the extent of nonlinear response.

Development of Acceleration Records


To perform NLRHA, the user must define sets of independent pairs for a minimum of 7 ground
motion records.

Input Histories
For this report, 7 pairs of strong motion records were chosen from the PEER Ground Motion
Database. The records were selected from strong motion data that can be expected at the site.
These records were then scaled to the target spectrum of Figure 7.3.2 using the method described
in Part 1.

@Seismicisolation
@Seismicisolation
86
Figure 7.3.5
Spectrally Matched Ground Motion Records

ANALYSIS RESULTS
Two models, with upper bound and lower bound damping coefficients, were analyzed for the
seven two-component spectrally matched ground motion suites. A load combination was defined
averaging the ground motion results for each model, including the effect of vertical loads. The
lower bound model was used to check design of the primary structural system, compute the SDR,
and check damper stroke requirements; the upper bound model was used to check the damper
force demands.

Lower Bound Analysis


Figure 7.3.6 presents the SDR computed from response history analysis for the building with the
damper and from the original model. It is noted that the dampers have reduced the maximum SDR
to approximately 1%, the target of analysis. If larger SDRS were obtained, then the damping
constant would be increased, and analysis repeated. In typical cases, only one of two iterations
would be necessary to finalize the damper properties.

@Seismicisolation
@Seismicisolation
87
Figure 7.3.6
Reduction in SDR After the Addition of Dampers

Table 9-4 of ASCE 41-17 provides information on the m-factors for beams, columns, and
connections. Although the m-factors for beams and columns are high for compact sections. The
design is governed by the m-factor for the WUF connection at live safety performance and is given
by:
Eq. 1. m=4.3-0.083d

This factor could be further reduced per Section 9.4.2.4 of the standard as follows:

Eq. 1. m=compactness factor*panel zone factor* continuity plate factor* beam span factor

For example, with this building, for connections of W30 beam, compact beams and col beams and
columns, non-compliant panel zone, complaint, beam span-to-depth ratio, and connection without
continuity plates, the m-factor is computed from:

Eq. 2. m=(1.0)*(0.8)*(1.0)*(0.8)*(4.3-0.083*30)=1.2

Note: if the m-factor is less than one, use one instead. Therefore, the building will need to remain
essentially elastic at the BSE 1N level to meet the life safety performance. Figure 7.3.7 presents the
DCR for the load combination of the average of the NLRHA cases. For clarity only data for the SMF
elements are shown. As seen in the figure, the retrofitted building met its performance goal, as all
the members remained elastic. The seismic retrofit protected the vulnerable WUF connections
without the need to repair these connections.

@Seismicisolation
@Seismicisolation
88
Building

Figure 7.3.7a
DCR for BSE 1N (Allowable m-factor 1.2 for Life Safety)

Figure 7.3.7b
DCR for BSE 1N (Allowable m-factor 1.2 for Life Safety)

Upper Bound Analysis


Figure 7.3.8 presents the force-displacement relation for one of the dampers. For this acceleration
record, the maximum damper force is approximately 160 kips. As seen in the figure, the dampers
dissipate significant amount of seismic energy (cumulative area in the hysteresis loops).

@Seismicisolation
@Seismicisolation
89
Figure 7.3.8
Force-Displacement Plot - Second Floor Damper

DCR values are smaller for this analysis but the damper forces are larger. As such, components such
as diaphragms, collectors, and foundation elements need to be checked for the damper force
computed from the upper bound analysis. Table 7.3.2 presents the damper responses.

Floor Damper Force Damper displacement


Level (kips) (in)
ROOF -- --
L4 -- --
L3 162 1.3
L2 163 1.2
L1 145 1.5

Table 7.3.2
Computed Damper Responses

One size of damper can be specified for this building retrofit since all the dampers have the same
properties and similar force and stroke demands. When specifying damper properties with Taylor
Devices, Inc. the drawings and specifications should indicate the following:
• Number of dampers, n = 24
• Nominal Damping constant, C = 70 kip-sec/inα
• Velocity Exponent, α = 0.4
• 200% BSE-1E Damper Stroke = +/-3 in
Note: The maximum computed displacement at BSE-1E is approximately 1.5 in. ASCE 41-17
requires that the damper displacement specification be 200% of the BSE-1E level demand.
• 200% BSE-1E Damper Force = 330 kips
Note: The maximum damper force is approximately 165 kips from BSE-1E. ASCE 41-17 requires
that the damper force specification be 200% of the BSE-1E level demand.
@Seismicisolation
@Seismicisolation
90
Figure 7.3.9
Fluid Viscous Damper Sizes (Taylor Devices)

@Seismicisolation
@Seismicisolation
91
DESIGN CONSIDERATIONS
Amongst other details, the designer should consider the following for seismic assessment :
• Check the existing elements such as columns, diaphragms, collectors, and foundations for program
output forces (including seismic and demand from dampers) and strengthen as required.
• Design all new members and connections connected to damping devices for the specified damper
forces.
• Design of the extender brace. The extender brace serves to attach the damper unit to the
structure. It must be strong enough to withstand the damper force without buckling and
adequately stiff as discussed in Part 1. The actual extender brace stiffness is computed using the
size and length of the brace.

TESTING AND QUALITY CONTROL


ASCE 41-17 requires prototype testing of new devices. These tests are extensive and time
consuming and can therefore add unnecessary cost/time to a project. However, since in almost all
cases, a damper similar to a unit previously tested by Taylor Devices is specified, these tests can be
eliminated.

ASCE 41-17 also requires QC production testing of dampers. Taylor Devices tests 100% of their
units prior to shipment to the job and these tests can be witnessed by the design engineer, if desired.
Refer to Taylor Devices sample specifications for typical damper production testing.

@Seismicisolation
@Seismicisolation
92
REFERENCES

American Society of Civil Engineers (ASCE) 2018, ASCE/SEI 41-17, Seismic Evaluation and Retrofit of
Existing Buildings Arlington, VA

American Society of Civil Engineers (ASCE) 2017, ASCE/SEI7‐16, Minimum Design Loads For
Buildings and Other Structures, Arlington, VA

Computes and Structures Inc. (CSI) 2017, ETABS : Integrated Design, Analysis, and Drafting of Building
Systems, Walnut Creek, CA https://www.csiamerica.com/products/etabs

Federal Emergency Management Agency (FEMA), FEMA 351, Recommended Seismic Evaluation and
Upgrade Criteria for Existing Welded Steel Moment-Frame Buildings, Washington DC.

Taylor Devices Inc. 2017 http://taylordevices.com and www.seismicdampers.com

MCEER 2001, Development and Evaluation of Simplified Procedures for Analysis and Design of
Buildings with Passive Energy Dissipation Systems, Technical Report, MCEER 00-0010

NEHRP 2015 Recommended Seismic Provisions: Design Examples

2012 IBC – SEAOC Structural Seismic Design Manual – Volume 5, Examples for Seismically Isolated
Buildings and Buildings with Supplemental Damping

@Seismicisolation
@Seismicisolation
93
@Seismicisolation
@Seismicisolation
94
8 Fluid Damper Performance vs.
Other Technologies

With an increasing number of applications of Taylor fluid viscous dampers (FVD) in a wide variety of
projects, the superior performance and applicability of these velocity-dependent devices for seismic-
and wind-resistant design are more recognized by structural engineers. This chapter highlights the
critical characteristics of FVDs by comparing their characteristic with several other commonly used
passive energy-dissipating devices, including buckling restrained braces (BRB), friction dampers,
yielding dampers, visco-elastic dampers (VED), viscous wall dampers (VWD), and tuned mass dampers
(TMD). Several key advantages of FVDs include: reduced seismic demands due to period shifting;
minimal interaction with primary structural system; less dependency on frequency and temperature;
more flexibility in terms of locations, configurations and size selection. Moreover, FVDs are easier to
implement, require less effort for maintenance during service period and need no external power,
control actuators or sensors.

Buckling Restrained Braces


Buckling restrained brace (BRB) is a commonly used energy-dissipation device worldwide. Statistics
[1] has shown that hysteretic dampers (including BRBs) occupies the largest percentage of use in the
seismic protective devices market, which is partly due to their somewhat inexpensive, initial price.
Another factor that might explain the trend lies in the fact that in most countries except Japan, BRBs
are treated as ordinary braces, thus conventional design methods such as the equivalent lateral force
analysis method or modal response spectral analysis method can be used, consequently there are no
requirements for peer review.

As an alternative, using Taylor Device FVDs would be an equally straight-forward way to be


implemented in a structure since they have been extensively tested, researched, and applied in
thousands of projects. More importantly, Taylor FVDs have several advantageous characteristics
than BRBs, as listed below.

Reduce floor accelerations


The mechanical properties of a BRB are displacement-dependent, and a BRB exhibits bilinear
hysteretic behavior. Contrary to a conventional brace that buckles in compression, a BRB avoids
buckling with its encasing unit, and usually has slightly greater strength in compression than tension
(Figure 8.1).

@Seismicisolation
@Seismicisolation
95
Figure 8.1
Typical axial force-displacement behavior of
a conventional brace versus a BRB [2].

A noted feature of a BRB system is that it provides static stiffness to a bare frame, increasing its lateral
stiffness and thus attracting larger seismic forces due to period shifting. This could be illustrated by
Figure 8.2, which shows a code-complied design spectrum. For a typical low-rise to high-rise building,
the fundamental period usually falls into the velocity-constant or displacement-constant range,
stiffening the structure will shift its fundamental period to the left (from blue dashed line to the red
dashed line), and the seismic force demand would be amplified; see the changes from the blue arrow
to the red arrow.

Figure 8.2
Typical design response spectrum shape.

@Seismicisolation
@Seismicisolation
96
Consequently, though the additional damping BRBs provide that helps reduce the drift ratio of a
structure, BRBs are less efficient in reducing the base shear and floor accelerations of a building
compared to FVDs. Take one case study for example. In a recent study [3], the cost-efficiency of three
different energy dissipation devices to upgrade an existing tall steel moment frame was examined,
including the case using (1) FVDs; (2) BRBs; and (3) viscous wall dampers (VWDs). Fig. 3 shows the
distribution of peak floor accelerations for different cases, including: (a) the original building before
adding any supplemental energy dissipation devices (black line); (b) the building incorporating FVDs
(blue line); (c) the building with BRBs (pink line); and (d) the building with VWDs (red line). The
locations for installation and effective damping ratios were kept the same for case (b), (c) and (d).
It is clearly shown that the addition of BRBs stiffened the building and increased the seismic force
demand. This has led to larger accelerations throughout floors, with a majority of floors exceeding
the values of the original building.

Figure 8.3
Distributions of peak floor accelerations [3].

On the contrary, FVDs are velocity-dependent devices. The typical hysteresis behavior of a linear
FVD (Figure 8.4) indicated that such a device does not provide static stiffness, avoiding additional
seismic forces in the structure. Moreover, FVDs provide damper forces that are out-of-phase with
displacement, hence damper forces would not increase the story forces and floor accelerations; see
blue line of Figure 8.3.

Figure 8.4
Typical axial force-displacement
behavior of a FVD.

@Seismicisolation
@Seismicisolation
97
Control residual drift
As we can see from the hysteresis loop of a BRB (Figure 8.1), this device has low post-yield stiffness
and lacks re-centering capability. Consequently, the residual drift ratios of a building with BRBs after
an earthquake event would be large. In contrast, a FVD has much better re-centering capability, and
could result in much reduced residual drift ratios.

Limit interactions with existing columns


BRBs are purely displacement-dependent, and they develop forces that are concurrently with
structural movement, i.e., the damper forces are in-phase with structural displacement. Therefore,
a large portion of forces will be transferred to elements connected with these dampers. Although
the additional damping provided by BRBs help reduce the drift ratios of the structure, and hence
decreasing the forces on columns due to overturning, the additional damper forces transferred to
connected columns would compensate for such a reduction, which might end up with even larger
column axial forces. For example, the case-study building examined in [3] indicated that only the
scheme using FVDs (blue lines) was able to reduce the column axial force Demand/Capacity (D/C)
ratios, whereas BRBs (pink lines) increased the column D/C ratios by about 20-30% for columns
located at lower floors; see Figure 8.5.

Figure 8.5
Distributions of peak column axial D/C
in a case-study building: (+) tension (-)
compression [3].

Friction Dampers
A friction damper is another kind of displacement-dependent energy dissipation device. Fig. 6(a)
shows a friction device (also known as “Pall device”), which is placed at the intersection of the cross
braces. When an earthquake occurs, the brace in tension forces the damper at the joint link to slip,
which activate the four links and force the compression brace to shorten. In this way, the brace
buckling could be avoided, energy could be dissipated in both tension and compression braces. The
hysteresis loop of a friction damper is illustrated in Figure 8.6(b).

@Seismicisolation
@Seismicisolation
98
Figure 8.6
Illustration of (a): a friction damper (photo courtesy of James Kelly); and (b) hysteresis loop of a friction damper.

Similar as a BRB, a friction damper exhibits a few disadvantages compared to a FVD, including:
• Inducing larger seismic forces;
• Increasing the peak floor accelerations;
• Increasing base shears;
• Resulting in larger residual drift ratios;
• Increasing the column axial forces connected with them.

In addition to above-mentioned drawbacks, a friction damper exerts a constant force for all levels of
earthquake excitations, and thus the friction force at each story level needs to be carefully selected
to achieve the optimal performance. Besides, a friction damper usually comes in the X-crossing
configuration, taking up more spacing than a diagonal or a chevron-type configuration.

Yielding Dampers
A yielding damper dissipates energy through the yielding properties of mild steel. The typical
hysteresis loop is shown in Figure 8.7. As with a BRB or a friction damper, a yielding damper
moves in-phase with displacement, incurring larger seismic forces and likely leading to larger floor
accelerations and base shears. Moreover, they develop forces that are concurrent with structural
movement, and cause a large portion of forces to be transferred to structural members connected
with these dampers. Whittaker et al. [4] tested on 3-story model with one kind of yielding damper:
added damping and stiffness devices (known as “ADAS”), and revealed a 14% increase of column axial
load compared to the bare frame without ADAS. Also, these dampers lack re-centering capabilities
and would result in large residual drift ratios of a building after an earthquake excitation.

Additionally, most yielding dampers utilize metallic materials, and would require replacement after a
major event, thus increasing the associated life-cycle cost.

@Seismicisolation
@Seismicisolation
99
Figure 8.7
Hysteresis loop of a yielding damper

Viscoelastic Dampers
Besides devices depend on purely displacement, there are other kinds of devices that perform based
on both displacement and velocity. One of such devices is viscoelastic damper (VED). A VED consists
of layers of viscoelastic solid, and sandwiched between steel plates. It dissipates energy through
relative shear deformation of viscoelastic material. In general, viscoelastic solid materials exhibit
both storage stiffness and loss stiffness, and their mechanical properties would depend on both
frequency and temperature [5]. As a comparison, a FVD system out-performs a VED system in the
following aspects.

Limit seismic force and in-phase components


Figure 8.8 shows a single degree of freedom (SDOF) system model representation of a viscoelastic
damper system, with the driving braces (Kb) in series with damper and the frame (Kf) included.
When a VED system is subject to the harmonic displacement u(t)=u max sin (~t) (~ is the circular
frequency), the damper force has components that are both in-phase and out-of-phase with the
displacement , as expressed below:

Fd(t) = Ka'u max sin ~t + Ka"u max cos ~t (1)

Where Ka' and Ka" are the storage stiffness (provide additional stiffening) and loss stiffness (provide
additional damping) of the brace-damper subsystem, and their mathematical expressions are
illustrated as [6]:

(Kb+Kd ') KbKd '+KbKd " 2 Kb 2 Kd" 2


Ka' = Ka"= (2)
(Kb+Kd')2Kd" 2 (Kb+Kd')2Kd" 2

Note: Kd"=C~

In the case of a FVD system, no additional brace stiffness exists, i.e., Kd '=0. As such, the above
expressions are reduced to:

KbKd " 2 Kb 2 Kd" 2


Ka' = Ka"= (3)
Kb 2 +Kd" 2 Kb 2 +Kd" 2

Comparing Eqn. (3) and Eqn. (2), we can see that the static stiffness (Ka') is zero for a FVD system

@Seismicisolation
@Seismicisolation
100
when ~ is zero, but not the case for a VED system. A direct result of zero static stiffness is that the
modal properties of a structure after adding FVDs keep the same. On the contrary, incorporation of
VEDs stiffens the structure, and increases its seismic input. Besides, under a dynamic loading case
(i.e., ~ is greater than zero), as long as driving braces stiffness Kb is fairly large, the storage stiffness
(Ka') in Eqn. (3) is approximately zero, indicating that the in-phase component in a FVD system is
negligible when subject to external excitations. However, in a VED system, both the storage stiffness
and loss stiffness exist regardless of the size of driving braces. This in-phase components enlarge the
interaction between VEDs and the structural members, and result in larger forces within elements
that are connected with the dampers, as discussed in previous sections.

Figure 8.8
a SDOF system with a viscoelastic damper [6].

Provide stable mechanical properties


Additionally, Taylor FVDs provide stable mechanical properties over a wide range of the frequency
and temperature, and are almost completely independent of the amplitude of motion. Early research
reported by Constantinou and Symans [5] indicated that a FVD exhibited relative stable mechanical
properties over a frequency range of 0.1 HZ to 4 HZ, and could operate over a wide temperature
range (-40 °C to 70 °C). The peak damper forces in the tests with frequency of 2 and 4 HZ were almost
identical, and the loss stiffness of dampers reduced by a factor smaller than 2 when the temperature
was in the range of 0 °C to 50 °C. Such a range of change is very small when compared to a VED.

A VED could exhibit a significant dependency on frequency and temperature. For example, it was
reported that a VED changed its storage and loss shear stiffness by a factor larger than 7 when
the frequency changes from 0.1 HZ to 4 HZ; and exhibit a close to 50-fold decrease in stiffness in
the temperature range of about 0 °C to 50 °C [5]. Such a large change of mechanical properties of
VEDs when subjected to environmental settings (e.g., temperature) would diminish their additional
damping effect, and adversely impact the structure. Consider when VEDs are used in a tall building,
the building might have asymmetric stiffness due to dramatic changes of damper properties over
story height.

Viscous Wall dampers


Another device that exhibits both displacement and velocity dependency is viscous wall dampers
(VWDs). As with VEDs, adding VWDs to a building would change the modal properties of a building,
and hence increasing the seismic forces; VEDs provide in-phase components, and thus having more
interactions with the primary structural system. This could also be illustrated in the case study done
by Wang and Mahin [3]; see Fig. 5 of the comparison of column axial D/C ratios for different systems.

Moreover, VWDs were found to cause critical issues when used in an existing building with vulnerable
members (e.g., Pre-Northridge connections). Wang and Mahin [3] identified that using wall-type
of dampers on a case-study Pre-Northridge steel moment frame could change the typical beam
deflection shapes, increase shear or/and moment on vulnerable beam-to-column connections, and
thus making these exiting connections more likely to fail. Consequently, the damping effect these
VWDs could provide would be significantly diminished.

@Seismicisolation
@Seismicisolation
101
On the contrary, a velocity-dependent FVD provides a dampening force that is out-of-phase with
displacement, a much more desirable solution to retrofit existing vulnerable buildings. Moreover,
additional strengthening can be provided easily when connecting FVDs to the beam-to-column
connections by utilizing their end bracings, thus vulnerable connections could be addressed at the
same time as when the dampers are installed.

Tuned Mass Dampers


A tuned mass damper (TMD) is kind of energy dissipation device that employs a mass (steel, concrete
or water) to oscillate out of phase with the building movement. With regard to a TMD system, a FVD
system has the advantages in the following aspects.

No constraints on damper size


In a TMD system, the mass of the damper is specially designed in order to tune the damper’s natural
frequency to that of the building. Nevertheless, when FVDs are used, dampers could come in various
sizes to meet the damping demand in a specific structure, with little constraints on any specific
mechanical properties.

Additionally, later system tuning could be required should the fundamental period of a building
change, e.g., a building softens due to yielding of some members under a large seismic event.

More flexibility to choose locations and configurations


A TMD is usually placed at the top of a building, taking up large portions of valuable top floors and
requiring special design of devices to hang a large damper. Whereas, FVDs could be distributed
at various locations in a building based on the functional and architectural requirements of each
building. For example, if the exterior of a building could not be altered, FVDs could be located inside
the building. Whereas if interference with office use or egress is of a concern, the FVDs could then be
placed outside the building. Moreover, FVDs could be distributed across multiple bays or frames in a
building, with the benefit to reduce accumulated damper forces to beams/columns. Since FVDs are
usually distributed in a variety of places across a building, it also increases the redundancy such that
a few dampers’ malfunctioning would not affect the overall damping effect.

Moreover, manufacturing a TMD is usually complicated and expensive. It is shown that the cost to
producing the device constitutes to about 82% of the entire design and construction procedure
in terms of using TMD [7]. Whereas, the FVDs are not only much cheaper in unit cost, but also the
manufacturing would contribute to a smaller amount in the whole procedure.

In terms of the configuration of placing FVDs, dampers could be arranged in a single diagonal form to
make the construction process simpler. Alternative configurations such as the chevron configuration
or V-shape configuration help maximize a damper’s deformation and could allow for more space to
include doors or windows. On the other hand, if a structure is relative stiff, a toggle-brace-damper
system could be utilized to amplify the damper deformation. If, in some cases the structure is quite
large, dampers could be installed across multiple stories to increase their deformation and thus
energy dissipation capacities. Besides in the superstructure in a building, FVDs are often used in
bridges to control their vibrations when subjected to external excitations. They could also be used in
combination with base isolation systems to reduce the isolator’s displacement or inducing additional
damping. When two structures are closely located, FVDs could be used in the gap between these
adjacent structures to eliminate potential ponding during an earthquake or wind event.

@Seismicisolation
@Seismicisolation
102
Summary
FVDs have found their wide applicability in engineering practice to improve the structural behavior
under a wind or earthquake event. They provide a list of attractive features than other types of
energy-dissipating devices.

When compared with a displacement-dependent device (e.g., BRB, friction damper, and yielding
damper etc.), a FVD could limit the increase of seismic forces, reduce floor accelerations in addition
to reducing story drift ratios, control residual drift ratios and limit interaction with other structural
members.

When compared with a device exhibiting both displacement-dependency and velocity-dependency,


a FVD has limited in-phase components which more efficiently limits the floor accelerations and
interactions with the primary structural system. FVDs are a more appealing solution for the retrofit
of existing buildings. Moreover, a FVD has more stable mechanical properties over a wide range of
frequencies and temperatures when compared with a device using visco-elastic materials.

Additionally, a FVD has more flexibility to select locations, configurations and sizes when compared
to a TMD system.

Reference
[1] Xie, Q. (2005). State of the art of buckling-restrained braces in Asia. Journal of Constructional
Steel Research, 61:727-748.

[2] ANSI/AISC. (2010). Seismic provisions for structural steel buildings. American Institute of Steel
Construction, Chicago, Illinois.

[3] Wang, S. and Mahin, S (2016). Seismic upgrade of an existing tall building by different energy
dissipation devices. Proceedings of 2016 SEAOC Convention, Paper No. 29, Oct. 12-15, 2016,
Maui, Hawaii.

[4] Whittaker, A. S., Bertero, V. V., Alonso, J. L. and Thompson, C. L. (1989). Earthquake simulator
testing of steel plate added damping and stiffness elements. Report No. UCB/EERC-89/02,
University of California, Berkeley.

[5] Constantinou, M.C., Symans, M.D. (1992). Experimental and analytical investigation of seismic
response of structures with supplemental fluid viscous dampers, NCEER-92-0032, Department
of Civil Engineering, State University of New York, Buffalo, NY.

[6] Fu, Y. M. (1996). Frame retrofit by using viscous and viscoelastic dampers, Proceedings of 11th
World Conference on Earthquake Engineering, No. 428.

[7] Tse, T.T.T., Kwok, K.C.S., and Tamura, Y. (2012). Performance and cost evaluation of a smart tuned
mass damper for suppressing wind-induced lateral-torsional motion of tall structures. Journal
of Structural Engineering, 138 (4), 514-525.

@Seismicisolation
@Seismicisolation
103
@Seismicisolation
@Seismicisolation
104
9 The Typical Process for
Incorporating Dampers into
Buildings

Overview
Code prescribed building seismic design implies extensive structural and nonstructural damage after
a design level earthquake, with loss of operational capability and likely major repair or replacement.
By contrast, building seismic design incorporating supplemental damping systems delivers buildings
requiring minimal post-earthquake inspection that are fully operational within hours at little or no
additional cost.

Buildings with a supplemental fluid viscous damping system with between 10 to 40 percent of critical
damping experience over 50% less displacement than a code prescribed building design; up to a 40%
smaller base shear; over a 50% reduction in floor accelerations; and a much lower damage level
due to the larger damping ratio. The structural and nonstructural components of the building are
better protected while reduced displacements and forces mean less steel and concrete offsetting the
damping system cost.

Fluid Viscous Dampers (FVDs) do not change the stiffness of the building, are velocity dependent and
therefore self-centering, have a property variation of less than ± 15%, are easy to install, and require
no maintenance.

SCOPE
The main sections of this document are as follows:
1. Key Aspects of Building Seismic Design incorporating Fluid Viscous Damping Devices:

a. Damping Device Properties;

b. Damping Device Aesthetics;

c. Damping Device Layout;

d. Damping Device Design Review;

e. Damping Device Testing; and

f. Damping Device Procurement.


2. The Optimal Design Process.

@Seismicisolation
@Seismicisolation
105
Key Aspects of Building Seismic Design Incorporating Fluid Viscous
Damping Devices
Fluid viscous damping devices (FVDs) can be incorporated into both new buildings and existing
buildings.

FVDs are placed between any two points where relative motion exists during a transient event such
as an earthquake or wind event. Diagonal brace mountings are popular as is the use of FVDs in a
chevron brace arrangement . Base-isolated structures are also optimized by using FVDs along the
plane of motion in conjunction with the base isolation bearings. For relatively stiff buildings toggle
brace assemblies can be used to magnify small displacements by as much as 5x while simultaneously
producing the required damping force. Many special custom configurations are also available.
Spherical bearings at each end of the FVD permit the damping device to angulate relative to the
structure without binding. These bearings permit the rotation in every direction. In many cases the
spherical bearings may be used at one end, only.

The standard that provides minimum load requirements for the design of buildings and other
structures that are subject to building code requirements, including the Seismic Design Requirements
for Structures with Damping Systems, is ASCE 7-10, Minimum Design Loads for Buildings and Other
Structures, Chapter 18.

ASCE 41-13, Seismic Evaluation and Retrofit of Existing Buildings, Chapter 14, Seismic Isolation
and Energy Absorption, Section 14.3, Passive Energy Dissipation Systems, and ASCE 41-17, Seismic
Evaluation and Retrofit of Existing Buildings, Chapter 15, Energy Absorption, specifies nationally
applicable provisions for the seismic evaluation and retrofit of buildings.

The design of Damping Devices and their connections are sized to resist the force demands,
displacements, and velocities from the maximum considered earthquake ground motions.

Elements of the Damping System are designed to remain elastic for design loads including unreduced
seismic forces of the Damping Devices.

Fluid Viscous Damping Device Properties


Diagonal brace damping devices are available in output force ratings up to 2,000 kips with stroke
capacities up to ± 20 inches. Base isolation damping devices are available in output force ratings up to
2,000 K with stroke capacities up to ± 84 inches. Standard sizes for building applications range from
55 K to 1,800 K with varying strokes from ± 3 inches to ± 5 inches.

A fluid viscous damping device is a velocity dependent device in which the resultant force is
proportional to some power of velocity. FVDs behave by converting kinetic energy into heat, typically
over multiple cycles of response. A FVD dissipates energy by pushing silicone fluid through a custom
orifice geometry to obtain either linear (α = 1.0) or non-linear (α ≠ 1.0) viscous damping.

FVD performance is characterized by the following force/velocity relationship:

F = CVα
where: F is the Damping Force, in pounds
V is the Relative Velocity, inches per second
C is the damping Coefficient, constant
α is the Velocity Exponent, constant

@Seismicisolation
@Seismicisolation
106
As α decreases the damping efficiency increases.

For example, and FVD with an α = 1.0 has a damping efficiency of 78.5%, while an FVD with an
α = 0.1 has a damping efficiency of 97.1%.

A low damping α provides a more consistent damping force over a larger range of input velocities.

For example, an FVD with α = 1.0 provides 2x the damping force if the actual velocity input is 2x the
design velocity. For an FVD with α = 0.1, if the actual velocity input is 2x the design velocity, the FVD
still provides approximately 1x the damping force.

Values from 0.3 to 0.9 typically work best for building applications. The standard α for starting a
building design is 0.5.

Nominal design properties of FVDs shall be established from either project-specific prototype test
data or prior prototype tests on a device of similar size and construction. These nominal design
properties are confirmed by prototype tests later in the design or construction phase of the project.
The use of a bounded analysis addressing the expected variation in nominal design properties enables
the design process for structures with FVDs to proceed in a similar fashion to a conventional project.

The typical practice for FVDs sourced from manufacturers, is to use upper-bound design property =
1.15× nominal design property and lower-bound design property = 0.85× nominal design property to
cover variations caused by specification, device characteristics, environment, and aging. The variation
is typically applied only to the C value and not to the α value. To provide an additional margin of safety,
ASCE 41-13, 14.3.1 requires FVDs to be capable of sustaining larger displacements, velocities, and
forces than the maxima calculated for the Design Basis Earthquakes.

Fluid Viscous Damping Device Aesthetics


Damped bays can either concealed or exposed. Two types of Taylor FVDs can be supplied, an FVD
with a spherical bearing at one end and a flange plate at the opposite end, or an FVD with spherical
bearings at each end and an integral extender in between the spherical bearings.

For example, in a chevron application either a flanged tube steel section or pipe section damping
device driver is bolted to the flange end of the damping device to achieve the pin to pin dimension
of the brace element. For an FVD with spherical bearings at each end and an integral extender in
between the spherical bearings the supplied device extends the full pin to pin dimension.

The aesthetic appearance of an FVD with an integral extender in an exposed condition may be more
architecturally pleasing and should be considered in the early layout stage.

Fluid Viscous Damping Device Layout


Possible layouts are usually dictated by the architectural form and function of the building. FVD
locations on the perimeter of buildings are the most effective. However, locations that may be
architecturally exposed or concealed on the interior of the building are also common. More damping
devices may be required for an interior layout compared to an equivalent perimeter layout.

Each damped story should have at least four energy dissipation devices in each principal direction of

@Seismicisolation
@Seismicisolation
107
the building, with at least two devices located on each side of the center of stiffness of the story in
the direction under consideration. When less damping devices are used, an increase in displacement
and velocity capacity is required by code due to the reduced level of redundancy in the damping
system. The damped stories may not need to extend the full height of the building depending upon
the building configuration.

In the case of a soft story application, the damping devices may be located at the soft story, only,
however overall building performance is improved if the damping devices are additionally located
either one or two levels above the soft story.

Damped bays do not always stack in adjacent stories. The load path should be complete but flexibility
in the location of the damped bays is an advantage in laying out the supplemental damping system.

The use of a larger number of smaller damping devices will be more cost effective than a smaller
number of larger damping devices.

Fluid Viscous Damping Device Design Review


In 1993, the Energy Dissipation Working Group (EDWG) of the Base Isolation Subcommittee of
the Structural Engineers Association of Northern California (SEAONC), in response to the demand
for design guidance for passive energy dissipation systems, developed a document that proposed
tentative design requirements. These requirements were applicable to a wide range of system
hardware and recommended a testing program to verify device performance. The scope included
metallic, friction, viscoelastic, and viscous devices.

The general philosophy of the EDWG document was to confine inelastic deformation primarily to
the energy dissipation devices, while the main structural members remained elastic for the Design
Basis Earthquake. Furthermore, since passive energy dissipation technology was still relatively
new, a conservative approach was taken on many issues. For example, an experienced independent
engineering review panel was required for all projects to conduct a review of the energy dissipation
system design and the associated prototype testing programs.

A simpler approach was included as Appendix to Chapter 2 of FEMA 222A NEHRP Recommended
Provisions for Seismic Regulations for New Buildings, 1994 Edition, also recommending a testing
program similar to that proposed by EDWG.

Concurrently a significant effort funded by FEMA was underway to create technical guidelines for
the seismic upgrading of buildings. Energy dissipation systems were included in the range of available
techniques to improve seismic performance. The results of these efforts were published in Chapter 9
of FEMA 273 NEHRP Guidelines for the Seismic Rehabilitation of Buildings, October 1997.

This guideline was more extensive than the EDWG guideline and was more extensive than the
FEMA 222A approach, but it could not be referenced or quoted for the proposed FEMA 302 NEHRP
Recommended Provisions for Seismic Regulations for New Buildings and Other Structures, 1997
Edition, because FEMA 273 had not been published or made generally available at the time FEMA
302 went to ballot.

As a result, the FEMA 302 Appendix to Chapter 13, Passive Energy Dissipation Systems, only provided
brief statements as to the benefits of damping for improved performance, suggested rational design
procedures be used, and recommended an independent panel for design and test program review.

@Seismicisolation
@Seismicisolation
108
It was recognized that this Appendix was only a placeholder for more thorough requirements in the
2000 edition of FEMA 368.

The Appendix to Chapter 13 of FEMA 368 NEHRP Recommended Provisions for Seismic Regulations
for New Buildings and Other Structures, 2000 Edition, Structures with Damping Systems, was
published in March 2001. It was intended to be applicable to all types of energy dissipation systems,
to provide design criteria comparable to conventional design performance, to provide design criteria
for enhanced seismic performance, to distinguish between the design of members that are part of the
energy dissipation system and the design of members independent of that system. It also required an
independent engineering review of the design and testing programs.

Since then, many of the provisions have been updated and incorporated into the current codes and
standards, however the independent panel for design review remains. Although, the number of panel
reviewers now is specified as “one or more”.

As described in ASCE 7-10, Section 18.8 or ASCE 14-13, Section 14.3.7, a design review of the damping
system and related test programs shall be performed by an independent team of registered design
professionals in the appropriate disciplines and others experienced in seismic analysis methods and
the theory and application of energy dissipation systems. The design review shall include, but need
not be limited to, the following:
1. Review of site-specific seismic criteria including the development of the site-specific
spectra and ground motion histories and all other project-specific design criteria;
2. Review of the preliminary design of the seismic force resisting system and the
damping system, including design parameters of the damping devices and prototype
testing;
3. Review of the final design of the seismic force-resisting system and the damping
system and all supporting analyses; and
4. Review of damping device test requirements, device manufacturing quality control
and assurance, and scheduled maintenance and inspection requirements.

In the early applications of energy dissipation, many design review panels included three individuals
to cover the range of expertise required in the design review, including the site-specific seismic and
other criteria and the prototype testing of the devices. Design review may now be performed by just
one individual. For more significant structures, a local jurisdiction may require a design review panel
with two or three individuals, but for many structures incorporating energy dissipation devices, one
well qualified and experienced design reviewer is adequate.

Although review of the prototype test program is mandated, the design reviewer is no longer required
to witness the prototype tests.

The design review is not a difficult, timely or costly process. Many suitably qualified design
professionals are available to perform the review. If required, Taylor Devices, Inc. can provide a list of
design professionals that have served in this capacity on previous projects.

Fluid Viscous Damping Device Testing


The force–velocity–displacement relations and damping values assumed as the nominal design
properties of the FVD shall be confirmed by tests before production of devices for construction, or
they shall be based on prior tests of devices of a similar size, as described in ASCE 7-10, Section 18.9
or ASCE 14-13, Section 14.3.2.

@Seismicisolation
@Seismicisolation
109
The fabrication and quality control procedures used for all prototype and production devices shall
be identical. These procedures shall be approved by the design professional before the fabrication of
prototype devices.

Prior to installation in the building, damping devices shall be tested to ensure that their force-
velocity- displacement characteristics fall within the limits set by the registered design professional
responsible for the design of the structure. The scope and frequency of the production-testing
program shall be determined by the registered design professional responsible for the design of the
structure.

Fluid Viscous Damping Device Procurement


Damping devices are manufactured and fabricated to meet the specific requirements of each project.
Although, some of the components are in stock for certain sizes of devices, the availability of the
damping devices should be verified for each project.

The estimated time of FVD production should always be verified with Taylor Devices. The production
time is heavily dependent on the orders being produced at the time of request and the size of the
damping device.

The critical information required to place an order or to receive a quotation includes but is not limited
to, the damping device size, stroke, C value, and α value.

Taylor Devices typical payment terms are 1/3 deposit due with order placement, 1/3 due when
machining parts begins and 1/3 due when units are ready for shipment. The duration of FVD
production commences with the receipt of the first payment installment.

The Optimal Design Process


The optimal design process should incorporate and balance all the issues summarized in the previous
section.

It is recommended that the registered design professional responsible for the design of the structure
develop and subsequently manage a design process schedule.

At the inception of the project meet with the authority having jurisdiction (AHJ) and determine
their design review requirements, process, and timeline. The design review may be conducted by
the AHJ. Alternatively, the review is performed by an outside consultant(s). The consultant(s) may
be a bench consultant(s) already under contract with the AHJ or may be hired specifically for the
project requiring an RFP and selection process. It is important to gain a clear understanding of the
arrangement and timing of the Design Review as soon as possible.

Typically, the design review is conducted by two consultants, one to review the site seismic hazards
and associated geotechnical design parameters and one to review the design and other aspects of the
damped seismic force resisting system (DSFRS). However, each AHJ may have a different approach
to the design review.

The design review should be conducted concurrently with the design. Early approval of the site
seismic hazards and associated geotechnical design parameters, the approach to establishing the
nominal design properties through prototype testing or prior tests of devices of a similar size, and
the Basis of Seismic Design of the DSFRS document is recommended.

The layout of the proposed damping device bay configuration should be finalized, and the design
completed.

@Seismicisolation
@Seismicisolation
110
A phased review to the structural design of the project is preferred. The design of the DSFRS can be
separated from the balance of the project. Once the design of the DSFRS has been accepted,

the damping devices can be ordered. The earlier the damping devices can be ordered the more
flexibility in the construction phase schedule will be realized.

With the implementation of a phased review, no additional design time compared to a traditional
design project is incurred.

The AHJ may not agree to a phased review so it is important to reconcile the timing of the design
schedule, the acceptance of the design and the delivery of the damping devices at the jobsite as early
as possible.

Taylor devices tests 100% of their units prior to shipment to the job and these tests can be witnessed
by the registered design professional responsible for the design of the structure. Current typical
practice is to perform quality control testing on all devices, but there is no codified requirement to
do so.

Definitions
Damping System: The collection of structural elements that includes all the individual damping
devices, all structural elements or bracing required to transfer forces from damping devices to the
base of the structure, and the structural elements required to transfer forces from damping devices
to the seismic force-resisting system.

Damping Device: A flexible structural element of the damping system that dissipates energy due to
relative motion of each end of the device. Damping devices include all pins, bolts, gusset plates, brace
extensions, and other components required to connect damping devices to the other elements of the
structure.

References
FEMA 222A, 1994, NEHRP Recommended Provisions for Seismic Regulations for New Buildings,
1994 Edition, Federal Emergency Management Agency, Washington, DC.

FEMA 273, 1997, NEHRP Guidelines for the Seismic Rehabilitation of Buildings, Federal Emergency
Management Agency, Washington, DC.

FEMA 302, 1998, NEHRP Recommended Provisions for Seismic Regulations for New Buildings and Other
Structures, 1997 Edition, Federal Emergency Management Agency, Washington, DC.

FEMA 368, 2001, NEHRP Recommended Provisions for Seismic Regulations for New Buildings and Other
Structures, 2000 Edition, Federal Emergency Management Agency, Washington, DC.

American Society of Civil Engineers (ASCE) 2013, ASCE/SEI 7-10, Minimum Design Loads for
Buildings and Other Structures, Reston, VA.

American Society of Civil Engineers (ASCE) 2014, ASCE/SEI 41-13, Seismic Evaluation and Retrofit of
Existing Buildings, Reston, VA.

American Society of Civil Engineers (ASCE) 2018, ASCE/SEI 41-17, Seismic Evaluation and Retrofit of
Existing Buildings, Reston, VA.

@Seismicisolation
@Seismicisolation
111
@Seismicisolation
@Seismicisolation
114
112
10 Photographs

A. Dampers in Diagonal Braces

Colorado Place – Santa Monica, CA - Retrofit

999 Sepulveda – El Segundo, CA - Retrofit

@Seismicisolation
@Seismicisolation
113
Bremerton Naval Hospital – Bremerton, WA - Retrofit Close-up of a damper at 999 Sepulveda - Retrofit

Exterior dampers at 999 Sepulveda – Retrofit

@Seismicisolation
@Seismicisolation
114
San Francisco Civic Center Complex – San Francisco, CA – New Build

Jinbocho Office Building – Tokyo, Japan – New Build

@Seismicisolation
@Seismicisolation
115
San Bernardino Justice Center – San Bernardino, CA – New Build

CSUS AIRC – Sacramento, CA – New Build

@Seismicisolation
@Seismicisolation
116
Caltrans District 4 Office Building – Oakland, CA - Retrofit

Glass Factory – Taichung, Taiwan - Retrofit

@Seismicisolation
@Seismicisolation
117
San Miguel Mall – Lima, Peru – New Build

Amgen Building – Seattle, WA – New Build

@Seismicisolation
@Seismicisolation
118
Public Safety Building – Salt Lake City, UT – New Build

Cal Poly Pomona Library – Pomona, CA – Retrofit

@Seismicisolation
@Seismicisolation
119
B. Dampers in Chevron Braces

Portland State University Student Union – Portland, OR – Retrofit (Ext)

Portland Galleria – Portland, OR – Retrofit

@Seismicisolation
@Seismicisolation
120
Patio Mayor Building – Santiago, Chile – Retrofit Hotel Woodland – Woodland, CA – Retrofit

Nordstrom Southcoast Plaza – Costa Mesa, CA – Retrofit

@Seismicisolation
@Seismicisolation
121
C. Base Isolation Dampers

KDDI Building – Tokyo, Japan – New Build

LARTMC - Glendale, CA – New Build

@Seismicisolation
@Seismicisolation
122
Hayward City Hall – Hayward, CA – New Build

Arrowhead Regional Medical Center – San Bernardino, CA – New Build

@Seismicisolation
@Seismicisolation
123
Tokyo Rinkai Hospital – Tokyo Japan – New Build

Fujisawa City Hall – Fujisawa, Japan – New Build

@Seismicisolation
@Seismicisolation
124
D. Dampers with Motion Amplification Devices

111 Huntington – Boston, MA – New Build

Yerba Buena Tower – San Francisco, CA – New Build

@Seismicisolation
@Seismicisolation
125
Olympic Building – Cyprus – New Build

E. Bridge Dampers

San Francisco- Oakland Bay Bridge - Retrofit

@Seismicisolation
@Seismicisolation
126
Hangzhou East Bridge – China – New Build

Santiago Creek Bridge – Irvine, CA – Retrofit

@Seismicisolation
@Seismicisolation
127
Tianjin Railway Bridge – Tianjin, China – New Build

Gang Hwa Bridge – S. Korea – Retrofit

@Seismicisolation
@Seismicisolation
128
Las Vegas Pedestrian Bridge TMD – Las Vegas, NV – New Build

Millennium Bridge – London, UK

@Seismicisolation
@Seismicisolation
129
F. Miscellaneous Damper Installations

Toranomon 5 Office Building – Tokyo, Japan – New Build

Safeco Field – Seattle, WA – New Build

@Seismicisolation
@Seismicisolation
130
181 Fremont Street – San Francisco, CA – New Build

10th & K Street – Sacramento, CA - Retrofit

@Seismicisolation
@Seismicisolation
131
Ichilov Hospital – Israel – Retrofit

@Seismicisolation
@Seismicisolation
132
11 Mounting Hardware

Taylor Devices fluid viscous dampers are provided with a spherical bearing in each end that provides at
least ±5° of rotation in both directions to account for misalignment of installation and for movement
during an event. This misalignment can be seen in Figure 11.1.

Figure 11.1

@Seismicisolation
@Seismicisolation
133
In order to maintain the ±5° rotation, it is important to make sure that no surrounding geometry
interferes with the movement of the clevises. A common interference can come from the bracket that
is housing the end of the damper. To prevent this, it is important for the bracket manufacturer and the
damper manufacturer to coordinate their efforts. Taylor Devices has produced the following chart
drawing (Figure 11.2) to aid in the design and manufacture of brackets that will allow ±5° minimum
rotational clearance.

Figure 11.2

If a gusset plate already exists at the installation location, Tang Plates can be a good substitute for a
bracket. With tang plates, the pin, shims and tang plates can be assembled prior to lifting the damper
into position. With the tangs attached to the damper, only nuts, bolts and washers need to be used to
attach the assembly to the structure. An example of a tang plate can be seen in Figure 11.3.

To ensure that the bearing stays in the middle of the bracket or tang plates, Taylor Devices provides
two shims along with each pin kit. The shims are to be placed on either side of the bracket such that
it always has enough space on either side to rotate and provide the required 5° of rotation. The
assembly can be seen in Figure 11.4.

@Seismicisolation
@Seismicisolation
134
Figure 11.3

Figure 11.4
Pin Kit Arrangement and Installation Details

@Seismicisolation
@Seismicisolation
135
To facilitate a close fit and ensure that the dampers can function at smaller displacements, Taylor
Devices holds a very tight tolerance on its bearings and pin kits. It is common to have tolerances of
±0.001 inches on pin kits. Because the tolerance is so tight, bracket holes must be bored after any
welding of plates to ensure the holes are perfectly aligned. Examples of some typical pin kits can be
seen in Figure 11.5 and 11.6.

Figure 11.5

Taylor Devices strives to meet its customer's needs and will design unique solutions to unique
problems. If there are concerns that a major earthquake could hit during the construction or retrofit
of a building, Taylor Devices can provide a system that can be activated within a short period of time.
Using our Pin-in-a-Pin system, the dampers can be installed in the structure with a thin walled pin
that is designed to hold the unit in place but would buckle under the load from an earthquake. Once
all dampers are installed, it is a simple manner to quickly go to each installation site and slide a solid
center pin into the thin walled pin activating the system and providing instant damping protection.
An example of one of these thin walled/solid core pins can be seen in Figure 11.7.

@Seismicisolation
@Seismicisolation
136
Figure 11.6

Figure 11.7

@Seismicisolation
@Seismicisolation
137
Taylor Devices provides two different systems for retaining pins in the assembly. The first is a
retaining ring system that are wound onto each end of the pins. The second system includes washers
and cotter pins. Examples of the two systems can be seen in Figure 11.8, 11.9, and 11.10.

Figure 11.8
Pin Kit Retaining Ring Assembly

@Seismicisolation
@Seismicisolation
138
TYPICAL PIN KIT WITH COTTER PINS

Figure 11.9
Un-installed position

Figure 11.10
Installed position

@Seismicisolation
@Seismicisolation
139
@Seismicisolation
@Seismicisolation
140
12 Available Sizes and Dimensions

Depicted below are four Taylor Devices’ brochures that represent the available sizes and dimensions
for our Fluid Viscous Dampers and Lock-Up Devices.

Figure 12.1
Dampers with Clevis – Base Plate Configuration (Metric)

@Seismicisolation
@Seismicisolation
141
Figure 12.2
Dampers with Clevis – Base Plate Configuration

@Seismicisolation
@Seismicisolation
142
Figure 12.3
Dampers with Clevis – Clevis Configuration (Metric)

@Seismicisolation
@Seismicisolation
143
Figure 12.4
Dampers with Clevis – Clevis Configuration

@Seismicisolation
@Seismicisolation
144
13 Case Studies

The following is a list of select case studies. Full versions of the studies are located in Appendix A.

1. Cost Delta for Achieving Higher Structural Performance Levels – Major cities in California are
now mandating that existing buildings with the highest seismic risks be retrofitted in conformance
with local ordinances. As a consequence to the recent engineering community’s push for cities
and building owners to become more resilient against seismic events, many owners are now more
conscious about their buildings’ anticipated seismic performance. Owners are now beginning
to ask engineers what it means to design above the minimum code standards. For an owner to
make an educated decision on building design, engineers need to convey the increased cost of a
higher structural performance in simple terms. This paper covers a scenario in which an existing
Pre-Northridge Steel Moment Frame building was evaluated for three different performance
objectives under California’s hospital building standards. This paper highlights the differences
in structural scope between each performance level as well as the expected percent increase
in construction costs. Engineers can use this case study as an example when speaking to their
clients about relative costs between different seismic performance levels.

2. Seismic retrofit and FEMA P-58 risk assessment of mid-rise soft-story concrete towers – The
two frame office towers, constructed in the 1970s per the 1967 edition of the UBC, use perimeter
reinforced concrete moment frames to resist seismic loading. The buildings are rectangular
in plan and have certain characteristics that adversely affect their seismic performance, in
particular the presence of a soft-story response at the first floor (approximately 50% taller than
typical floors), and limited ductility typical of buildings of that era. Risk analysis showed that for
the towers the PML exceeded 20%. Nonlinear response history analysis (NLRHA) of the towers
was conducted and showed that in the existing configuration, the story drift ratios (SDRs) at
the first floor exceeded 2%, shear hinging of the first floor beams was expected and that the
SDRs would need to be reduced to approximately 1.4% for the first floor to limit the extent
of nonlinear response. Seismic retrofit included addition of 300-kip viscous dampers in both
directions to the first floor of the building. Analysis showed that the retrofitted structure had a
first floor SDR of approximately 1.3% and that the soft story response and plastic shear hinging
of first floor beams were mitigated. FEMA P-58 analysis of the retrofitted buildings were then
conducted using the results—SDR, story acceleration, and residual drifts—from the NLRHA. It
was seen that the 90th percentile repair cost (PML) was significantly reduced and was now less
than 15%.

3. Seismic Rehabilitation of Historic Concrete Structure With Fluid Visco-Elastic Dampers – This
paper presents the nonlinear seismic analysis, development, and implementation of an innovative
seismic retrofit strategy for a six-story nonductile reinforced concrete 145,000-sf (13,470 m2)
historic building. Dynamic and nonlinear static analytical results verified that the building had a

@Seismicisolation
@Seismicisolation
145
weak soft-story with inadequate post-yield capacity, and large torsional response. The analysis
indicated that the existing building is not seismically adequate to withstand anticipated lateral
forces generated by earthquake excitations at the site. A “collapse prevention” performance
upgrade for a 475-year return event was desired. Nonlinear fluid viscous dampers were placed
at the first story level to reduce the seismic demand and obtain a more uniform response. Visco-
elastic fluid viscous dampers were strategically placed at one side of the building to reduce the
torsional irregularity of the building. The proposed cost effective, state-of-the-art retrofit will
improve the seismic performance of the building.

4. Seismic Upgrade of an Existing Tall Building by Different Energy Dissipation Devices – The
Pacific Earthquake Engineering Research (PEER) Center has expanded its Tall Building Initiative
(TBI) program to include the seismic performance of existing tall buildings. A 35-story steel
moment resisting frame, designed in 1968, and had representative details of buildings between
1960 to 1990 was selected for detailed seismic evaluation in the framework of Performance
Based Earthquake Engineering (PBEE). It was identified that the case study building failed
to meet the performance objectives suggested by ASCE 41-13, and had a number of seismic
vulnerabilities that endangered its structural integrity at two basic safety earthquake hazard
levels (BSE): BSE-1E and BSE-2E. Therefore, exploration of retrofit strategies and their cost-
effectiveness are fostered. In this paper, three kinds of supplemental energy dissipation devices
are investigated to upgrade the seismic performance of the case study building, including fluid
viscous dampers (FVDs), viscous wall dampers (VWDs) and buckling restrained braces (BRBs).
The retrofit design started by selecting locations to install supplemental devices. Then the total
effective damping ratios needed to achieve the target roof displacements in two directions
were estimated based on a damping scale factor (DSF). One retrofit strategy by using FVDs
was investigated as a first trail, and the mechanical characteristics of each damper device were
calculated based on the overall effective damping ratio and the story wise distributions of
dampers. Next, other two retrofit strategies by using VWDs or BRBs were investigated. Sizing
of different devices at one location was performed following the principle of equal energy
dissipation. The effectiveness of each strategy to meet the retrofit intent of ensuring structural
stability at BSE-2E were compared. Moreover, probabilistic damage and loss analysis were
conducted using Performance Assessment Calculation Tool (PACT) to relate the structural
responses to economic losses. After a detailed examination, it was found that upgrading the
case study tall building using FVDs was the most effective retrofit strategy to control structural
responses, and reduce damage and economic losses after BSE-2E events.

5. Practical Implementation of ASCE-41 and NLRHA Procedures for the Design of the LLUMC
Replacement Hospital – The Loma Linda University Medical Center Campus Transformation
Project (LLUMC CTP) is a new 17 story base-isolated 1,000,000 square foot replacement
acute care hospital located less than 1 km from the San Jacinto Fault. The seismic design and
analysis of the structure used LS-DYNA to efficiently perform nonlinear response history
analysis (NLRHA) with 110 individual ground motion analyses incorporating DE, MCE, upper
bound, lower bound, and varying ground motion direction. Implementation of triple pendulum
isolators, fluid viscous dampers, buckling restrained braces, and SidePlate moment frames in LS-
DYNA will be described. As required for OSHPD-1 facilities, the NLRHA results demonstrated
Immediate Occupancy performance at DE and Life Safety performance at MCE using element
backbone curves and acceptance criteria from ASCE 41 as amended by the California Building
Code. Inconsistencies in element acceptance criteria for combined lateral systems and other
code implementation challenges will be discussed. A cloud computing and database framework,
using Penguin-on-Demand and Amazon Web Services, was developed to manage the 8
terabytes of data generated from each set of 110 ground motion analyses performed on each

@Seismicisolation
@Seismicisolation
146
design iteration. Automated processes enabled the team to reduce the time between design
iterations to 2 weeks for the complete suite of NLRHA, post-processing, report generation, and
design optimization. The team’s approach to analysis data management, design optimization
procedures based on NLRHA results, automated post-processing, and automated report
generation will be detailed.

6. Viscous Damper with Motion Amplification Devices for High Rise Building Applications –
Adding damping by the use of various damping devices has become an accepted method to
reduce wind-induced vibrations in tall buildings. An interesting example of a 39-storey office
tower is presented where large projected accelerations are the result of the vortex shedding
off an adjacent existing 52-storey building. Viscous dampers and a motion amplification system
are utilized to suppress the anticipated accelerations. A description of the damping system and
its analytical complexities are discussed. Non-linear analysis of the tower, with time history
forcing functions, derived from the wind tunnel, is presented. The dichotomy of the wind and
seismic design requirements on the damping system is discussed. Cost data for the system is
also presented.

7. Integrated Design and Construction at the 250 West 55th Street Tower – The recently
completed 40 story office tower at 250 west 55th street in Manhattan demonstrates
the best in innovative structural design, and use of 3-D coordination tools for design
and construction. This paper describes the integrated process that was followed
and some of the challenges that were met along the way, and will be of interest to
design professionals and others interested in integrated construction processes.

The integrated 3-D process started with the use of Revit from the Schematic design stage, and
was followed through the design, procurement, and construction phases, with all major sub
contractors producing 3-D or 4-D models. These models were carefully integrated by the general
contractor, and enabled savings in schedule, reduction of field conflicts, and reduced project risks.

The progress of the project was further complicated by a suspension of construction for two
years after completion of the foundations. This paper describes some of the unusual steps taken
to manage this process and allow for an accelerated schedule upon restart of construction.

8. Design of an Essential Facility with Steel Moment Frames and Viscous Dampers Using 2000
NEHRP – This new 2-story, 40,000ft2 police headquarters becomes the first building in the
United States to apply 2000 NEHRP procedure to design an essential facility with Fluid Viscous
Dampers (FVDs). The structure is located in Vacaville, California, which is in a region of high
seismic activity and classified as zone 4 per 1997 Uniform Building Code. The lateral force
resisting system (LFRS) consists of special steel moment frames with FVDs. In accordance with
2000 NEHRP, the LFRS is sized and designed with strength requirements of the code level force.
FVDs are provided to control displacement of the structure. This design philosophy leads to a
low frequency structure with low acceleration. FVDs reduce the displacement level to less than
0.01 story drift ratio. Earthquake performance and cost effectiveness are the primary concerns
in designing this building. Site specific response spectra and time histories are synthesized for
a 500-year and a 2,500-year return event. Performance Based Design using both linear and
nonlinear time history analyses is conducted to ensure “immediate occupancy” performance. A
cost study shows that much of the FVD’s cost is offset by reducing the weight of the LFRS while
providing a far superior performance than the “code-compliant” structures.

9. Seismic Retrofit of the Tower of Hope – Preservation of a Masterwork of Mid-Century Modernism


- Richard Neutra’s iconic Tower of Hope on the Christ Cathedral (formerly “Crystal Cathedral”)

@Seismicisolation
@Seismicisolation
147
campus in Garden Grove, California has been an important Orange County landmark since it was
built in 1968. The thirteen-story tower – the tallest building in Orange County when it was built
– has been called an “overlooked masterwork in Neutra’s oeuvre” by architectural historians.

Like many concrete buildings built prior to the 1971 Sylmar Earthquake in California,
the Tower of Hope’s concrete frames lack the ductility needed to safely dissipate seismic
energy. After acquiring the Crystal Cathedral campus in 2012 the Roman Catholic Diocese
of Orange undertook a comprehensive renovation and seismic retrofit project to provide
21st century seismic resilience to the historic tower. This challenging seismic retrofit and
renovation project was completed in 2015. The retrofit work included the installation
of fluid viscous dampers on the second through fifth floors of the tower in combination
with fiber-reinforced polymer strengthening of targeted concrete columns and walls.

This paper focuses on two challenges unique to the Tower of Hope. First, it was imperative
that the retrofit design respect the historically significant mid-century modernist architecture,
preserving those features that were emblematic of that period of significance. Seismic retrofit
construction was limited to areas that didn’t affect Neutra’s open floor plate design aesthetic
or lessen the inside-outside connectivity of each of the spaces. This openness was particularly
challenging to preserve in the glass-walled first floor lobby where seismic forces are at their
most intense. The second unique challenge was the large damper connection forces that had
to be developed into the existing cast-in-place concrete frames without damaging the existing
steel rebar. The strategies described by the authors are generally applicable to other historic
buildings from the mid-century modernist movement and to the use of fluid viscous dampers to
retrofit concrete frames.

10. Strong Medicine – An article from Modern Steel Construction magazine discussing the use of
viscous dampers for the seismic upgrade of the Naval Hospital located Bremerton, WA.

@Seismicisolation
@Seismicisolation
148
14 Taylor Devices’ Literature

Structural Protection Products Brochure (8 pages).............................152-159

Protect Your Investment Brochure (4 pages)..........................................160-163

Seismic Dampers—Buildings (2 pages)....................................................... 164-165

Seismic Dampers—Bridges (2 pages)........................................................... 166-167

Metal Bellows Dampers (2 pages)................................................................. 168-169

Open Space Damper System (2 pages).......................................................170-171

@Seismicisolation
@Seismicisolation
149
Structural Protection Products

Since 1955

@Seismicisolation
@Seismicisolation
150
We’re on top of structural control

250 West 55th Street


7 custom high capacity metal bellows
dampers at the top of the building in
an outrigger conԴguration
Output Forces:  lbs
Stroke Capacity:   inches
Owner: Boston Properties
Architect: Skidmore, Owings &
Merrill LLP
Engineer: ARUP Group
General Contractor: Turner
Steel Contractor: Owen Steel
Company

@Seismicisolation
@Seismicisolation
151
As The World Churns: A New Level of Seismic Protection

Originally developed for NASA in the 1960s, Taylor Fluid Viscous Dampers are now used by civil
engineers worldwide to protect their work and the people who rely upon its safety Over 600
buildings, bridges and other crucial structures around the world are now using Taylor dampers

Taylor Devices  a 01 inductee into the Space Technology +all of Fame  has been recogni]ed
and certiԴed by NASA for these space program innovations that now save lives on (arth

Taylor Devices has established a track record for reliability and stateoftheart performance
Our unshakeable reputation has been validated through rigorous research, full scale testing,
and published work by organi]ations such as
■ The Multidisciplinary Center for (arthTuake (ngineering Research MC((R
along with S((SL & N((S at the State University of New <ork at Buffalo
■ (arthTuake (ngineering Research Center ((RC at UC Berkeley
■ The +ighway ,nnovative Technology (valuation Center +,T(C
Taylor Devices’ Seismic Dampers and our
■ The National Center for Research on (arthTuake (ngineering NCR(( located in Taiwan
President, Douglas Taylor, were inducted into
■ The Caltrans Seismic Response ModiԴcation Device SRMD test facility at UCSD
the Space Technology +all of Fame in 201
Our Vice-President, Richard +ill, received the ■ (Defense test facility located in -apan
Corporate Award
Why seismic dampers?
There are simply no disadvantages to using Taylor Seismic Dampers for structural control
due to their viscous behavior Building codes throughout the world now recogni]e and even
encourage their use A straightforward dynamic analysis of a structure using standard software
demonstrates the signiԴcant beneԴt of seismic dampers, and the technology is now being
widely used to optimi]e structures Taylor Fluid Viscous Seismic Dampers provide a reliable,
ine[pensive, environmentallyfriendly and easytoinstall solution

How seismic dampers work


Taylor Seismic and Wind Dampers are available in output force ratings up to 2 million pounds
and stroke amplitude capacities up to 2 inches These highly efԴcient devices absorb a
tremendous amount of energy during an earthTuake in order to prevent that energy from ever
reaching the structure itself This earthTuake energy is transferred to heat energy and is safely
dissipated after the event is over This allows designers to reduce the cost of the structure by
utili]ing smaller structural elements and less comple[ foundations while improving the dynamic
performance of the structure

Using seismic dampers


Taylor Seismic Dampers are easy to incorporate into a structure Dampers can be placed
between any two points where relative motion exists during a transient event such as an earth-
Tuake or wind event Diagonal brace mounting is popular, as is the use of dampers in a chevron
brace arrangement Base-isolated structures are also optimi]ed by using seismic dampers along
the plane of motion in conMunction with base isolatorsbearings Special custom conԴgurations
are also available

Dampers that produce a linear relationship with respect to velocity are available Please note
that standard damping exponents can be set anywhere between  and 20, as reTuired by the
speciԴc application

@Seismicisolation
@Seismicisolation
152
Seismic Dampers
Ready, Steady, Reliable.
Seismic dampers soak up earthTuake-
induced motion to prevent structural
damage Compact yet powerful, our Եuid
viscous dampers increase structural damping
levels to as much as 0 percent of critical

Our Եuid viscous dampers are based on


aerospace technology developed for the
MX missile and B-2 Spirit “Stealth
Bomberϱ These dampers have been
proven in extensive tests at internationally
accredited laboratories Along with a
standard  year warranty, these dampers
utili]e dependable machined Եuid Եow
passages for ultimate reliability
and longevity

San Bernardino Justice Center


1 Seismic Dampers throughout structure
Output Force: 0,000 lbs
Stroke Capacity: -  inches
Engineer: Skidmore, Owings & Merrill LLP
General Contractor: Rudolph & Sletten
Steel Contractor: Schuff Steel

Protect new or existing structures against earthTuakes simply and inexpensively Taylor Devices’ Fluid
Viscous Dampers provide complete protection for buildings, bridges, towers, elevated freeways -
virtually any structure that is subMect to earthTuake damage
■ Diagonal brace dampers are available in output levels up to 2 million pounds force
with strokes of up to 20 inches
■ Base isolation dampers are available in output levels up to 2 million pounds force
with strokes of up to  inches

For unusual seismic applications, motion ampliԴcation devices and open space frames can be
incorporated These products also have non-seismic applications as wind dampers on tall buildings,
bridges and other structures

Taylor Seismic Dampers reTuire no maintenance - ever Our exclusive design uses a minimum number
of moving parts and our patented dry seal has over 0 years of successful performance in some of
the world’s most demanding applications

@Seismicisolation
@Seismicisolation
153
Tuned Mass Dampers

Against the Wind.


Tuned Mass Dampers TMD are often
used in tall buildings and consist of a
large, internally-lumped mass suspended
from the uppermost Եoors The mass is
supported with cables, steel arms
or springs combined with airԵuid
mechanical slider bearings The mass
is centered within the building on these
lateral spring elements

Fluid dampers inset photo which


typically have long strokes in the 2 to 6
feet range, control the response of the
tuned mass When the building is
subjected to wind inputs, the building
sways from side to side The tuned
mass, due to its inertia and long period
attachment, moves out of phase with
the building (nergy is dissipated by
the Եuid dampers

432 Park Ave. - New York City


16 Long Stroke Taylor Dampers for Tuned
Mass Dampers at the top of the building
Damper Force: 30,000 lbs
Stroke Capacity: - 0 inches
Damper Power Capacity: Greater than 0 +P
Wind Consultant and TMD Designer: RWD,
Motioneering

The tuned mass damper reduces deԵection of the building under wind inputs by essentially
applying force to the building in a direction always opposite to the wind-induced motion

Tuned mass dampers are not normally used to provide seismic protection

TMDs are also in pedestrian bridges, ballrooms, stadiums and other structures to reduce vertical,
lateral andor torsional vibrations caused by human activities andor wind

This photograph shows a 7,600 lb TMD designed and manufactured


by Taylor Devices used to mitigate lateral movements on a long
pedestrian bridge The mass is suspended by steel cables, tuned
to the lateral freTuency of the bridge, and is controlled by Եuid
dampers One of these dampers is shown in the foreground with
heat dissipating Դns on its outside diameter

Vertical vibrations on this bridge were mitigated by similar tuned


mass dampers that were mounted on springs and designed to move
out of phase in a vertical orientation

@Seismicisolation
@Seismicisolation
154
Lock-Up Devices
Instant Protection.
Lock-up devices LUD and snubbers are
dampers with extremely restrictive oriԴces
that function much like a seatbelt These
products are used to limit relative motion
between masses during a transient event
while allowing free motion in the normal
mode This allows dynamic loads to be
shared at multiple locations rather than
concentrated at Դxed locations

LUDs can also be used to prevent two


adjacent buildings from colliding or to
tie multiple masses together during
an earthTuake

Seattle Central Light Rail


16 Lock-Up Devices with various force and
stroke capacity for seismic protection
Output Force: Up to 60,000 lbs
Stroke Capacity: +/- 6 inches
Engineer: ,nternational Bridge Technologies
Various Contractors: PCL, Mowat

,n the 1990s, Taylor Devices began using snubbers, originally developed during the 1960’s for
protection of large steam pipes in nuclear power plants, as locking devices to limit the relative
motion of highway bridge sections under various types of transient motion - usually seismic in
origin When used on civil engineering structures, the product is usually called a lock-up device
LUD or shock transmission unit STU 

Most applications have been in regions of low or moderate seismic risk The devices allow
essentially unrestricted motion when the bridge structure expands and contracts thermally,
but locks the structural masses together under seismic or wind storm conditions

Our lock-up devices have the same basic dimensions as our Եuid viscous dampers Available
in force ranges of 10 kip to 2000 kip with thermal stroke capability of up to  inches

@Seismicisolation
@Seismicisolation
155
Facilities

Made in the USA

More than All Taylor products are made in the United States at our two manufacturing centers located in
North Tonawanda, New <ork Our facilities cover 1 total acres and over 10,000 sTuare feet

600 buildings and are minutes from both the Buffalo Niagara ,nternational Airport and Niagara Falls
,nternational Airport

and bridges Small Parts Manufacturing


are now using The Small Parts Division includes a small machine shop and tool room with separate
assembly and pre-test room Assembly of space-TualiԴed products is performed in a controlled
Taylor Dampers atmosphere clean room with a laminar Եow assembly bench certiԴed to Class 100 reTuirements
of US Federal Standard 209(

Large Parts Manufacturing


The Large Parts Division consists of a complete machine shop, using predominantly
custom-built machinery for boring, deep hole drilling, gun drilling and turning of large parts
Two assembly ]ones are used - both are deep pit areas capable of assembling and pre-testing
products up to  feet in length

Common Facilities
Both Taylor manufacturing centers are capable of centerless and center grinding, honing,
diamond lapping, bar and plate handling, in-house welding and heat treatments

Quality Assurance
Taylor facilities and business management systems are registered to the following standards
■ ,SO 9001 - audited and certiԴed by NSF ,nternational Strategic Registrations for AS910200
■ AS9100 - audited and certiԴed by NSF ,nternational Strategic Registrations for AS910200
■ ,SO 1001 - environmental management standard audited and certiԴed by NSF ,nternational
Strategic Registrations
■ N+B-300 1C - NASA Tuality standard for spacecraft and satellite applications

■ (N1129 - (uropean standard for anti-seismic device

Testing Facilities
Taylor facilities set the world’s standard for testing of shock absorber and shock isolation
systems Facilities include
■ Vertical drop rails ■ +ydraulic multipliers

■ +eavyweight vertical drop rail ■ Static testing

■ Super-heavyweight vertical drop rail ■ Thermal testing

■ +ydraulic vibration testing ■ Velocity testing

■ +igh freTuency vibration testing ■ Data acTuisition

@Seismicisolation
@Seismicisolation
156
Frequently asked questions

Q :KDW DUH WKH EHQHζWV RI XVLQJ 7D\ORU 'HYLFHV̢ 6HLVPLF 'DPSHUV Q :K\ DUH 7D\ORU )OXLG 'DPSHUV EHWWHU WKDQ RWKHU VROXWLRQV VXFK DV
LQ D VWUXFWXUH" %XFNOLQJ 5HVWUDLQHG %UDFHV %5%̢V "
A Since dampers resist dynamic motion and remove energy from a A Taylor Fluid Dampers are reliable, predictable, reusable and have
structure during wind or seismic events, the resulting structural stress been proven through decades of use Our dampers allow perfect
and deԵection will always be mitigated in an efԴcient and reliable re-centering of the structure due to near-]ero resistance force at
manner This enables the structure to withstand the harsh input energy near-]ero velocities ,n fact, they are instantly ready for the next
associated with these events and allows for a simple retroԴt of existing aftershock Automobile suspensions converged to viscous damping-
structures without costly foundation work type shock absorbers over 100 years ago for the same reason
to be instantly ready for the next bump
Q :K\ GR 7D\ORU 'DPSHUV RIIHU D EHWWHU VROXWLRQ WKDQ RWKHU GHYLFHV"
There are several reasons our dampers are superior to BRBs
A Since damping force in a Taylor Damper is dependent on velocity, this
force is out-of-phase with the normal structural stresses during a ■ Taylor Dampers allow the relief of forces that could
dynamic event such as an earthTuake otherwise build up in BRB’s as a result of creep and
shrinkage as a building ages
When a structure moves, the maximum structural stress occurs at the
point of maximum deԵection At this point the velocity is ]ero, and the ■ BRB’s are yielding members, and their characteristics
damping force is therefore ]ero Then, the maximum damping force and life are unpredictable after use
occurs when the structural stress is lowest, because this is the precise
instant in time that the velocity is the highest This results in the ability ■ BRB’s need to be replaced after a signiԴcant seismic event
of the damper to precisely remove energy from the structure at optimal
points of deԵection, while not needing any additional structural strength ■ BRB’s cannot provide energy dissipation for low-level
to withstand damping forces earthTuakes A BRB functions only as a brace until a
larger earthTuake strikes, when they may or may not yield
The result forces and accelerations are reduced simultaneously with
deԵection reductions, thereby allowing the structure to protect not only ■ BRBs increase acceleration levels in the structure
the structural frame, but contents and occupants as well
Q :KDW W\SH RI ηXLG LV XVHG LQ D GDPSHU"
Q +RZ UHOLDEOH DUH 7D\ORU 'DPSHUV"
A Silicone Եuid ,t is non-toxic and is cosmetically and chemically inert
A Taylor Devices has been manufacturing Եuid dampers since 19 This Եuid is thermally stable, does not experience viscosity break-
We have extensive experience and a proven track record over many down and will not settle out ,ts Եashpoint is greater than 600qF and
decades Our structural dampers have been tested to provide is non-Եammable and non-combustible The Եuid is manufactured and
completely leak-free operation for millions of cycles We use our own certiԴed in accordance with US Federal standards
proprietary seals manufactured from structural plastic combined with
solid high strength stainless steel piston rods polished to a mirror-like Q +RZ PXFK GDPSLQJ LV XVXDOO\ QHHGHG WR SURWHFW VWUXFWXUHV"
Դnish of less than  micro-inches These types of products have been
successfully used in thousands of applications, beneԴting buildings, A Structures have different reTuirements Sometimes the structure calls
bridges, steel mills, chemical plants, military ships, submarines, aircraft for a reduction in deԵection, stress, or acceleration or a combination
and missiles Taylor Devices Dampers do not degrade with age, do thereof  Typical structures can experience a dramatic improvement
not utili]e any moving parts for Եuid Եow oriԴces and are completely with 10 to 0 percent of critical damping added A dynamic analysis
maintenance-free These units do not need service or even inspection demonstrates the best solution through a relatively simple iterative
after a seismic event and maintain their reTuired characteristics for the approach +owever, due to the fact that the optimi]ed solution reTuires
life of the structure non-linear damping performance, it is not accurate to Tuantify the
reTuired damping in terms of percent of critical A dynamic analysis
Q :KDW W\SHV RI PDWHULDOV DUH XVHG IRU GDPSHUV" demonstrates the best solution through a relatively simple iterative
approach
A All materials used are corrosion-protected, usually with a combination
of plating and paint Special paints and colors are used if reTuested Q +RZ FDQ , EH VXUH WKDW 7D\ORU )OXLG 'DPSHUV ZLOO RSHUDWH DV
Stainless steel is used exclusively in the piston rod Stainless steel is UHTXLUHG GXULQJ DQ HDUWKTXDNH RU ZLQG HYHQW"
also available for any external part if needed
A To ensure exact performance, prior to shipment, each and every Taylor
Taylor Devices does not use commercially available seals, but instead Devices structural damper is dynamically tested to maximum rated force
rely on our own proprietary machined seal design using high strength at maximum velocity using our world-class in-plant test machines with
structural polymers rather than soft elastomers This seal does not instrumentation certiԴed to exact US government standards (ach
degrade with age and we have test units that date back to 19 that damper is proof-pressure-tested to ensure the integrity of its parts and
operate perfectly today with ]ero leakage and no reԴlling or seal assembly A test report on each damper
changes of any type needed Our seals have been tested to provide is provided with every shipment
completely leak-free operation for millions of cycles

90 Taylor Drive / PO Box 7


North Tonawanda, N<, USA 1120-07
Phone 71669000 / Fax 71669601
wwwseismicdampercom

@Seismicisolation
@Seismicisolation
157
PROTECT YOUR
INVESTMENT
WITH THE ULTIMATE CHOICE
IN SEISMIC PROTECTION

@Seismicisolation
@Seismicisolation
158
COST EFFECTIVE
Originally developed for NASA and the US Military,
leading structural engineers use Taylor Dampers as
a cost effective way to protect all types of buildings,
their contents and their occupants from damage
caused by earthquakes.
Since the early 90s, modern-day damping technology
has been protecting buildings and saving lives in
high-seismic areas throughout the world. Taylor
Dampers, much like shock absorbers on a car, absorb
energy to protect a structure. Without them, you’re in
for a bumpy ride.

EXCEED INDUSTRY EXPECTATIONS


Current building codes only require that a building
should not collapse during a major earthquake. Heavy
structural damage, is acceptable under the current
building codes. That means that a building could still
sustain a total loss after an earthquake and occupants
FRXOGEHIRUFHGWRԴQGDQRWKHUUHVLGHQFHRUZRUNSODFH
Not with Taylor Dampers.

@Seismicisolation
@Seismicisolation
159
ADDED VALUE
Taylor Dampers provide much better building
performance — even to the point of allowing immediate
occupancy after a major earthquake — for little or no
H[WUDFRVW7D\ORU'DPSHUVUHGXFHVWUHVVGHԵHFWLRQ
and acceleration, protecting the structure and content.

KNOW YOUR OPTIONS


Conventional wisdom would indicate that it is better
to add more steel and concrete to make a building
more earthquake resistant. However, heavier buildings
or buildings with yielding braces actually attract more
earthquake load. Furthermore, these elements result in
permanent offset of the building after an earthquake,
and need to be replaced.

TAYLOR DAMPERS: THE SAFE CHOICE


There are no disadvantages to using Taylor Dampers.
For one, these dampers offer increased safety for
building residents, businesses and equipment. They
also enable immediate occupancy status after even
the largest predicted earthquakes and allow you to
increase rent and decrease insurance premiums.

PROTECT YOUR BUILDING WITH


STATE-OF-THE-ART TAYLOR
DEVICES FLUID VISCOUS DAMPERS
• 7D\ORU'DPSHUVDUHTXDOLԴHGIRUPLOOLRQVRIF\FOHV
and will last the life of the building, regardless of the
number of earthquakes and aftershocks it experiences.
• Each and every damper is tested prior to shipping
from the factory.
• Buildings with Taylor Dampers will passively reset to
their original upright position after an earthquake.
• Taylor Dampers will absorb the energy of an
earthquake so the building doesn’t have to.
• Taylor Dampers come with a 35-year product warranty
IRUDOOEXLOGLQJGDPSHUVRIIHUHGZLWKWKHFRQԴGHQFH
of 60+ years of experience.
• Taylor Dampers are maintenance free.

@Seismicisolation
@Seismicisolation
160
DESIGN OPTIONS FOR OWNERS,
ARCHITECTS AND ENGINEERS
Leading engineers and architects are taking advantage
RIWKHEHQHԴWVRI7D\ORU'DPSHUVLQWKHLUSURMHFWV
You can arrange Taylor Dampers in many ways, whether
you want to display them for all to see or hide them
inside the walls. Here are some examples:
0DQ\RWKHUFRQԴJXUDWLRQVDUHDYDLODEOH$GGLWLRQDOO\
we can supply dampers with any type of paint system,
FRORUDQGDYDULHW\RIDUFKLWHFWXUDOԴQLVKHV

COMMITTED TO THE INDUSTRY


Taylor Devices is proud to be a founding member
of the U.S. Resiliency Council (USRC). The USRC
strives to:
• Establish and implement meaningful rating systems
that describe the performance of buildings during
earthquakes and other natural hazard events.
• Educate the general public to understand
these risks.
• Help building stakeholders better plan for, recover
from and ultimately improve the resilience of their
communities, one building at a time.

TAYLOR DAMPERS —
WHY SETTLE FOR LESS?

90 Taylor Drive, P.O. Box 748


North Tonawanda, NY 14120-0748
P 716 694 0800
F 716 695 6015
taylordevices.com
seismicdamper.com
shockandvibration.com

@Seismicisolation
@Seismicisolation
161
SEISMIC DAMPERS Originally developed for military and
aerospace applications in the 1960’s,
Building Applications Taylor Fluid Viscous Dampers are now
used by structural engineers worldwide
to protect their work and the people who
BY TAYLOR DEVICES rely upon its safety.

• With over 700 projects world-wide, we are the • Our policy is to test every single seismic damper
world leader in providing seismic and wind dampers we produce to maximum output requirements
for any structure
• Taylor dampers are designed and tested in output
• Our dampers have been validated through rigorous force ratings up to 2 million pounds (8900 kN) and
research, full scale testing and published work by amplitudes up to +/-42 inches (1.06 m)
highly reputable organizations
• Highly efficient damper design can absorb
• Proprietary dry-running seals have been tremendous amounts of energy during an
qualified and tested for millions of cycles and are earthquake thus minimizing or eliminating damage
manufactured only by Taylor Devices to the structure

• High strength, mirror polished, stainless steel • Unequalled ability to provide a damper design that
piston rods is accurate, controllable, efficient, temperature and
frequency independent and maintenance free
• Strict control over our design and manufacturing
processes with 3rd party certification to ISO 9001, • Taylor dampers have unequalled ability to respond
ISO 14001 and the stringent aerospace quality to extremely small or very large motions
standard AS9100
• The only technology that can reduce stress and
• Identical quality standards for all our products deflection simultaneously; damper forces are out of
whether they are used for spaceflight, military phase with structural dynamic forces
equipment, buildings or bridges
• 35 year warranty
• Products certified by NASA are now being used to
ISO 9001
save lives on earth AS9100
ISO 14001

www.taylordevices.com | www.seismicdamper.com
90 Taylor Drive | North Tonawanda, NY 14120 -0748 | Phone: (716) 694-0800 | Fax: (716) 695-6015

@Seismicisolation
@Seismicisolation
162
90 Taylor Drive | North Tonawanda, NY 14120 -0748 www.taylordevices.com
Phone: (716) 694-0800 | Fax: (716) 695-6015 www.seismicdamper.com
05/2017

@Seismicisolation
@Seismicisolation
163
SEISMIC DAMPERS Originally developed for military and
aerospace applications in the 1960’s, Taylor
Bridge Applications Fluid Viscous Dampers are now used by
structural engineers worldwide to protect
bridges and the people who rely upon them
BY TAYLOR DEVICES to have a functional structure/lifeline after
an earthquake.

• Taylor Devices is the world leader in providing • Our policy is to test every single bridge damper we
dampers to control vibrations caused by produce to maximum output requirements
earthquakes, wind, traffic and pedestrians
• Taylor dampers are designed and tested in output
• Our dampers have been validated through rigorous force ratings up to 2 million pounds (8900 kN) and
research, full scale testing and published work by amplitudes up to +/-42 inches (1.06 m)
highly reputable organizations
• Unequalled ability to provide a damper design that
• Proprietary dry-running seals have been is accurate, controllable, efficient, temperature and
qualified and tested for millions of cycles and are frequency independent and maintenance free
manufactured only by Taylor Devices
• Taylor dampers have unequalled ability to respond
• High strength, mirror polished, stainless steel to extremely small or very large motions
piston rods and 3 part heavy-duty paint system on
external components • The only technology that can reduce stress and
deflection simultaneously; damper forces are out-
• Damper output force is proportional to velocity, of-phase with structural dynamic forces
Force = CV! ; ! between 0.2 and 2.0; unlimited C
"

values available for optimal structural performance


" • Bridge dampers are available with “Lost Motion
Devices or Fuse Elements or Brake Elements” to
• Long stroke bridge dampers have a heavy-wall steel eliminate the continuous response to small, every-
guide sleeve day traffic and wind induced vibrations

• Identical quality standards for all our products • Dampers can also be provided with special stroke
whether they are used for spaceflight, military limiting devices or end-of-travel bumpers
equipment, buildings or bridges
ISO 9001
AS9100
ISO 14001

www.taylordevices.com | www.seismicdamper.com
90 Taylor Drive | North Tonawanda, NY 14120 -0748 | Phone: (716) 694-0800 | Fax: (716) 695-6015

@Seismicisolation
@Seismicisolation
164
90 Taylor Drive | North Tonawanda, NY 14120 -0748 www.taylordevices.com
Phone: (716) 694-0800 | Fax: (716) 695-6015 www.seismicdamper.com
06/2017

@Seismicisolation
@Seismicisolation
165
METAL BELLOWS DAMPERS
BY TAYLOR DEVICES

The Ultimate in Fluid Damper Design • ,QĆQLWHOLIH – The only damper design in the world
WKDWLVGHVLJQHGIRULQĆQLWHOLIH3DUWVDUHVWUHVVHG
Unique hermetically sealed dampers were below the material fatigue endurance limit ensuring
originally developed by Taylor Devices the dampers will survive for billions of cycles
during the 1980’s for use on platforms based • Virtually zero friction – There is no source of friction
LQRXWHUVSDFHZKHUHćXLGOHDNDJHFRXOG typical of standard elastomeric seals
cause catastrophic failure to critical space
• Effective damping at extremely small amplitudes
missions. Conventional sliding surfaces that
were sealed acceptably on earth proved • ([WUHPHO\KLJKĆGHOLW\UHVSRQVHWRVKRFNDQG
vibration over a frequency band of 0-500 Hz
unacceptable for spacecraft use. The reason and amplitudes as low as +/- 0.001 inches (+/-
was simply that even the tiniest amount of 0.00254mm) or less
ćXLGOHDNDJHSDVWFRQYHQWLRQDOVHDOVWXUQV
into a dense fog in a vacuum, contaminating • High series stiffness for high damping effectiveness

optics and electronic systems. • Various damping functions are achieved based on
customer requirements
Now, these same dampers have been
successfully adapted for use in structures for • Taylor Metal Bellows Dampers are designed and
tested in output force ratings up to 450 KIP or more
protection against wind and seismic events.
DQGVWURNHVXSWRLQFKHVRUPRUH
Additionally, metal bellows dampers are ideal
for applications requiring damping at very • No maintenance over the life of the structure or
dynamic system
low displacements, including human-induced
vibration for bridges and buildings. • $EVROXWH]HUROHDNDJHGXHWRKHUPHWLFDOO\VHDOHG
construction
ISO 9001
AS9100
ISO 14001

www.taylordevices.com
om | www.seismicdamper.com
90 Taylor Drive | North Tonawanda, NY 14120 -0748 | Phone:
ne: (716) 694
694-0800
080 | Fax: (716) 695-6015

@Seismicisolation
@Seismicisolation
166
90 Taylor Drive | North Tonawanda, NY 14120 -0748 www.taylordevices.com
Phone: (716) 694-0800 | Fax: (716) 695-6015 www.seismicdamper.com
12/2017

@Seismicisolation
@Seismicisolation
167
Open Space Damper System
BY TAYLOR DEVICES

This new application of proven fluid damper technology provides all


the benefits associated with adding damping for seismic and wind
protection of structures with minimal blockage of building bays.

• Until now, all damping systems such as diagonal braces, chevron braces, wall
dampers and toggles braces would visually and physically obstruct the bays
where they are installed, thus resulting in occasional rejection of the damping
system by the architect or owner

• The Open Space Damping System utilizes simple mechanical linkages to allow the
system to be located around the perimeter of the bay, thus resulting in virtually
no change in the appearance of the bay

• Simple linkages and mechanisms provide a damping system to remove vibration


energy that is effective as a diagonal brace without consuming an entire bay for
implementation of the system

• Extensive seismic testing on the large scale shake table located at the State
University at Buffalo SEESL Laboratory has validated the performance, theory
and computational models for the system. Technical Report MCEER-16-0007
is available upon request and provides complete testing results as well as the
system theory and computational models to be used for analysis of structures
with this system

• The drawing above shows one configuration that is available. Several variations
of this configuration are also available

• Taylor Devices provides complete theoretical and analytical support for system
implementation. Contact us today to engage our services at no cost

• Patent Number 9,580,924


ISO 9001
AS9100
ISO 14001

www.taylordevices.com | www.seismicdamper.com
90 Taylor Drive | North Tonawanda, NY 14120 -0748 | Phone: (716) 694-0800 | Fax: (716) 695-6015

@Seismicisolation
@Seismicisolation
168
90 Taylor Drive | North Tonawanda, NY 14120 -0748 www.taylordevices.com
Phone: (716) 694-0800 | Fax: (716) 695-6015 www.seismicdamper.com
06/2017

@Seismicisolation
@Seismicisolation
169
Standing Tall
Taylor Solutions for High-rise Structures
Taylor Devices manufacturers damping systems for wind and
seismic control of the most demanding high-rise projects.

taylor devices www.taylordevices.com


@Seismicisolation
@Seismicisolation
170
MEGABRACES

Damped Mega Brace System Benefits:


A novel strategy for dynamic control of high-rise towers is the damped megabrace • Fewer dampers in the
system. The system provides damping within the length of megabraces that span structure.
multiple floors. The damped megabrace works by placing a fluid viscous damper
or a group of dampers at the end of a long bracing element which forms the main
lateral resisting system. Since the viscous dampers provide no resistance to static • Less steel, reducing cost &
loading, a parallel system of bracing is also used. improving performance.

The 181 Fremont Street Tower, located in Downtown San Francisco, is arguably the • Reduced weight is
most resilient tall building on the West Coast of the United States. In the heart of advantageous in areas where
San Francisco, adjacent to the Transbay transit center, it is a prime real estate in the soft soil conditions and
city. Given the relative slenderness and the weight of the structure (steel framed), liquefaction are a concern.
the building is relatively susceptible to wind-induced vibration. Aside from the
damped megabrace, both TMD and TSD were considered. While some strategies • Damping can be integrated as
for reducing accelerations have used TMDs, the location of the building, in a seismic an architectural feature of a
region, favored a robust, distributed damping system. While feasible, TMD and TSD
building.
systems were not pursued, in part because of the high value placed upon floor
space at the top of the building and the additional steel required to support the
weight of the TMD or TSD.

The viscous dampers were placed in eight groups of four within the bracing system.
The damping is approximately 8% in two directions for the service level wind. For
larger wind events and earthquakes, this effectively increases to 10%–20% because
of the beneficial nonlinearity in the dampers. The added damping significantly
reduced the overall steel tonnage for the project.

taylor devices 2 www.taylordevices.com


@Seismicisolation
@Seismicisolation
171
DAMPED OUTRIGGERS

Damped Outrigger Systems Benefits:


A damped outrigger system is highly efficient in reducing dynamic movements from • No temperature sensitive
hazardous winds. This system is based on the concept of total gross movement of viscoelastic materials.
the structure, applying vertical tension/compression forces into perimeter columns.

Outrigger damping can be accomplished by constructing a rigid cantilever off the • No failure prone components
core of the building within a specific floor level, or levels, and connecting fluid like valves.
viscous dampers between the end of the cantilever and the outer columns. The
solution takes advantage of the tension and compression on the opposing outer • No need for re-centering,
columns of the building which amplifies the movements of the central core at the repair or replacement.
location of the outriggers.
• Provides energy dissipation
250 West 55th Street in New York City is a 40-story office tower. This structure to all frequencies of input
required additional damping based on the potential for higher than anticipated vibration.
building accelerations due largely in part to both the shielding and wake buffeting
caused by the surrounding buildings. The engineers at ARUP opted for the damped
• Components are small and can
outrigger system as it offered the same level of damping as a TMD but without
sacrificing valuable real estate at the top of the structure or added weight to the be hidden in various locations
building. The final building design had 10% less steel tonnage than a configuration throughout the structure.
with no damping, effectively saving the client millions of dollars.
• Avoids large potentially
dangerous masses in the
Damped outrigger configuration for concrete structures
structure.

• Can be used to reduce service


and strength level demands.

taylor devices 3 www.taylordevices.com


@Seismicisolation
@Seismicisolation
172
Tuned Mass Dampers

Tuned Mass Dampers Benefits:


Tuned mass dampers (TMDs) are attached to high-rise towers to reduce • They do not depend on
dynamic response caused by hazardous winds. A TMD is a system composed of
external power source for their
a mass, springs and dampers tuned to a specific frequency. When a frequency of
wind loading causes dynamic amplification of a tower, the TMD will resonate out of operation.
phase with the building, energy will be dissipated by the dampers, and the tower’s
dynamic response is improved. • They can respond to small
levels of excitation.
432 Park Avenue is an 88-story, 1396ft (425m) tall residential skyscraper with an
aspect ratio of 14:1 that overlooks Central Park in New York City. When it was • Their properties can be
completed in 2015 it was the tallest residential building in the world, and the third adjusted in the field.
tallest building in the United States.
• They can also be introduced in
A traditional pendulum type TMD would have taken up roughly eight stories of structural upgrades or retrofits.
valuable space at the top of the structure, so another solution had to be found.
Engineers at RWDI proposed an opposed-pendulum design that would reduce the
required vertical space to three floors by splitting the mass into two equal parts and • They require low maintenance.
placing them on either side of the central tower core.

Taylor Devices supplied a total of 16 fluid viscous dampers as a part of this project
aimed to improve occupant comfort. The outcome was a design that reduced the
required space while maintaining travel of 136 inches peak to peak.

taylor devices 4 www.taylordevices.com


@Seismicisolation
@Seismicisolation
173
Direct Acting Damping
Direct Acting Damping
While many methods exist to implement distributed damping in a structure, the underlying concept is to connect the
dampers where motion will occur, such as between beam and column joints or between floor levels which deform relative
to one another in a shearing-type motion. Some common configurations are listed below.

Open Space
While not specifically used for high-rise buildings,
Taylor Devices does offer an open space configuration
designed to offer more open bays.

Toggle
Toggle frames can be used as a mechanism to amplify
deflections into the damper in otherwise stiff, or tiny
deflection situations, creating a more efficient damping
system. Toggle Frames utilize a bent-brace mechanism
theory to capture deflections in one plane and translate
the deflections into another plane and therefore provide
very efficient damping,

Chevron
In this configuration, the dampers are placed horizontally,
and connected to a frame (chevron) that is intended to be
near-rigid with the floor it is connected to. The advantage
with this direct damping orientation is that the horizontal
flexibility of the structure injects this full movement directly
into the horizontal orientation of the damper. However, a
small amount of motion can be lost due to the constraints
of the attainable stiffness of an economical chevron frame.

Diagonal
A very common method of applying distributed damping
to a structure is to connect the dampers to diagonal
corners or center of a structural frame or bay. In this
orientation, the horizontal movement of the structure
allows an angular component of the full deflection to go
into the damper. This takes the motion directly to the next
floor level through a strong tension/compression member.

taylor devices 5 www.taylordevices.com


@Seismicisolation
@Seismicisolation
174
Project References
Chicony HQ - Taiwan Millennium Place - Boston

A 40-story commercial high-rise in New A 37-Story luxury residential high-rise in


Taipei City. Boston, Massachusetts.

Solution: Direct (Chevron Brace) Solution: Direct (Toggle Brace)

Completion in 2015 Completion in 2000

111 Huntington Ave - Boston Torre Mayor - Mexico City

A 36-story commercial high-rise in Bos- A 55-story mixed-use high-rise located


ton, Massachusetts. in Mexico City, Mexico

Solution: Direct (Toggle & Diagonal Braces) Solution: Damped Mega Brace

Completion in 2002 Completion in 2003

28 State Street - Boston Farglory the ONE - Taiwan

A 40-story commercial high-rise in A 68-story mixed-use high-rise located


Boston, Massachusetts in Kaohsiung, Taiwan.

Solution: Direct (Diagonal Brace) Solution: Direct (Chevron Brace)

Completion in 1970 Completion in 2019


Retrofitted in 1996

217 West 57th Street - New York City Park Tower - Chicago

A 100-story mixed-use high-rise located A 70-story mixed-use high-rise located


in New York, New York. in Chicago, Illinois.

Solution: Tuned Mass Damper with Solution: Tuned Mass Damper with
Taylor VDDs Taylor Dampers.

Completion in 2020 Completion in 2000

About Quality Standards


Since 1955, Taylor Devices, Inc has been a world leader in Taylor Devices holds itself to the strictest national and international standards. Our
the shock and vibration control industry. Over 700 buildings, facilities and Business Management Systems are registered to the current versions of
bridges, stadiums and other structures around the world rely ISO 9001, ISO 4001 and AS 9100. The same level of quality is standard on every product
on the quality and durability of our products. we manufacture, regardless of industry or application, with all structural products being
fully tested prior to shipment.

Taylor Devices, Inc. Headquarters Find Us Online


90 Taylor Drive Visit us at www.taylordevices.com for more information
North Tonowanda, NY 14125 on all our products and services. Be sure to subscribe to
our newsletter or find us on youtube and social media for
Contact Us all the latest news and information.
Tel: +1 (716) 694-0800
Fax: +1 (716) 695-6015
[email protected]
taylor devices www.taylordevices.com
@Seismicisolation
@Seismicisolation
175
@Seismicisolation
@Seismicisolation
176
15 Sample Technical Manual

Building & Bridge Projects


1.0 Technical Description
Taylor Devices’ Fluid Viscous Dampers are constructed only from the highest quality materials. Finish
machining and fabrication of parts is performed exclusively at Taylor Devices’ North Tonawanda,
NY facility. Shown at the bottom of this page is a typical Fluid Viscous Damper with the following
components:
1. Piston Rod: Solid 17-4 PH stainless steel, billet machined, through hardened, then
hand polished to a mirror finish of less than 4 microinches.
2. Piston Head: Solid steel construction, machined from billet. Contains fluid flow
channels that provide the orificing for the damping function (F=CVα ).
3. Seals/Seal Bearings: Dynamic seals and seal bearings are manufactured by Taylor
Devices to patented and proprietary specifications from acetyl resin and Virgin
Teflon.
4. Fluid: Silicone fluid, per Federal Standard VV-D-1078. This fluid is nonflammable and
noncombustible under current North American/OSHA standards and is classified as
cosmetically inert by the U.S. Food and Drug Administration.
5. Cylinder: Heat-treated alloy steel, machined from pierced billet or solid, corrosion
protected by painting.
6. End Cap: Heat-treated alloy steel, billet machined from wrought condition, through
heat treated, and corrosion protected by painting.
7. Extender: Carbon steel, machined from wrought billet, tube, or pipe, then painted for
corrosion protection.

Figure 15.1
Typical fluid viscous damper

@Seismicisolation
@Seismicisolation
177
8. End Clevis: Heat-treated alloy steel; painted for corrosion protection.
9. Spherical Bearing: Forged from aircraft quality alloy steel, hand fitted and checked
for clearance.
10. Outer Sleeve: Carbon steel, painted for corrosion protection.

2.0 Handling
Although Taylor Devices dampers are very robust in design, care should be taken when handling a
damper. This section gives the recommended handling procedures.

2.1 Shipping
Each damper shall be placed on a wooden skid for shipping and secured with metal banding and blue
shrink-wrap. The skid must remain upright (The 4x4 wooden feet down) at all times to help prevent
damage to the dampers. DO NOT STACK THE SKIDS.

2.2 Storage
The dampers may be stored on their shipping skids for an indefinite period if they are kept in a
suitable environment. The wooden shipping skids will deteriorate over time unless they are kept in a
relatively dry environment.

2.3 Lifting
The dampers will be shipped on wooden skids. See section 7.0 of this manual for the weight of the
dampers. Each skid must be lifted in two locations with an evenly distributed force as typically
performed with a fork truck.

After carefully removing the shrink-wrap and metal banding, the damper may be lifted from its skid
by a single strap located at its balancing point. Care must always be taken when moving a damper so
that there is no damage to the factory fill port that protrudes out of the cylinder.

2.4 Unit Care


All surfaces on the damper are painted for corrosion protection (See section 5.2 of this manual). Any
damage to the painted or coated surface may cause superficial surface corrosion, and must be touched
up with paint. Only damage to the underlying steel will be cause for evaluation or corrective action.
The factory fill port that protrudes out of the cylinder is a relatively delicate component, compared
to the rest of the damper. Cognizance of this port at all times is necessary to prevent damage to it.
This port seals a high internal fluid pressure and is wired in place for safety. DO NOT REMOVE THIS
PORT OR SAFETY WIRE UNDER ANY CIRCUMSTANCES.

3.0 Installation
Taylor Devices' Fluid Viscous Dampers are easily installed with a minimum of tools required. For
a Viscous Damper used in a structure, connection points of the damper are as indicated in Section
1.0 and shown in Section 7.0. Refer to Section 7.0 of this manual for drawings of the Fluid Viscous
Dampers and their mounting hardware.

3.1 Tools Required


• Tape measure

@Seismicisolation
@Seismicisolation
178
• Hammer (2 to 4 lb.)
• Light marine grease (Taylor Blu-Grease supplied with dampers)
• Hoist device with straps

3.2 Installation Procedure - Fluid Viscous Damper.


1. Measure the eye-to-eye (center of first spherical bearing to the center of the second
spherical bearing) length of the Viscous Damper and compare it with the mid-stroke
length of the Viscous Damper shown in Section 7.0 of this manual. If the dimensions
do not match to within 3 mm, consult a representative from Taylor Devices, Inc. The
Viscous Damper may need to be adjusted. Refer to Section 3.3 & 3.4 of this manual
for more details.
2. Place the mounting brackets (not provided by Taylor Devices) into position using
a jack or hoist as necessary, per the contract documents and drawings. Attach the
mounting brackets with bolts or weld per the requirements in the contract documents
and drawings making sure the brackets are properly aligned towards each other.
3. Hoist the Fluid Viscous Damper into position, using a jack or hoist as necessary. The
clevis ends should easily fit into the mounting bracket assemblies with extra play.
4. Clean the mounting pins with commercially available parts cleaner and let them dry
completely. Lubricate the mounting pins with supplied Taylor Blu-Grease.
5. Align the spherical bearing at one end of the Viscous Damper with the hole in the
mounting bracket using a tapered pin, wood wedges or other means as necessary.
6. Position the mounting pin by hand with the beveled edge leading into the mounting
bracket hole and push through the entire assembly as shown in section 7.0 of this
manual. Re-align and adjust the Viscous Damper as necessary. If difficulty is
encountered in pushing the mounting pin through, light tapping with a hammer
may be used. If unreasonable difficulty is encountered, consult a Taylor Devices'
representative.
7. Install the Retaining Rings on both ends of the mounting pin, as shown in the manual
assembly method diagram found in section 7.0 of this manual.
8. Repeat steps 2 through 7 for the opposite end of the Fluid Viscous damper. If
alignment difficulty occurs, contact a Taylor Devices’ representative.

3.3 Damper length adjustment using the available adjustment for Taylor Devices Part
Number 67DP-17170-01
The dampers will be shipped in a mid-stroke position that is shown on the drawings in section 7.0 of
this manual. This is the length that should be set between the centers of the holes in the clevis plates
where the damper mounting pins will be attached through. The damper units are equipped with an
adjustment nut that allows for mechanical adjustment of ±½ in. The clevis is free to rotate for ±½
in of adjustment (each 180o rotation is equal to 0.083 in). After the damper has been adjusted to
the correct length tighten the adjustment set screw to lock the threads. DO NOT exceed ±½ inch
adjustment in either direction. If further length adjustment is required, the damper must be stroked.
The Engineer of Record must approve any use of damper stroke for length adjustment.

@Seismicisolation
@Seismicisolation
179
3.4 Using Damper Stroke for Adjustment Consult the Engineer of Record

THIS METHOD OF ADJUSTMENT IS ONLY TO BE PERFORMED IF DIRECTED


TO DO SO BY THE E.O.R.
Extension - Stroking a damper in extension can most easily be accomplished by lifting the FVD with
a strap until the FVD is vertical. This may extend the unit. If the unit does not extend under its own
weight, additional weight may be hung from the other end as necessary. Alternatively, a come-along,
or a winching device may be used with one end of the damper pinned into its bracket, and the other
end pulled by the device.

Compression - To compress the FVD is slightly more difficult. The FVD can be lifted, as previously
described, by one end and then set down on the ground on the opposite end. Additional weight may
need to be placed on the top clevis as necessary. Care must be taken to avoid damage to the unit, or
toppling/collapse of this setup.

Note that a force of approximately 2%-4% of the maximum rated force of the damper may be
necessary to overcome internal seal gripping forces and allow stroking of the damper.

4.0 INSPECTION

4.1 Periodic Inspection


Taylor Devices' Fluid Viscous Dampers are designed to be completely maintenance free for the
life of the damper. No periodic maintenance, inspection or spare parts are required, desired, or
recommended for Taylor Devices Fluid Viscous Dampers. However, if the owner’s in-house safety
manual for seismic protection elements requires periodic inspection, the following procedure is
suggested:
1. If the dampers are not located in open access bays, provide access panels at damper
locations for ease of visual inspection. Access panels should be so located as to
provide clear visual inspection of at least one-half of the whole damper and if possible,
the entire damper.
2. Inspection intervals of approximately two (2) years, or after any sizable earthquake,
with a written record of damper condition (i.e. clean, dusty, etc.).
3. Visual inspection of the damper bay (surroundings) to determine if any foreign object
is interfering with or restricting the damper from functioning properly and that the
bracket alignment is properly accommodated.
4. Visual inspection of the damper units. If the damper has been overloaded in stroke or
force, one of the following observations will be easily noted:
a. Obvious substantial and persistent fluid leakage, observable on the outside of
the damper.
b. Discernable looseness or play in the end attachment bearings of the damper.
3. No other inspection techniques are necessary.

4.2 Other Inspection


Visual inspection of the dampers prior to installation is recommended to determine if any obvious
damage has occurred from shipping, storage or handling. Post-earthquake inspection is recommended
and consists of the procedure listed above in Section 4.1 of this manual. If any discernable or
unordinary situation is observed at any time, please consult the factory:

@Seismicisolation
@Seismicisolation
180
Taylor Devices, Inc.
90 Taylor Drive
North Tonawanda, NY 14120-0748

PHONE: 716-694-0800
FACSIMILE: 716-695-6015

5.0 MAINTENANCE

5.1 Periodic Maintenance


Taylor Devices Fluid Viscous Dampers are intended for 100% maintenance free operation and hence,
no fluid refill ports are provided. Under normal operation, no maintenance whatsoever is required
for the life of the damper.

All Taylor Devices Fluid Viscous Dampers have the main cylinder cartridge sealed at the factory and
no field repair or maintenance can be performed. If the damper sustains any damage, the damper
must be returned to the address below:

Taylor Devices, Inc.


90 Taylor Drive
North Tonawanda, NY 14120-0748

Taylor Devices' personnel will then contact the owner of the damper concerning the status of
the device.

5.2 Painting
For aesthetic value, the dampers can be repainted if a different color is desired or if any surface
irregularities develop on painted surfaces. The spherical bearings and rod should not be painted.
External surfaces that can be repainted have been factory painted with a corrosion resistant primer,
top color black. All other surfaces are to be carefully covered and masked to prevent paint spray from
contacting these surfaces and from entering the sleeve or spherical bearing assemblies. Always be
sure to remove all masking tape after paint has dried. Failure to remove masking, or painting surfaces
not recommended for painting; could cause improper function of the damper. Feel free to consult the
factory, should there be any additional questions or concerns. Paint information is included in section
7.0 of this manual for reference.

@Seismicisolation
@Seismicisolation
181
WARRANTY
Taylor Devices Fluid Viscous Dampers
SELLER warrants the fluid viscous dampers against defects in materials and workmanship
under normal use and service for a period of thirty-five (35) years from the date of
installation. Surface coatings are excluded from this warranty.

SELLER warrants that the damping coefficient of the dampers will be within ±15% of
the design value at 70°F within the warranty period. Any obligation under this warranty
terminates if the fluid dampers are directly exposed to fire, flood, or if an earthquake
occurs which exceeds the specified Design Earthquake level. In such instance, BUYER may
reactivate the warranty by hiring SELLER to inspect all fluid dampers and for BUYER to
pay SELLER to recondition any fluid damper that is deemed required.

This warranty shall not apply to any of the SELLER's products which must be replaced
because of normal wear, which have been subject to misuse, negligence or accident or
which shall have been repaired or altered outside of the SELLER's factory unless authorized
in writing by the SELLER.

The benefits to BUYER of this warranty are expressly conditioned upon BUYER's care,
inspection, and maintenance of the fluid damper in strict accordance with SELLER's
inspection and maintenance requirements.

DISCLAIMER OF WARRANTIES
THIS WARRANTY IS IN LIEU OF ALL OTHER WARRANTIES, EXPRESSED,
IMPLIED, WRITTEN, OR ORAL, INCLUDING BUT NOT LIMITED TO ANY IMPLIED
WARRANTY OR MERCHANTABILITY OR FITNESS AND IS PROVIDED IN LIEU OF
ALL OBLIGATIONS AND LIABILITIES OF SELLER WITH RESPECT TO DEFECTS IN
MATERIALS OR WORKMANSHIP. THE RIGHTS AND REMEDIES CONTAINED IN THIS
WARRANTY EXCLUDES AND WAIVES ANY RIGHT OF BUYER TO CONSEQUENTIAL
OR INCIDENTAL DAMAGES.

7.0 SPECIFICATION DRAWINGS


Outline drawings of the damper and hardware supplied by Taylor Devices are following this page.

@Seismicisolation
@Seismicisolation
182
Figure 15.2
Lock Up Device Drawing

Figure 15.3
Pin Kit Drawing

@Seismicisolation
@Seismicisolation
183
Figure 15.4
Pin Kit Assembly

@Seismicisolation
@Seismicisolation
184
Figure 15.5
Retaining Ring Installation

@Seismicisolation
@Seismicisolation
185
Figure 15.6
Typical Paint Technical Data Sheet

@Seismicisolation
@Seismicisolation
186
Appendix A

Case
Studies

@Seismicisolation
@Seismicisolation
187
CASE STUDY: 1

2017 SEAOC CONVENTION PROCEEDINGS

Case Study: Cost Delta for Achieving Higher Structural


Performance Levels

Kamalpreet Kalsi, Project Engineer


Daniel Zepeda, Principal
Degenkolb Engineers
Los Angeles, CA

Abstract improved seismic performance of acute care facilities after a


major seismic events. These Performance Objectives are
Major cities in California are now mandating that existing categorized into five categories, SPC-1, SPC-2, SPC-3, SPC-4
buildings with the highest seismic risks be retrofitted in and SPC5 (See Figure 1c). SPC-1 is not considered safe under
conformance with local ordinances. As a consequence to the a seismic event, while SPC-2 and above will have different
recent engineering community’s push for cities and building levels of safety under a seismic event. Figure 1a and Figure 1b
owners to become more resilient against seismic events, show the timeline of major milestones associated with
many owners are now more conscious about their buildings’ California legislation for hospitals regulated by OSHPD.
anticipated seismic performance. Owners are now beginning
to ask engineers what it means to design above the minimum
code standards. For an owner to make an educated decision on
building design, engineers need to convey the increased cost
of a higher structural performance in simple terms. This paper
covers a scenario in which an existing Pre-Northridge Steel
Moment Frame building was evaluated for three different
performance objectives under California’s hospital building
standards. This paper highlights the differences in structural
scope between each performance level as well as the expected
percent increase in construction costs. Engineers can use this
case study as an example when speaking to their clients about
relative costs between different seismic performance levels.

Introduction and Background Figure 1a: Timeline of Hospital Upgrades in California


Note: Figure re-printed from webinar by California Hospital
The case study project is a hospital located in California and Commission on May 2015.
falls under Office of Statewide Health Planning and
Development (OSHPD) jurisdiction.

Under the state code, any essential facility (in this case an acute
care hospital) requires the importance factor “I” to be 1.5 when
designing a new building. Similarly, when evaluating or
retrofitting an existing building, a higher performance criteria
is utilized. The higher seismic criteria result in more extensive
retrofits compared to similar buildings under the same hazard.
California has always been in the forefront of seismic
protection by demanding stricter regulations to protect lives
and property. Essential facilities, such as hospitals, came under
regulation after passage of the Alfred E. Alquist Hospital
Seismic Safety Act post 1971’s Sylmar earthquake. Senate Bill
(SB) 1953 and subsequent regulations have provided timelines Figure 1b: Timeline of Hospital Upgrades in California
and performance objectives that hospitals should meet to contd.
comply with California State’s overall goal of ensuring the

@Seismicisolation
@Seismicisolation
188
2017 SEAOC CONVENTION PROCEEDINGS

Note: Figure re-printed from webinar by California Hospital category provides an alternate path to compliance beyond
Commission on May 2015 2030.

Table 2.5.3 of the California Administrative Code (CAC) American Society of Civil Engineers (ASCE) 41-13 defines
defines the Structural Performance Categories (SPC) in more the different performance levels and CBC 2016 outlines the
detail (Figure 1c). It is worth noting that an existing building objectives for each performance level. As part of this case
could only be upgraded from an SPC-1 to an SPC-2 or SPC-5. study these objectives and criteria were performed for three
However, in recent years OSHPD has defined an intermediate different performance categories: SPC-2, SPC-4D, and SPC-5.
category SPC-4D that can be used to retrofit an existing
building. This new category is expected to yield a similar • SPC-2: Life Safety structural performance level in
seismic performance level as SPC-4. accordance with § 2.3.1.3 of ASCE 41-13 at BSE-1E
hazard. BSE-1E hazard corresponds to a return period
The retrofit example is a six story steel moment frame of 225 year.
building, which was erected in 1972. OSHPD designated the
building as an SPC-1 due to significant structural deficiencies. • SPC-4D: Damage Control structural performance
The project’s general contractor provided pricing to upgrade level in accordance with § 2.3.1.2.1 of ASCE 41-13
the building to structural performance categories: SPC-2, SPC- at BSE-1E hazard and Collapse Prevention structural
4D and SPC-5. See Table 1 for reference. What follows is an performance level in accordance with § 2.3.1.5 of
analysis of project’s potential seismic upgrades and their ASCE 41-13 at BSE-2E hazard. BSE-2E hazard
associated costs. This type of analysis is recommended for corresponds to a return period of 975 years
assisting building owners, as they make critical safety and
financial decision to make seismic upgrades. • SPC-5: Immediate Occupancy structural
performance level in accordance with § 2.3.1.1 of
Table 1: Snapshot from CAC Table 2.5.3 for SPC ASCE 41-13 at BSE-1N hazard and Life Safety
description performance level in accordance with ASCE 41-13 §
2.3.1.3 at BSE-2N hazard. BSE -1N corresponds to a
return period of 475 years, and BSE-2N corresponds
to a return period of 2,475 years.

ASCE 41-13 defines the post-earthquake damage state for Life


Safety as the state in which a structure has damaged
components but still regains a margin against the onset of
partial or total collapse. In addition, the definition of damage
state for a Damage Control performance objective is the
midway point between Life Safety and Immediate Occupancy.
It is intended to provide a structure with a greater reliability of
resisting collapse and being less damaged than a typical
structure but not to the extent required of a structure designed
to meet the Immediate Occupancy Performance Level. It is
important to note that, for this performance level, repairs may
Note: Reprinted from California Administrative Code 2016. be required before the building can be re-occupied after a
design level earthquake.
Performance Levels and Hazards
The post-earthquake damage state for Collapse Prevention is
As indicated in the table above, in 2016 California Hospital such that the building is on the verge of collapse with
Building Safety Board included a new performance category significant portions of the structural and non-structural
(SPC-4D) under California Building Code (CBC) to allow components damaged beyond repair. A building designed to
acute care hospitals another seismic upgrade option. The SPC- this level of performance will likely not be able to be
4D performance category is meant to be equivalent to the reoccupied after a maximum credible earthquake. Figure 2,
minimum prescriptive requirements of the 1980 CBC. Prior to shows different damage states for buildings under a seismic
the addition of this new performance category, an existing event.
building (SPC-1 or SPC-2) was required to upgrade to SPC-5.
Buildings not upgraded to these stringent SPC-5 requirements
would have to be removed or rebuilt before 2030. The SPC-4D

@Seismicisolation
@Seismicisolation
189
2017 SEAOC CONVENTION PROCEEDINGS

• Soft Story per § 3.3.2


• Adjacent Building per § 3.4
• Drift Check per § 4.3.2
• Strong Column Weak Beam per § 4.2.8
• Pre-Northridge Welded Moment Frames per § 4.2.10
• Untopped Diaphragms per § 7.3.1 and 7.3.2

Building Retrofit Scheme

The building retrofit consists of new primary external damped


3-bay moment frames at all three sides of the building. The
external frames are connected to the structure through
horizontal trusses. The existing perimeter beams are used as
drags and minor strengthening was required to increase the
Figure 2: Different damage stages for buildings drag capacities. Deep pile foundations support the new lateral
Note: Figure re-printed from webinar by California Hospital elements. Each new external frame is supported on a pile cap.
Commission on May 2015 The new piles consist of a battered, proprietary pile system to
resist seismic demands. Figure 3a shows a typicalplan view of
Existing Building Description the building and Figure 3b shows a typical retrofit elevation of
the external damped moment frame with foundation.
The project building is an existing six-story structure built in
1972 that is triangular in plan. It is located in San Bernardino,
California. The lateral force resisting system (LFRS) of the
existing structure consists of pre-Northridge steel moment
frames that extend to the roof level of the structure. The
moment frames are positioned such that one 6-bay moment
frame is located at each of the three sides of the structure along
the perimeter, and one 4-bay moment frame is located 20 ft.
directly inward and parallel with each 6-bay moment frame.
This structural configuration provides 6 frames total (three 6-
bay frames, and three 4-bay frames) within the building. The
moment frames were designed using the prescriptive pre-
Northridge welded unreinforced flange (WUF) moment
connection. The gravity system of the structure consists of a
flat one-way concrete slab that spans between wide flange steel
joists. These joists span between larger wide flange beams &
girders and are in turn supported by the wide flange columns. Figure 3a: Overall Plan of the Building
Above the roof level, there is an existing penthouse and
elevator machine room. The penthouse and elevator machine
room areas consist of un-topped metal deck over wide flange
steel beams supported by steel columns. The penthouse floor
plan is mostly triangular in plan as well. The lateral system for
the penthouse structure consists of diagonal steel plate-straps
in selected beam-column frames. At the foundation level, the
structure consists of a slab-on-grade system with cast-in-place
drilled piles and grade beams.

Existing Building Deficiencies

As part of the initial assessment, a Tier 1 evaluation was


conducted in accordance with Chapter 6 of the CAC. The
following deficiencies were identified by performing a
analysis:

@Seismicisolation
@Seismicisolation
190
2017 SEAOC CONVENTION PROCEEDINGS

returning Architectural-Mechanical-Electrical-Plumbing
(AMEP) fixtures back after necessary removals (if any) to
conduct the retrofit. Finally, ADA projects are upgrades that
are triggered by code for seismic upgrade projects.

Figure 4 shows the direct cost associated with the structural


upgrade of the building for three different structural
performance categories: SPC-2, SPC-4D, and SPC-5. As
illustrated, the incremental structural cost difference between
SPC-4D and SPC-5 is less than that of SPC-2 to SPC-4D. In
this case study, the low relative structural cost difference
between SPC-4D and SPC-5 is atypical, and related to the
selected retrofit configuration, specifically the use of external
dampers. Additionally, when selecting an upgrade option
based on costs, it is important to note that SPC-2 would still
require future upgrades before 2030 per current legislations
(implying additional costs in the near future).

STRUCTURAL COST
$30 $234 /sf
$222 /sf

$25 $192 /sf


Millions

Figure 3b: Elevation of External Damped Moment Frame $20


- Retrofit Scheme
$15
Cost Comparison
$10
Although the evaluation of multiple performance objectives
$5
was not the initial intention of the project, changes in state
legislation (i.e. introduction of SPC-4D) opened the $-
opportunity to explore them. In addition the owner was SPC-2 SPC-4D SPC-5
attracted to the idea of having a better structural performance
with an extended life cycle. As such the design team was
requested by the owner to explore various performance Figure 4: Direct structural cost of retrofit for three
categories. The various analysis provided the team with an different structural performance categories.
opportunity to understand the overall construction costs
associated with different upgrades. Although in a structural Figure 5 shows the cost associated with removing and re-
retrofit project, the structural costs are a major portion of the patching architectural, mechanical, electric and plumbing. For
overall budget, they are not the only deciding factor of the this project, this cost also includes the mandatory ADA
project scope. This is because there are other non-structural upgrade requirements per California Building Code totaling to
costs associated with a retrofit that require a large percentage 20% of the overall project. The total project cost differences
of the overall budget and can also be the determining factors between incremental upgrades from SPC-2 to SPC-4D to SPC-
when deciding how to proceed with a project. Major costs 5 are nominal as compared to the structural cost differences
associated with the structural retrofit explored in this paper (see Figure 4).
include structural upgrades, non-structural upgrades, make
ready projects, ADA projects, and patch back work costs. The
case study does not include costs the owner incurs for special
inspections, and operational downtime. This paper refers to
make ready projects as site preparation and relocation of staff
in and around the construction area to make space “ready” for
the retrofit. Patch back work refers to the costs associated with

@Seismicisolation
@Seismicisolation
191
2017 SEAOC CONVENTION PROCEEDINGS

cost jump between SPC-2 and SPC-4D upgrades (over $3.5


million), the total cost difference between SPC-2 and SPC-4D
AMEP COST is not much greater (just $4 million). Although the AMEP and
make ready costs are substantial, these costs do not increase
$147 /sf significantly between each upgrade scenario the way the
$18.0 structural costs increase between SPC-2 and SPC-4D. This
$17.8 holds true for the overall cost differences between SPC-4D and
$17.6 SPC-5. In addition, the total cost was approximately half of
$142 /sf what a new building would cost in today’s market.
$17.4
Millions

$17.2
$17.0
$16.8
$137 /sf TOTAL COST
$16.6 $800 /sf
$16.4 $100
$90
$16.2
$80
$16.0
$70
Millions

SPC-2 SPC-4D SPC-5


$438 /sf $456 /sf
$60 $402 /sf
$50
Figure 5: AMEP (patch back) cost of retrofit for three $40
different structural performance categories. $30
$20
Figure 6 shows make ready project costs. Similar to the AMEP $10
cost differences between upgrade scenarios, the make ready $-
costs differences between SPC-2 to SPC-4D to SPC-5
upgrades are minimal compared to the structural cost
comparisons (see Figure 4).
Figure 7: Overall (total) project cost of retrofit for three
different structural performance categories and new
MAKE READY COST construction comparison.

$9.2 $75 /sf Conclusion


$9.1 As seen in the cost comparison study of different structural
$9.1 upgrade performance levels, the total upgrade cost consists of
$74 /sf
Millions

many different aspects apart from just upgrading the structural


$9.0
components. A 15% increase in structural only upgrading cost
$9.0 from SPC-2 to SPC-4D results in only 8% increase in total cost
$73 /sf
$8.9 of retrofit. A 5% increase in structural only upgrading cost
from SPC-4D to SPC-5 results in only 4% increase in total cost
$8.9
of retrofit. A 22% increase in structural only upgrading cost
$8.8 from SPC-2 to SPC-5 results in only 13% increase in overall
$8.8 cost of retrofit.
SPC-2 SPC-4D SPC-5
As illustrated in Figure 5 and Figure 6, the AMEP and make
ready costs are substantial to the project; however, the cost
Figure 6: Make Ready projects cost of retrofit for three differences between the three upgrade scenarios are not large.
different structural performance categories. Therefore, when looking at the overall costs differences
between upgrade scenarios in Figure 7, the AMEP and make
Figure 7 compares the overall (total) cost for the three different ready costs dilute the larger structural cost differences between
structural upgrades and an approximate cost of a similarly each upgrade scenario. Having the cost data available for
sized new building. Whereas Figure 4 shows a large structural different performance categories, the owner ended up choosing

@Seismicisolation
@Seismicisolation
192
2017 SEAOC CONVENTION PROCEEDINGS

SPC-4D performance level for upgrade. Even though cost


delta was small to get SPC-5 performance level, the existing
adjacent buildings limited the size of new foundation. SPC-5
foundation size would have resulted it more disruption and
downtime.

We hope this study can be used to encourage building owners


to do a cost study on various performance levels prior to
proceeding with the retrofit design. A reasonable cost increase
can result in having a higher performance building.

References

2016 CBC, California Building Code, California Code of


Regulations, Title 24, Part 2, Volume 1 & 2 (based on the
2015 International Building Code), including Supplements,
California Building Standards Commission, Sacramento, CA.

2016 CAC, California Administrative Code, California Code


of Regulations, Title 24, Part 1

2015, California Hospital Association (CHA) webinar on new


seismic performance category (SPC-4D)

@Seismicisolation
@Seismicisolation
193
CASE STUDY: 2

Seismic retrofit and FEMA P-58 risk assessment of mid-rise


soft-story concrete towers
Kit Miyamoto, Ph.D., S.E., Amir SJ Gilani, Ph.D., S.E. and Lon Determan, S.E.
Miyamoto International
Los Angeles, CA

Abstract The lateral load is resisted by a system of reinforced concrete


perimeter moment frames. Figure 1 presents the plan view of
The two frame office towers, constructed in the 1970s per the a typical floor and the location of the lateral load resisting
1967 edition of the UBC, use perimeter reinforced concrete system. Columns are 28-in square and support 18x40 in. beams
moment frames to resist seismic loading. The buildings are on the second floor and 18x30 in. beams at the floors above. In
rectangular in plan and have certain characteristics that the longitudinal direction, there are five 29-ft long bays on
adversely affect their seismic performance, in particular the each side (10 bays total), and in the transverse direction, there
presence of a soft-story response at the first floor are three 26-ft long bays on each side (6 bays total).
(approximately 50% taller than typical floors), and limited
ductility typical of buildings of that era. Risk analysis showed
that for the towers the PML exceeded 20%. Nonlinear response
history analysis (NLRHA) of the towers was conducted and
showed that in the existing configuration, the story drift ratios
(SDRs) at the first floor exceeded 2%, shear hinging of the first
floor beams was expected and that the SDRs would need to be
reduced to approximately 1.4% for the first floor to limit the
extent of nonlinear response. Seismic retrofit included addition
of 300-kip viscous dampers in both directions to the first floor
of the building. Analysis showed that the retrofitted structure
had a first floor SDR of approximately 1.3% and that the soft
story response and plastic shear hinging of first floor beams Figure 1. Lateral load resisting system
were mitigated. FEMA P-58 analysis of the retrofitted
buildings were then conducted using the results—SDR, story Site seismicity
acceleration, and residual drifts—from the NLRHA. It was The building site is classified as Class D. For the building site
seen that the 90th percentile repair cost (PML) was and soil conditions, the USGS provides the following design
significantly reduced and was now less than 15%. parameters: short period acceleration of 0.53g, and 1-sec
acceleration of 0.31g. These values are used to construct the
Introduction seismic demand for the building. The seismicity at the site is
considered moderate.
Description of the buildings
Recent visits to the building has shown that the structure has
The office towers investigated, located in downtown not experienced any noticeable damage from past earthquakes,
Sacramento, CA, are of very similar construction. For the including the recent 2014 South Napa event.
purposes of this study, the west tower was specifically
modeled and evaluated. The west tower is a 14-story Building characteristics
reinforced concrete moment frame building constructed in
1971. The building has a plan dimension of approximately 150 Overview
x 81 ft. The building is rectangular in plan. Typical floors Reinforced concrete buildings constructed prior to adoption of
measure 11 ft in height, whereas, the first floor is 17 ft tall. new seismic codes are usually classified as having low
Gravity loading is resisted by a system consisting of 8-inch ductility. In other words, such buildings do not have the means
thick post-tensioned 2-way flat slab and reinforced concrete to resist the seismic energy that is imparted to them by
columns bearing on a pile foundation system. earthquakes. The low-ductile reinforced concrete buildings
have performed poorly in past earthquakes.

@Seismicisolation
@Seismicisolation
194
Factors enhancing seismic performance requirement is not met for the Type D columns shown in
the plans. This code requirement is intended to prevent
The structure under consideration has several key design buckling of longitudinal reinforcement at locations of high
features that enhance its earthquake resistance, including the seismic loading.
following:
 Joint eccentricity. Eccentricity in the line of action
between beams and columns will amplify loading on the
 Structural configuration. The building is regular in plan,
members.
with no re-entrant corners or vertical off-sets. Regular
buildings have performed well in past earthquakes. Seismic performance of existing building
 Close stirrup spacing at the beam-to-column joints. The
Overview
plans show stirrup spacing of 4-in. on center for beams
and tie spacing of 3.25-in. on center at the joints. The
ASCE 41-13 (ASCE 2014) provides comprehensive
reinforcement is shown with 135-degree hooks. Such
requirements for seismic evaluation and upgrade of existing
close spacing of transverse reinforcement would prevent
buildings and was used for this structure. Computer program
the buckling of reinforcement at the location of highest
ETABS (CSI 2016) was used to prepare a three-dimensional
seismic stress.
mathematical model of the building; see Figure 2. This model
was used to assess the performance of the existing building
Factors decreasing seismic performance moment frames. Nominal material properties, spans and
member sizes specified in the original construction documents
The structure under consideration has several key design were used in analysis. Dimensions were based on centerline
features that reduce its earthquake resistance, including the dimensions provided in the drawings. Gravity loading on the
following: building is composed of member self-weights, design live load
and additional dead load to account for non-structural
 Soft-story response. The first story of the building is elements such as flooring, ceiling, and duct work, which is
approximately 50% taller than the stories above. distributed uniformly on floor slabs. The concrete floor
Buildings with such configuration can be vulnerable to diaphragms are modeled as rigid, meshed shell elements. The
earthquake damage because the deformation and damage seismic loading was based on values obtained from the USGS
is concentrated at the first floor, while good design web site for the design earthquake (475-year event).
typically results in uniform distribution of lateral
deformation among all floors.
 Transverse reinforcement. The beams have stirrup spacing
of 18-in. and 13-in. on center near midspan at the second
level and above, respectively. The mid-height column tie
spacing is 12-in. on center. These values exceed the
current code limits and can lead to premature failure in
some members.
 Shear capacity of beams. Beams are constructed of
lightweight concrete and use No. 3 or 4 transverse bars
spaced 18 in. or 13 in. on center at middle third of the
members, thus having limited shear capacity. Modern
codes attempt to mitigate shear failure by requiring ductile
flexural damage prior to shear failure.
 Splices and development length. The tension lap splices
for the beams do not meet the current code requirements.
The column #18 to #14 longitudinal bar splices use cold-
welded couplers. Inadequate splice and development
length can lead to bar pullout and prevent reinforcement
from reaching its capacity.
 Column ties. The code requires that every other Figure 2. Mathematical model of the building
longitudinal reinforcement have a tie around it. This

@Seismicisolation
@Seismicisolation
195
Building codes allow for both linear and nonlinear analysis. measures should be considered for the first floor or two. For
When linear analysis is used, there are certain conservatisms multistory non-ductile reinforced concrete buildings, drifts
built into the results to account for the modeling and analysis need to be kept to 1.5% or lower.
assumptions. By contrast, nonlinear analysis attempts to
model the behavior of the building and its components in Pushover analysis
greater detail, resulting in greater accuracy of the results,
thusly requiring less conservatism. For this structure, Preliminary nonlinear analysis of the structure was conducted.
nonlinear analysis was utilized to compute capacities and the For this analysis, it was assumed that all reinforcement as
principal of equal displacement was applied to demands. In shown in the plans will be fully developed and that bending
other word, displacement-based (or performance-based), nonlinearity would only occur near the joints. Additionally,
rather than force-based, methodology was employed. given the low capacity of concrete beams in shear, the model
incorporated nonlinear elements at midspan of the beams. Key
Story drift ratios (SDRs) findings are summarized in Table 1.

Figure 3 presents the computed drift ratios at the design Key pushover analysis results
earthquakes. Drift ratios are one of the most telling parameters
in evaluating the response of a building, as they correlate Displacement, in SDR %
directly to the demand on structural members and drift-
sensitive structural components, such as partitions. The Step Level X- Y- X- Y-
building codes place limits on drifts at the design-level Existing Roof 17.0 20.0
1.9 2.2
earthquake. building L1 3.9 4.5
Onset of Roof 14.0 13.0
200
1.6 1.4
180
damage First 3.2 2.9
160 X‐
Y‐ The deformed shapes of the perimeter frame for the building
140
in its existing condition during design earthquake is shown in
120 Figure 4.
Elevation, ft

100

80

60

40

20

0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5%
Drift ratio, %

Figure 3. Drift ratios at design earthquake

Figure 3 reveals the following: Figure 4. Building deformed shape

 Drift ratio in the Y- (transverse) direction are larger than Examination of Table 1 and Figure 4 indicates the following:
in the X- (longitudinal) direction. This is because there are  The building in its existing condition (without upgrade)
fewer moment frame bays in the Y-direction. will experience damage when subjected to the design
 Drift ratios at the first floor are the largest because of the earthquake
soft story present at this level. First floor drift in the Y-  The major damage will be primarily limited to the first
direction exceeds 2%. floor. Since damage is concentrated at one level only, this
For multistory non- or low-ductile reinforced concrete can lead to instability and collapse.
moment frame buildings, the target drift ratio is typically set  If the first floor displacement is reduced below
at approximately 1% to 1.5%. At 1% or below, the structure approximately 2.9 in. (1.4% drift), then damage is
is unlikely to experience any damage. The 1.5% value is essentially eliminated.
referred to as nearly elastic—implying that there will be some
small level of nonlinearity but the damage is likely to be
localized and minor. A review of Figure 3 shows that upgrade

@Seismicisolation
@Seismicisolation
196
Seismic upgrade with fluid viscous dampers
Fluid viscous dampers, were used as the upgrade solution.
Dampers possess the following characteristics:

 Developed by the defense and aerospace industry, and


present state-of-the-art solution for civil applications.
 Are maintenance free and have been widely used in
upgrade of reinforced concrete buildings with proven
reliability.
 Minimize the need for strengthening of existing members
and foundations.
Figure 6. Damper placement, bottom story
 Can be aesthetically integrated into the building
architectural features. Findings
 Have been extensively tested.
Figure 7 presents the computed story displacements for the
 Are cost-effective. upgraded model. The displacement values of Table 1
corresponding to the approximate thresholds for linear
 Minimize disruption to building occupancy.
response (no or minor damage) at the first floor and roof levels
are also shown as arrows in the figure. It is noted that
For this project, dampers were selected with the following displacements of the upgraded model are within the acceptable
properties: limits.
200
 Velocity exponent = 0.5 180

 Damper nominal design force = 350 kips 160 X‐


Y‐
140
Limits
120
Elevation, ft

Seismic upgrade evaluation 100

80
Overview 60

40
The analytical model of the building was revised by adding 8
dampers per floor for the bottom story; see Figure 5 and Figure 20

6. Analysis was conducted to assess the efficacy of the 0


0 5 10 15 20
proposed upgrade. Three pairs of recorded accelerations from Stort displacement, in.
past California earthquakes were synthesized to correspond Figure 7. Computed story displacements
closely to the types of motions that can be anticipated at the
building site during a design-level earthquake. Maximum Figure 8 presents the computed drift ratios for the existing and
values were then selected for assessment of the upgraded upgraded building. The efficacy of the proposed upgrade can
building model. be evaluated by noting the following:
 The soft-story response at the first floor is significantly
reduced. The drift ratio at the first floor was on the order
of twice that of the typical floors above, and this
amplification is now reduced by approximately 60%.
 Drift ratio at first floor is approximately 1.2%. As such,
no or only minor yielding of concrete members is
expected.

Figure 5. Revised building model with dampers

@Seismicisolation
@Seismicisolation
197
200 400

180
300
160 Existing
Retrofitted 200
140
100
120
Elevation, ft

Force, kips
100 0

80
‐100
60
‐200
40
‐300
20

0 ‐400
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% ‐3 ‐2 ‐1 0 1 2 3
Drift ratio, % Displacement, in.

Figure 8. Computed drift ratios (Y direction) Figure 10. First floor damper force-displacement
response
The effectiveness of the damper upgrade solution can further
be seen in Figure 9, where the significant reduction in first Risk (PML) analysis
floor displacement and nonlinear structural damage can be
seen. Probabilistic risk analysis was conducted to compute the
5 probable maximum loss (90% confidence PML) and the
4
Existing scenario expected loss (50% confidence SEL) of the structure
Retrofitted
Limit before and after upgrading. A similar analysis was conducted
3
previously by URS Corporation. The results are presented in
2 Table 2 for both studies.
Displacement, in.

0 PML SEL
‐1 Existing 31 14
‐2
Upgraded 19 9
‐3
Scenario-based risk analysis (design
‐4
earthquake)
‐5
0 5 10 15 20
Time, sec
The following is noted:
Figure 9. Displacement of existing and upgraded
models, First floor, Y-direction
 The reduction in PML and SEL for the upgraded building
is more pronounced due to the differences in upgrade
As shown in Figure 10, viscous dampers dissipate a significant
approaches proposed. The URS proposed upgrade
amount of the earthquake input energy. In the absence of
consisted of wrapping the mid-sections of all frame beams
dampers, such energy must be absorbed by the existing
and columns. While this approach is effective in providing
reinforced concrete members through nonlinear action and
confinement and increasing shear capacity of the concrete
structural damage.
members, it does not reduce the seismic demand or soft-
story behavior. With the proposed upgrade using dampers,
the soft-story behavior is mitigated and demand on the
existing members is reduced to near-elastic levels. Thus
concerns regarding the member ductility are no longer
applicable.

@Seismicisolation
@Seismicisolation
198
Seismic Risk Analysis Procedure viscous dampers. The key input parameters include story drift
ratio, peak floor acceleration, and residual drift; see Figure 13
The FEMA P-58 (FEMA 2012 and SP3 2016) methodology is through Figure 14.
a probabilistic approach that combines the site-specific hazard,
building properties, and exposure to estimate key response
parameters, including the 90th-percentile repair cost in the
event of the design (475-year return period) earthquake. The
simulation for this project included 10,000 Monte Carlo
analyses.

Site Hazard

The site hazard was based on USGS data. The de-aggregation


data from contributing faults is presented in Figure 11 and the
hazard curve for the site is presented in Figure 12.

Figure 13. Story drift ratios

Figure 14. Peak floor accelerations


Figure 11. De-aggregation for the site seismic hazard
Building Capacity

Building capacity was based on the FEMA 154 checklist. The


default values for the collapse fragility were used.
Building Content and Fragility Functions

The building contents (structural and nonstructural) were


based on typical contents for this class of buildings. The P58
default fragility functions were used.
Figure 12. Site hazard curve Results
Building Properties The median and 90% repair costs for the design earthquake
were computed as approximately 6% and 12%, respectively;
The building was defined as a 14-story reinforced concrete see Table 3.
moment frame structure, with lateral frames located at the
building perimeter. It is of 1970s vintage, Risk Category II and
has office occupancy.

The input data for building properties were based on the results
from the structural analysis of the building retrofitted with

@Seismicisolation
@Seismicisolation
199
Median and 90% repair cost
Sa (3.6s) Mean Median 90th percentile
(g) repair cost repair cost repair cost
8% 12% 6% 12%

Project status

The construction for the seismic retrofitting of the towers has


been completed. Figure 15and Figure 16 present a portion of
construction plans showing the damper elevations and
connection to the existing concrete members, respectively. As
required by ASCE 41-13, all dampers were subjected to
production testing conducted by the manufacturer. Figure 17
shows the force-velocity relation for the 350-kip dampers.
Figure 18 presents photographs of retrofitted building
Figure 17. Force-velocity relation

Figure 15. Elevation view of damper placement

Figure 16. Connection details for dampers

Figure 18. Seismic retrofitting using viscous dampers

@Seismicisolation
@Seismicisolation
200
Conclusions

The preliminary probable response of the office towers in


Sacramento to earthquake loading was investigated using a
combination of advanced structural analysis and risk
assessment. Analysis and evaluation showed the following:

 The buildings in their existing configurations had PML


values that exceed 20 and will likely experience moderate
to significant damage in the event of a design-level
earthquake.
 The building structure had a moderate degree of
nonductile detailing as well as soft-story behavior at the
ground floor which pose significant hazards during
earthquakes.
 The buildings were effectively upgraded with fluid
viscous dampers.
 The upgraded solution mitigated the critical building
deficiencies and reduce the PML.

References

ASCE 41-13 (2014) Seismic Evaluation and Retrofit of


Existing Buildings, American Society of Civil Engineers,
Reston, VA.

ASTM E2557 (2016) Standard Practice for Probable


Maximum Loss (PML) Evaluations for Earthquake Due-
Diligence Assessments. American Society of Testing and
Material, West Conshohocken, PA

ETABS (2016), Integrated software package for the structural


analysis and design of buildings, Computers and Structures
Inc., Walnut Creek, CA

FEMA P-58 (2012) Seismic Performance Assessment of


Buildings Federal Emergency Management Agency
Washington, D.C

SP3 (2016) Seimic Performance Predication Program,


Hasleton and Baker, https://www.hbrisk.com/

@Seismicisolation
@Seismicisolation
201
CASE STUDY: 3

Seismic Rehabilitation of Historic Concrete


Structure
With Fluid Visco-Elastic Dampers

Kit Miyamoto, M.S., S.E., President & CEO


Lon M. Determan, S.E., Project Manager
Amir Gilani, PhD, P.E., Project Engineer
Marr Shaffer & Miyamoto, Inc.
West Sacramento, CA

Robert D. Hansen, PhD, P.E., Professor Emeritus


University of Michigan
Walnut Creek, CA

Abstract Introduction

This paper presents the nonlinear seismic analysis, This paper presents the performance-based evaluation
development, and implementation of an innovative and retrofit design of the Hotel Stockton. The 145,000-sf
seismic retrofit strategy for a six-story nonductile (13,470 m2) reinforced concrete building, built in 1910
reinforced concrete 145,000-sf (13,470 m2) historic in Stockton, California, is a torsionally irregular
building. Dynamic and nonlinear static analytical results structure comprised of a six-story portion connected to a
verified that the building had a weak soft-story with two-story portion. There was significant concern that
inadequate post-yield capacity, and large torsional the building will not be able to withstand the level of
response. The analysis indicated that the existing earthquake shaking expected at the site for two reasons:
building is not seismically adequate to withstand a weak and soft lateral force resisting system at the first
anticipated lateral forces generated by earthquake floor level, and the inadequate confinement of
excitations at the site. A “collapse prevention” reinforcement in the first story columns. To assess the
performance upgrade for a 475-year return event was performance of the structure, a detailed mathematical
desired. Nonlinear fluid viscous dampers were placed at model of the building was prepared based on FEMA 273
the first story level to reduce the seismic demand and guidelines. Dynamic and nonlinear static analytical
obtain a more uniform response. Visco-elastic fluid results verified the presence of the soft-story response,
viscous dampers were strategically placed at one side of inadequate post-yield capacity, and large torsional
the building to reduce the torsional irregularity of the response. The analyses indicated that the existing
building. The proposed cost effective, state-of-the-art building is not seismically adequate to withstand
retrofit will improve the seismic performance of the anticipated lateral forces generated by earthquake
building. excitations at the site. The existing structure will suffer
substantial damage and possible collapse in the event of
a major earthquake.

@Seismicisolation
@Seismicisolation
202
To address the above-mentioned inadequacies, the E-W direction and 100 ft (30.5 m) in the N-S direction.
Owner decided to undertake a voluntary seismic upgrade In elevation, it is comprised of a six-story portion on the
of this building. The focus of the seismic rehabilitation east side and a two-story portion on the west side, and
was to address the major deficiencies of the structure, has a full basement. The first story is 18 ft (5.5 m) high
namely the soft-story and torsional response of the and the remaining floors have a story height of 10’-3”
building. The main objective was to provide a “collapse (3.1 m). Figure 1 below shows a south elevation of the
prevention” performance goal during a 475-year return eastern portion of the building.
event. Nonlinear fluid viscous dampers were placed at
the first story level to reduce the seismic demand and In the E-W direction, the building consists of 15 bays at
obtain a more uniform response. Visco-elastic fluid approximately 20-ft (6.1-m) spacing. In the N-S
viscous dampers were strategically placed at one side of direction, there are five bays at approximately 20 ft (6.1
the building to reduce the torsional irregularity of the m) per bay, see Figure 2. The structure is a cast-in-place
building. Finally, the first story interior columns reinforced concrete building. Reinforced concrete
supporting the six-story portion of the building were columns, beams, and shear walls comprise the gravity
wrapped with a fiber-reinforced polymer composite and lateral load resisting system. The basement columns
(FRP). A new mathematical model was prepared are 18- and 20-in. (457 and 508 mm) square for the two-
incorporating the seismic upgrades, and was subjected to story and six-story segments of the building,
nonlinear time history analyses using three sets of two- respectively. At the ground floor and above, column
component, independent acceleration histories derived sizes vary from 18-in. (457 mm) square at the first story
from a site-specific acceleration spectrum. Evaluation of to 14-in. (356 mm) square at the fifth story. There is a
the analytical results of this model showed that the story full 9-in. (229 mm) thick concrete perimeter wall
drift for the first floor was significantly reduced, the between the basement and the first floor, and there are
torsional response was nearly eliminated, and all numerous 6-in. (152 mm) thick concrete walls between
structural members remained elastic. the floors above the second floor. However, there are no
structural walls between the ground and the second floor
Description of Structure levels. Typical floors consist of 4-in. (102 mm) concrete
slabs with a 2-in. (51 mm) topping slab supported by E-
The Hotel Stockton, built in 1910 as a 252-guest room W concrete beams, and N-S concrete girders.
hotel, is a historic landmark building in Stockton,
California. The building, also referred to as The Although the as-built plans of the structure are not
Stockton, measures approximately 300 ft (91.4 m) in the available, field investigations have shown that the

Figure 1: South Elevation

@Seismicisolation
@Seismicisolation
203
columns typically have four and eight longitudinal MPa), per FEMA 273. A yield value of 36 ksi (250
reinforcing bars around the perimeter of columns at the MPa) was used for the column ties.
two-story and six-story segments, respectively. Typical
• Frame elements. All columns were modeled as
minimum concrete cover for the reinforcement is
square sections with longitudinal bars in a circular
approximately 2 ½ to 3 in. (64 to 76 mm). The ground-
pattern. Girders and beams were modeled as
to-first story columns have eight 1-in (25 mm) square
rectangular sections with the section depth measured
bars. Typical transverse ties consist of 1/8-in (3.2 mm)
from the top of the topping slab. T-beam action from
thick by 1-in (25 mm) wide bars at 8 in. (203 mm)
the floor slab was neglected. All dimensions were
spacing.
specified as centerline-to-centerline – (i.e. no rigid
end offsets were specified). The perimeter basement
walls and wall segments between the floors were

Figure 2: Plan View at 2nd Floor


Analytical Model of Existing Structure modeled as shell elements. Similarly, the floor slabs
at all levels were modeled as shell elements. FEMA
The computer program ETABS (CSI 2001) was used to 273 recommends using a value of 50% of the gross
prepare a mathematical model of the building, see Figure moment of inertia (Ig) for the cracked moment of
3 for a schematic of the model. Key features of the inertia (Icr) of the flexural members. This reduction
mathematical model are summarized below. factor was applied to the beams, columns and
shearwalls.
• Material properties. A concrete compressive • Parameters for nonlinear analysis. For this soft-
strength of 3 ksi was used. This value is consistent story structure, the nonlinear behavior will be
for concrete strength of buildings constructed in the entirely limited to the first story columns. As such,
early part of the last century (FEMA 273), and nonlinear hinges were defined and placed on these
corresponds to the values obtained from field columns. To capture the complete nonlinear
investigations. Tensile testing of sample response of these columns, two types of hinges were
reinforcement indicated yield and tensile strengths of used: shear hinges placed at mid-height of the
approximately 65 and 72 ksi (450 and 500 MPa), columns, and biaxial-force (PMM) hinges near the
respectively. Field studies indicated that the column top and bottom of the columns. The location of the
longitudinal reinforcement splice lengths varied PMM hinges was determined by assuming that the
from 26 to 28 inches (660 to 711 mm) for 1-in (25 plastic hinges would form at a distance of 2b/3
mm) square bars. Therefore, the yield properties of (where “b” equals the width of the column) from the
the longitudinal bars were reduced to 42 ksi (290 top and bottom of column-to-floor connections. For
the PMM hinges, interaction curves based on ACI

@Seismicisolation
@Seismicisolation
204
318-99 were used to determine the axial force- columns should remain elastic.
biaxial moment yield surface. For the nonlinear
• Gravity loading. Gravity loads used in the model
analysis, the column plastic hinge properties are a
consisted of the self-weight of the structure, 0.02 ksf
function of column slenderness, transverse
(0.96 kPa) for partitions, 0.025 ksf (1.20 kPa) for
reinforcement (size, spacing, and anchorage), and
weight of the 2-in. (50-cm) topping slab, and 0.005
axial and shear demand. For the columns under
ksf (2.40 ksf) for miscellaneous (e.g., fans, vents,
consideration, the axial force ranges between 10-15
plaster). Live loads consisted of typical code
percent of the nominal compressive strength, and
prescribed floor loads.
flexure is the controlling response. The shear force is
less than three times the nominal shear strength, and • Inertial mass. The mass of the structure consisted of
the columns have poor confinement (transverse all structure dead loads and one half of the partition
reinforcement). Since the lap splices for the loads. The code-mandated 5-percent eccentricity
longitudinal reinforcement are not fully developed, was achieved by offsetting the floor mass. The total
sudden strength degradation may occur after the inertial weight (mass) of the structure is
onset of the nonlinear behavior. Therefore, hinge approximately 14,000 kips (64,050 kN).
formation (yielding) should be avoided, and subject

Figure 3: Schematic of The Mathematical Model of The Building

@Seismicisolation
@Seismicisolation
205
Earthquake Histories from the horizontal components of each of the three
recorded earthquake records to the target spectra, and
Site investigations were used to determine the site- then base-line correcting in the time domain. The
specific acceleration spectra. The Design Basis records were derived from the 1989 Loma Prieta
Earthquake (DBE) spectra (10% probability of earthquake (0- and 90-degree components recorded at

0.8

DBE (PGA = 0.22g) MCE (PGA = 0.34g)

0.6
Spectral Acceleration (g)

0.4

0.2

0.0
0.0 1.0 2.0 3.0 4.0
Period (sec)

Figure 4: Response Spectra


exceedence in 50 years) used for the studies presented Agnews, 0- and 90-degree components recorded at
herein is shown in Figure 4. In the same plot, the Gilroy, and 270- and 360-degree components recorded at
response spectrum for the Maximum Considered San Marino). Figure 5 shows the DBE acceleration
Earthquake (MCE) (2% probability of exceedence in 50 record and the computed acceleration spectrum for the x-
years) is also shown. component of the Agnews record. The y-component of
this record and the x- and y-components of the other two
Three sets of time histories were prepared by J. P. Singh records have similar acceleration spectra.
(Singh 2002) by matching the response spectra derived
0 .3
0 .2
0 .1
accel, g

0
-0 . 1
-0 . 2

0 5 10 15 20 25 30
t im e , s e c
0 .8

0 .6
sa, g

0 .4

0 .2

0
0 0 .5 1 1 .5 2 2 .5 3 3 .5 4
p e rio d , s e c

Figure 5: Acceleration Record & Spectrum

@Seismicisolation
@Seismicisolation
206
Analytical results of the existing building
building torsional component to this mode. In particular, the
largest deformation occurs at the far right (east) side of
Dynamic analysis. A modal analysis of the building was the building. This torsional response will place
conducted to determine the fundamental period and additional demand on the columns at this side of the
mode shapes of the structure. Table 1 summarizes the structure.
results for the first three modes obtained. The first two
mode shapes of the building are shown in Figure 6a and Nonlinear pushover analysis. To assess the performance
6b. It is noted that the response is that of a soft-story of the building to seismic loading, a nonlinear static
structure with nearly all the deformation concentrated in analysis was conducted. The structure was initially
the first story columns. loaded to a gravity loading equal to 110% of the dead
load and 27.5% of the unreduced live load. Next, step-
Table 1 by-step lateral loading in the x- and y- directions were
Mode Period (sec) Principal direction applied to the structure. Two separate and independent
1 1.2 Transverse (N-S) lateral load patterns were considered: (1) a force pattern
2 1.0 Longitudinal (E-W) matching the mode shapes with 100% and 30% loading
3 0.9 Torsion in each direction and (2) uniform force pattern with
100% and 30% loading in each direction. For the

Ground Level

Figure 6: Modes 1 & 2 (Elevations)


Figure 7 shows the deformed shape of the second floor governing load case, the demand and capacity curves do
in plan for the first mode. It is noted that due to the lack not intersect. Therefore, collapse of the structure is
of symmetry in the N-S direction, there is a large predicted.

Figure 7: Mode 1 (Plan)

@Seismicisolation
@Seismicisolation
207
Two particular events of interest were studied: (1) when
does the first plastic hinge form in each direction, and The displaced structure at a story displacement of 1.5 in.
(2) what is the ultimate configuration of the plastic (38 mm) for loading along the longitudinal-axis (Figure
hinge? (The plastic hinges are identified by circles on 9a), and 2.5 in. (64 mm) for loading along the
the columns.) The displaced shape of the structure at the Transverse-axis are shown (Figure 9b). Note that many
formation of the first column plastic hinge is shown in of the first story columns have formed plastic hinges at
Figure 8. The frame elevation on the left corresponds to the top and bottom. The soft-story behavior of the

First Hinge First Hinge

Figure 8: Hinge Initiation

the formation of the first plastic hinge when the structure building is made clear in the figures; all the floors above
is pushed along the longitudal-direction. This yielding the second floor have a nearly rigid behavior, while the

Figure 9a & 9b: Progressive Hinge Formation (elevations)

response occurs at a displacement of 0.84 in. (21 mm), first story columns experience substantial deformation.
measured at the second floor level. The frame elevation
on the right corresponds to the formation of the first Figure 10 shows the second floor plan view of the
plastic hinge when the structure is pushed along the structure at the deformation level of 2.5 in. (64 mm) as
transverse-direction. This yielding response occurs at a the structure is pushed in the transverse-direction. It is
displacement of 1.44 in. (37 mm), measured at the right noted that all the nonlinear behavior is concentrated at or
(east) side of the second floor level. In summary, as close to the right (east) side of the building. As
long as the second floor displacements are limited to the previously noted, the building is torsionally irregular in
values specified above, it is expected that the column the transverse direction.
response for the critical first story columns will remain
in the elastic range.

@Seismicisolation
@Seismicisolation
208
Figure 10: Progressive Hinge Formation (Plan)

Linear time history analyses. To investigate the Performance Goal. Since this is a voluntary seismic
performance of the building during a 475-year return upgrade, the focus of the seismic rehabilitation is to
event (DBE), the structure was subjected to acceleration address the major deficiencies of the structure, namely,
time histories. Study of the three motions revealed that the soft-story and torsional response of the building.
the San Marino record produced the most severe test for The retrofit will limit the response of the structure to
the structure (i.e., the largest values of column stress and linear elastic behavior; that is, limiting the maximum x-
story drift). As such, this record will be used for the and y-components of the second floor displacement to
remainder of this paper for comparison purposes. 0.85 in. (22 mm) and 1.44 in. (37 mm), respectively.
This will give an adequate level of confidence against
The three dimensional linear model was subjected to this collapse of the structure. The main performance goal is
accelerogram. Figure 11 shows the second floor to provide a cost-effective “collapse prevention”
displacements as measured at the lower-right (S.E.) performance upgrade during a 475-year return event
corner of the building. Using equal displacement (DBE).
assumption, a comparison of the time history response of
the existing building with that of the nonlinear pushover Retrofit method
analysis indicates that the story drifts will cause
significant plastic rotation in the hinge regions of the To meet the selected performance goals for the upgrade
columns, and cause probable collapse of the building.

5.0

2.5
Displ, in.

0.0

-2.5

-5.0
0 10 20
Time, sec

Figure 11: Displacement History of Existing Structure (Linear Model)

@Seismicisolation
@Seismicisolation
209
of this structure, a retrofit approach combining several were considered for the upgrade, however, this approach
state-of-the-art strategies was utilized. necessitated using relatively large devices to meet the
performance criteria. In addition, this did not address
1. Reduce the soft-weak story effects by increasing the
the torsional irregularity of the building. To mitigate
effective damping of the structure. This objective
these problems, two types of devices were utilized:
was achieved by employing Fluid Viscous Dampers
nonlinear fluid viscous dampers were used in 16 braced
(FVD) at the first floor.
bays, and a combination of nonlinear fluid viscous
2. Reduce the torsional response of the building dampers in parallel with elastic elements (herein referred
without increasing acceleration demand of the super to as fluid visco-elastic dampers, or FVEDs) was utilized
structure. This was achieved by adding fluid visco- in four braced bays. The table below summarizes the
elastic dampers at the east side of the structure. pertinent properties of the devices.
3. Provide a more redundant story shear capacity in the
Table 2: Damper Properties
upper floor transverse direction. In the transverse
Device No. DBE c, k-sec/in α K, k/in
direction, the building has structural walls at the
Capacity, (kN- (kN/mm)
exterior walls only. Therefore, wood shear walls
kip (kN) sec/mm)
were added for the upper six story portion of the
FVD 16 210 (934) 100 (35) 0.5 None
building. These walls will act in a fashion
FVED 4 300 (1334) 125 (44) 0.5 144 (50)
analogous to cross-walls in an unreinforced masonry
(URM) bearing wall building. Additional columns at either end of the diagonal devices
will prevent the transfer of the damper forces to the
4. Provide redundancy for the gravity load-carrying existing building columns. Figure 12 shows a typical
capacity of the columns along the right (east) side of damper frame elevation.
the structure. Addition of steel columns for the FVD
braces adjacent to all the columns along this gridline
met this goal.
5. Increase ductility of all the interior first story
columns for the 6-story segment of the building. To
meet this criterion, fiber-reinforced polymer
composite (FRP) was wrapped around the hinge
regions (top and bottom) of the columns.

Structural upgrade

FVDs have been extensively researched (Constantinou


and Symans, 1992) and implemented in the upgrade of
many structures, including the seismic retrofit of the
historic Hotel Woodland (Miyamoto and Scholl 1996).
FVDs provide an economical way of improving the Figure 12: Damper Frame Elevation
structural response without losing any floor space. This
was the chosen seismic improvement method for this
building for two reasons: (1) it reduces the second floor Fluid Visco-Elastic Damper
displacement by increasing viscous damping, and (2) it
reduces the seismic demand of the superstructure. A combination of fluid viscous dampers and poly-
urethane elastomers have been successfully used in the
Damper selection. FVDs were strategically placed in aerospace industry. The mechanical characteristics of
the structure to optimize their effectiveness without this elastomer are as follows: (Gallagher Corporation,
blocking access to the architecturally sensitive areas of 2002)
the ground floor. A total of 20 damper bays were
utilized. Initially, only linear fluid viscous dampers

@Seismicisolation
@Seismicisolation
210
1. Urethane Elastomers provides consistent mechanical incorporating the dampers were performed. The
properties through a temperature range of 0ºF to mathematical model of the existing building was
225ºF (-18ºC to 107ºC). modified by adding the sixteen FVDs and the four
FVEDs. Two time history cases were considered. In
2. Urethane exhibits compressive capacity of 80 ksi
one case, the mathematical model was preloaded by a
(552 MPa) without molecular damage and elasticity.
static load equal to 90% of the total dead load prior to
3. Aging under static stress has no effect on mechanical being subjected to the lateral accelerations. In the
properties if protected from ultraviolet light. second case, the preload equaled 110% of the dead load
4. Flame resistance is sufficient to meet Federal and 27.5% of the unreduced live load. The envelope of
Aerospace Regulation 25.853B. response quantities was then obtained by selecting the
maximum values from the two load cases.
See Figure 13 for FVED and FVD construction.
Response evaluation. To evaluate the seismic response
Prototype testing per FEMA 273 will be conducted to
of the upgraded structure, the displacement response of
verify response and durability.
the second floor was examined and a stress check of all

FVED FVD
Figure 13: FVED & FVD Devices
Response of the retrofitted structure first story columns was performed. Figure 14 shows the
second floor displacement responses for the lower-right
To assess the effectiveness of the proposed building (S.E.) corner. It is noted that the maximum computed
upgrade, nonlinear time history analyses of the structure displacements are approximately 0.56 in. (14 mm) and

Displacement History
1.0

0.8

0.6
0.56

0.4
Displacement, in.

0.2

0.0

-0.2

-0.4
x-Direction y-Direction
-0.6

-0.8
-0.85

-1.0
0 5 10 15 20
Time, sec

Figure 14: Displacement History of Damped Structure

@Seismicisolation
@Seismicisolation
211
0.85 in. (22 mm) in the longitudal transverse directions, structure was upgraded with a combination of sixteen
respectively, which is well below their target values. nonlinear fluid viscous dampers, four nonlinear fluid
This corresponds to story drift ratios of approximately visco-elastic dampers, and fiber reinforced polymer wrap
0.003 and 0.004, respectively. A comparison of the at selected columns. The analytical studies predict that
displacement response for the original structure and this the retrofitted structure will have a significantly
figure shows that the maximum response was reduced by improved performance when compared to the existing
more than a factor of five by the addition of FVD and structure. In particular, the upgrade will limit the
FVED elements. response of the existing members to the linear range by
limiting the expected seismic demand on the structure.
Finally, the computed axial force in the columns was This upgrade will reduce the risk of building collapse.
examined. No net axial tension was found in the Total seismic upgrade cost was $1.3 million ($9/ft2,
existing columns. The maximum force in the FVD was $96/m2), which was about 5% of total construction
less than 200 kips (890 kN). As such, the 200-kip (890 budget ($24 million, $165/ft2, $1780/m2).
kN) dampers used are adequate for these 16 damper

Typical FVED Response


300
291

200

100 Elastomer 86
Response
Force, kip

-100 Combined

Viscous
-200 Response

-2 4 9

-300
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Displacement, in.
Figure 15: Typical FVED response
bays. Figure 15 depicts the response of a typical FVED. References
It is noted that the maximum damper and spring forces
are approximately 250 kips ( 1112 kN) and 90 kips (400 Computer and Structures, Inc., 2002, “ETABS 7.2.2,
kN), respectively. Spectra acceleration of this structure Linear and nonlinear static and dynamic analysis and
was 0.17g. design of building systems,” Berkeley, CA
Conclusion FEMA 273, October 1997, “NEHRP Guidelines for the
Seismic Rehabilitation of Buildings,” Building Seismic
Analytical studies of the Hotel Stockton revealed that the Safety Council, Washington, D.C.
structure would not be able to withstand the seismic
loading resulting from the anticipated site-specific Miyamoto, H.K. and Scholl, 2002, “Seismic Rehabilitation
earthquakes. To mitigate this seismic deficiency, the of an Historic Non-Ductile Soft-Story Concrete Structure

@Seismicisolation
@Seismicisolation
212
using Fluid Viscous Dampers,” Proceedings of the 11th
World Conference on Earthquake Engineering

Constantinou, M. and Symans, M., 1992, “Experimental &


Analytical Investigation of Seismic Response of Structures
with Supplemental Fluid Viscous Dampers,” State
University Of New York at Buffalo, Buffalo, New York,
NCEER-92-0032

Sing, 2002, Site Specific Time Histories for Hotel


Stockton, Richmond, CA

Gallagher Corporation, 2002, “The Engineering


Properties of Urethane Elastomers,” Gurnee, IL

@Seismicisolation
@Seismicisolation
213
CASE STUDY: 4

Seismic Upgrade of an Existing Tall Building by Different


Energy Dissipation Devices

Shanshan Wang, Graduate Student Researcher


Stephen Mahin, Professor
University of California, Berkeley
Berkeley, CA

Abstract 01. Introduction

The Pacific Earthquake Engineering Research (PEER) Center In traditional design where seismic energy is mainly
has expanded its Tall Building Initiative (TBI) program to dissipated by irrecoverable inelastic deformation of structural
include the seismic performance of existing tall buildings. A elements, the building safety is maintained at the compromise
35-story steel moment resisting frame, designed in 1968, and of components’ damage, leading to direct and indirect
had representative details of buildings between 1960 to 1990 economic losses. This has been highlighted in recent
was selected for detailed seismic evaluation in the framework earthquakes in Chile, Japan, China and New Zealand. As
of Performance Based Earthquake Engineering (PBEE). It such, the development of seismic protection systems has been
was identified that the case study building failed to meet the spurred, which includes base isolation, active control and
performance objectives suggested by ASCE 41-13, and had a passive energy dissipation systems by large (Soong and
number of seismic vulnerabilities that endangered its Spencer, 2002). Of these, passive energy dissipation systems
structural integrity at two basic safety earthquake hazard do not require external power source, and are relatively easy
levels (BSE): BSE-1E and BSE-2E. Therefore, exploration to install, and thus considered as a better choice to upgrade
of retrofit strategies and their cost-effectiveness are fostered. existing structures. Three kinds of devices are investigated in
In this paper, three kinds of supplemental energy dissipation this paper: fluid viscous dampers (FVDs), viscous wall
devices are investigated to upgrade the seismic performance dampers (VWDs) and buckling restrained braces (BRBs), and
of the case study building, including fluid viscous dampers they are used in combination with preliminary retrofit
(FVDs), viscous wall dampers (VWDs) and buckling methods to upgrade an existing 35-story Pre-Northridge steel
restrained braces (BRBs). The retrofit design started by moment resisting frame. The investigations focus on
selecting locations to install supplemental devices. Then the comparing the cost-effectiveness of each retrofit method, and
total effective damping ratios needed to achieve the target also raise critical design considerations that appear for each
roof displacements in two directions were estimated based on strategy. Fig. 1 illustrates the applications of these three
a damping scale factor (DSF). One retrofit strategy by using devices.
FVDs was investigated as a first trail, and the mechanical
characteristics of each damper device were calculated based
on the overall effective damping ratio and the story wise
distributions of dampers. Next, other two retrofit strategies
by using VWDs or BRBs were investigated. Sizing of
different devices at one location was performed following the
principle of equal energy dissipation. The effectiveness of
each strategy to meet the retrofit intent of ensuring structural
stability at BSE-2E were compared. Moreover, probabilistic
damage and loss analysis were conducted using Performance
Assessment Calculation Tool (PACT) to relate the structural (a). FVDs
responses to economic losses. After a detailed examination, Figure 1. Supplemental energy dissipation devices
it was found that upgrading the case study tall building using
FVDs was the most effective retrofit strategy to control
structural responses, and reduce damage and economic losses
after BSE-2E events.

@Seismicisolation
@Seismicisolation
214
A systematic structural evaluation of the case study building
indicated that the building failed to meet the performance
objectives suggested by ASCE/SEI 41-13, Seismic
Evaluation and Retrofit of Existing Buildings (ASCE 41,
2013). Similar outcomes were found using FEMA 351,
Recommended Seismic Evaluation and Upgrade Criteria for
Existing Welded Steel Moment-Frame Buildings (FEMA
2000), and FEMA P-58, Seismic Performance Assessment of
Buildings (FEMA 2012a-c), showing that the building had a
(b). VWDs variety of seismic vulnerabilities, and suffered great damage
and economic losses after basic safety earthquake hazard
level (BSE) events. Consequently, feasible upgrade
strategies are necessary to enhance the seismic performance
of the building. The intent of the retrofit is to reduce the
overall drifts of the structure to a level where brittle fracture
of the beam-to-column connections would not seriously
jeopardize the overall stability of the structure at a basic
safety earthquake, level 2 (BSE-2E) hazard (with a
probability of exceedance of 5% in 50 years). To achieve this,
a “two-stage” retrofit plan was proposed. In “Stage-1”, the
(c). BRBs prevalent brittle column splices were fixed everywhere, and
Figure 1 (continued). Supplemental energy dissipation heavy concrete claddings were removed. The retrofit resulted
devices in a change of elastic model periods, that were 4.33 sec. (X-
direction translation), 4.18 sec. (Y-direction translation), and
02. Evaluation of the Case Study Building 3.59 sec. (rotation). Nevertheless, these strategies were
demonstrated insufficient to achieve the target performance
A 35-story steel moment resisting frame that had goal, and thus additional strategies are necessary. In “Stage-
representative design details from the period of 1960 to 1990 2”, three different kinds of supplemental energy dissipation
was selected for systematic seismic evaluations. The case devices were used in conjunction with strategies used in
study building is about 490 ft. tall, with a typical floor height “Stage-1” for further improvement, including FVDs, VWDs
of 13 ft. It spans about 185 ft. by 135 ft. in plan, and has a and BRBs. This paper focuses on the “Stage-2” retrofits, and
typical beam span of 30 ft. The building completed compares the cost-effectiveness of upgrading the case study
construction in San Francisco in 1971, consisting of complete tall building using selected devices.
three-dimensional moment-resisting space frames in both
longitudinal direction (X) and transverse direction (Y). Fig. 2
shows a frame elevation and a floor plan of the building
model.

Beam-to-column moment connections used typical pre-


Northridge details and the column slices were erected using
partial joint penetration welds, both considered to be quite
brittle. A three-dimensional (3D) model was generated using
the program: Open System for Earthquake Engineering
Simulation (OpenSees, Mckenna et. al., 2010) to investigate
its nonlinear dynamic behavior; see Fig. 3. All aboveground
main framing members contributing to the seismic lateral
force resisting systems were included (see Fig. 2 and Fig. 3).
Given the nature of this investigation, certain modelling
simplifications were made. For instance, the basement levels
were disregarded. The first three elastic modal periods of the
structure were: 4.70 sec. (X-direction translation), 4.53 sec.
(Y-direction translation), and 4.15 sec. (rotation). More (a) (b)
details about the building information and numerical Figure 2. Illustration of the building model: (a) plane view of
modeling could be found in a Lai et. al. (2015). a typical frame in X-direction; (b) floor elevation

@Seismicisolation
@Seismicisolation
215
efficiency. Effective damping ratios were estimated based on
the target roof displacement in each direction at BSE-2E
events. The mechanical properties of dampers were selected
using an equation related to the effective damping ratio, and
story wise distributions of dampers. Several design
considerations and viable alternatives to address these
considerations were raised at the end of this section.

4.1 Damper locations

To initialize the design within an existing building, the first


consideration was selecting proper locations to install
supplemental devices. Fig. 5 shows the plane view of a
typical floor based on the architectural drawing, where the
black boxes indicate column locations. The interior frames
are usually adjacent to stairs and elevator locations, and
putting dampers there would interfere with office space and
egress. Therefore, the perimeter frames are considered as
Figure 3. Illustration of the building model better options to add dampers. At first few trials, FVDs were
installed along all stories with three different distribution
03. Analysis Method patterns of the damping constant C: (I) uniform; (II)
proportional to story shear; and (III) proportional to story
Nonlinear response history analysis (NRHA) was used to stiffness. A preliminary analysis found that scheme (III),
examine the seismic responses of the existing building using with damping constant C values proportional to story
OpenSees. The retrofit feasibility study focused on BSE-2E stiffness, was the most efficient among the three and thus
hazard level. Three ground motions were selected at this selected for continued refinement. The initial design was
hazard level for preliminary design and comparison, and the refined by removing dampers, or adjusting damper sizes
selection criteria was the closeness of their pseudo- based on the control effectiveness of dampers at different
acceleration spectra to the target spectra near the fundamental regions. Consequently, a refined design scheme with
period of the original building, as shown in Fig. 4. “Stage-2” concentrated dampers at lower two-thirds of the building was
retrofit started with a numerical model with “Stage-1” retrofit, proposed; see Fig. 6. Damping exponent α was set to be 0.35
i.e., the column splices were fixed everywhere, and concrete to ensure adequate control effectiveness without excessive
cladding were removed. This baseline model was denoted as damper forces, based on a parametric study. In a single
“as-built” hereafter. frame, dampers were distributed across multiple bays to
minimize accumulation of forces transferred to the adjacent
columns.

Figure 4. Target response spectrum and selected ground


motions at BSE-2E

04. Retrofit with FVDs

The “Stage-2” retrofit plan started with design of FVDs. The


locations for installing devices were selected considering the Figure 5. Openings at building floor
architectural restraints, constructability as well as damper

@Seismicisolation
@Seismicisolation
216
during dynamic vibration is neglected (Constantinou and
Symans 1992; Reinhorn and Constantnou 1995):
Fd = Cvα·sign(v) (1)
where C is the damping constant, α is the damping exponent,
and sign (v) is the sign function of relative velocity of the
piston end with respect to the damper housing. In earthquake
engineering, α is generally in the range of 0.3 to 1.0 (Lee and
Taylor 2001). Eq. (1) could predict behaviors of a FVD well
for low rate excitations, but the frequency-dependent contents
need to be accounted as the operating frequency increases.

To model the FVD in OpenSees, a viscous damper material


was used to represent the damper sub-assemblage: a dashpot
and a spring in series. The dashpot resembled the pure
viscous behavior, as described by Eq. (1); the elastic spring
element represented the driving braces. Researches (Fu and
Kasai 1998; Takewaki and Yoshitomi 1998) have found that
the brace flexibility would influence the damper behavior
significantly and shall be fully accounted for. In this study,
Figure 6. Story wise damper distributions the total brace stiffness per story was equal to twice of the
story stiffness, which was proven to be rigid enough to ensure
4.2 Effective damping ratios adequate effect of FVDs.

Estimating the overall effective damping (including the 4.4 Design considerations
intrinsic damping and supplemental damping) needed to
reduce the overall drifts is a prerequisite to estimate For a high-rise building, fairly large dampers are usually
additional damping demand. A non-iterative approach was required to achieve the target performance goal, and this
used based on researches of Rezaeian et al. (2012). In this poses great challenges to existing buildings. Issues such as
approach, a Damping Scale Factor (DSF) was developed to delivering heavy devices to multiple stories and clearing
adjust the 5% damped spectral ordinates to damping ratio structural/non-structural components would increase
ranging between 0.5% and 30%, which is defined as the ratio construction difficulty and retrofit costs, and need careful
between the target overall displacement to the current considerations. Alternatives such as using two dampers per
displacement demand. The target roof displacement at each driver, more damped bays at selected stories, and utilizing
direction was selected based on the static pushover curves toggle-brace mechanisms to magnify the effective force of a
when the original building abruptly lost more than 70% force damping device (Taylor and Constantinou 1998) might be
resistance capacity. Meanwhile the current displacement considered. On the other perspective, reduced performance
demands were estimated from the displacement spectrum at objectives might be used.
BSE-2E event. The DSF was then related to a regression
relation derived based on the entire NGA-W2 earthquake Another critical design consideration is the vulnerable
record set (Rezaeian et al. 2012). Variations of magnitude, columns. After implementing “Stage-1” retrofits, the
source-to-site distance and local site conditions have been columns are anticipated to have adequate tension capacities,
considered in the regression relation. With a calculated DSF but they might still be overloaded in compression. Using
of each direction, the required damping ratios at a BSE-2E FVDs could bring down the drift ratios and reduce the axial
event were estimated, which were 8% for X-direction and 13% forces and bending moments in columns. Nevertheless, an
for Y-direction. excessive accumulation of damper forces on adjoining
columns would cause problems if the structure enters into
4.3 Mathematical modeling inelastic range, and the damper forces are large. Other
factors such as the flexibility of connecting elements (e.g.,
General fractional derivative Maxwell model was described driving braces, girders, connections and columns) would
by Makris and Constantinou (1990) to capture the behavior of drive the dampers to act more in-phase with peak
FVDs, whereas a simplified mathematical model (Eq. (1)) displacement and add up to the total forces in columns.
could be used if the operating frequency is under the cut-off Additional retrofit methods such as filling the columns with
frequency of a FVD, that is, the stiffening effect of a FVD concrete, constructing mega columns at the corners might be
investigated.

@Seismicisolation
@Seismicisolation
217
05. Retrofit with VWDs

The retrofit design scheme of using VWDs followed that of


FVDs, and used consistent installation locations. The
mathematical model and mechanical properties of VWDs are
presented in this section. One of a unique design issue of
VDWs is highlighted in the end.

5.1 Mathematical modeling

A Kelvin material model (Eq. 2) with linearized parameter


could be used to simulate a viscoelastic damper with mild
frequency-dependence (Lobo et. al. 1993):
Fd = Ku + Cv α (2)
where C is the damping constant, α is the damping exponent, Figure 7. VWD modeling
and these two parameters of each VWD are kept the same as
that of a FVD at the same location. u is the relative 5.2 Design considerations
displacement of the two steel plates. Another parameter K,
representing the storage stiffness of VWD tanks is introduced, VWDs could provide more architectural flexibility than the
which represents the capability of a VWD to provide brace-type dampers (e.g., FVDs, BRBs), and they could
additional stiffness. Recent tests on VWDs in the United provide both additional damping and additional stiffness.
States (Newell et. al. 2011) showed the stiffening of the However, several considerations call for special attention. As
structure due to wall dampers was about 5%, and thus the with the case to install FVDs, using VWDs would bring
stiffness parameter K of 1000 k/in was used for all VWD about similar design considerations such as the large damper
elements in this study for simplicity. sizes, vulnerable columns etc. Additionally, the effect of a
VWD in the frame would be quite sensitive to the behaviors
In the numerical model, a beam was modified if a VWD of beams connected with it. For an existing steel building
element was inserted into its middle span; see Fig. 7. First, having Pre-Northridge moment connection details, this
an additional node was created in the central bay of a beam, problem is a bigger concern and thus presented below.
and the beam was discretized into two elements. The
alternation was made for both the upper and lower beams First of all, the effect of a VWD will be greatly influenced by
connected to a VWD. To model the VWD element, a two the behaviors of beams connected with it. Fig. 8 shows the
node link element was generated in OpenSees that connected deformed shapes of a frame with two different cases. In case
the two middle nodes at upper and lower beams, and a 1, if a beam is strong and could provide full restraint against
viscous damper material model and an elastic material model bending, it would deform as an end fixed beam, and its
were used in parallel to represent Eq. (2). The parameters of deformed shape could produce large relative movement
a viscous damper material were identical to the FVD element between two steel plates; see Fig. 8(a). Consequently, a large
at each location, and the elastic material used a stiffness K amount of energy dissipation would be produced. On the
equaling 1000 k/in, as discussed before. The materials were other hand, in case 2, the beams-to-column connections are
applied in the direction of in-plane movement of the VWD. released, and the beam would deform without rotational
With introduced additional stiffness, the overall fundamental constraints. As such, the deformed shape of the upper and
period of the case study building shifted from 4.33 sec. (have lower beams would limit the relative displacement between
“Stage-1” retrofit only) to 4.10 sec. two steel plates, and result in insufficient energy dissipation
of the VWD; see Fig. 8(b). Thereby, in order to ensure
adequate energy dissipation provided by VWDs, the beams
need to be strong and provide enough rotational constraints.
However, the Pre-Northridge beam-to-column connections
used in the case study building make it hard to meet this
criterion, since a great many of these connections exhibited
brittle failure (Lai et. al. 2015). Once the beam end
connections fail, the beam would deform like case 2,
significantly diminishing the energy dissipation capacity of
the connected VWD element. It should also be highlighted
that a sudden change of the deformed shapes upon the beam

@Seismicisolation
@Seismicisolation
218
failure might produce a spike of force and deformation on the
VWD, and bring about the rupture of viscous material and the
failure of the wall damper.

Figure 9. Deformed shape of a frame when the storage


stiffness of a VWD is large

Thirdly, the seismic demands on the structural components


(a). Case 1: deformed shape with fixed-end connection would be increased with the additional stiffening effect,
which in turn would cause an earlier fracture of beam-to-
column connections. Both the bending moment demands and
the shear demands would be increased. It should be noted
that the increase of the shears also depends on the aspect ratio
(defined as its width-to-depth ratio) of a VWD. The shear
forces are amplified by the inverse of the aspect ratio. Thus
if a small aspect ratio is used, beam would end up with large
shear force, and suffer from shear yielding. Nevertheless, the
numerical model did not include the shear yielding, and the
results would be optimistic.

06. Retrofit with BRBs

(b). Case 2: deformed shape with pin-end connection BRBs are cheaper than FVDs or VWDs, and they are
Figure 8. Deformed shape of a frame with different boundary considered as ordinary braces in the U.S. design code, which
conditions make their design and analysis procedures less complicated
than other supplemental energy dissipation devices. As with
Secondly, the storage stiffness of a VWD would affect the previous two retrofit methods, the distributions of BRBs in
beam deformations. If the storage stiffness of a VWD under the existing building followed the pattern with FVDs and
a dynamic loading is large, it would prevent the two steel VWDs. In this section, the mathematic modeling of a BRB
plates moving freely (see Fig. 9), and resulted in reduced in OpenSees and the major design considerations would be
energy dissipation capacity of a VWD. Note that the storage discussed.
stiffness of a VWD is not an exclusive contribution from the
steel plates; the frequency-dependent part of the viscous 6.1 Mathematical modeling
material would also influence the storage stiffness (Fu and
Kasai, 1998). More research is needed to understand the BRB is a kind of displacement-dependent devices, and it
relation between the VWD storage stiffness, beam stiffness, dissipates energy through the yielding of the brace. The basic
and their effect on the behavior of the VWD. force-displacing relation of a BRB is expressed as:
Fd = Ku (3)
where K is the effective stiffness of the brace, and u is the
relative displacement between two ends of a brace. When a
brace is in its elastic range, K represents the elastic stiffness.
After it yields, a post-yield stiffness in the order of 0.001 of
the elastic stiffness is used to represent its force resistant
capacity. This strain hardening value of 0.001 is
recommended in the OpenSees manual (Mazzoni et al. 2006),

@Seismicisolation
@Seismicisolation
219
which could control the transition from elastic to plastic direction follow a similar trend. It should be noted that
branches and accounts for isotropic hardening. during the simulation (entire ground motion duration plus 15-
second free vibrations), most numerical simulations were
To simulate the behavior of a BRB, a co-rotational truss successful; however, in the case with VWDs, several VWDs
element was used in OpenSees. The material model used a were broken under one ground motion excitation after the
Giuffre-Menegotto-Pinto model (Steel02), and was assigned connected beams failed, and the structure had a peak drift
in the axial direction of the element. The effective stiffness ratio in excess of 10%. In this case, the numerical analysis
K0 in the elastic range was estimated using the principle of was arbitrarily terminated since the building was most likely
equal energy dissipation. This was calibrated by assuming to collapse.
that that the peak force F0 of a FVD and a BRB would be the
same when they reached a same peak displacement U0 ; see 7.1 Global responses
Fig. 10. The proposed simple model was adequate to capture
the primary characteristics of BRBs, e.g., the Bauchinger The peak displacement distributions shown in Fig. 11
effect and strain hardening effect, and thus selected for this indicate that all cases incorporating different devices could
study. The stiffening effect of BRBs changed the help reduce the structural deformations by a large amount,
fundamental period of the building from 4.33 sec. to 4.05 sec. ranging from 20% to 40%. With a same effective damping
ratio, they help bring down the peak roof displacement to a
similar value, and the value is close to the selected target roof
displacement, i.e., 38 inches in the X-direction. This
demonstrates that the DSF method discussed in Section 4.2 to
estimate the effective damping ratio is adequate for
preliminary design of FVDs.

Figure 10. Illustration of hysteresis loops of a BRB and a


FVD

6.2 Design considerations

When use BRBs, design considerations such as large peak


forces of BRB devices, and overloaded columns exist, as with
other retrofit methods. Nevertheless, their displacement-
Figure 11. Distributions of peak displacement in X-dir.
dependency would force BRBs to act more in-phase with
peak displacement, and the peak force demands on existing
Among three retrofit schemes, the strategy of using FVDs is
beams and columns would be increased at a larger extend
considered as the most effective to eliminate the concentrated
than other two methods.
drift ratios at floor level 2-10, and contributes to a more
uniform distribution of the peak deformations; see Fig. 12.
07. Comparison of Control Effect
On the other hand, the retrofitted schemes using VWDs or
BRBs could not achieve satisfied control effect and still have
The results of global structural responses, damper/BRB
a peak drift ratio in excess of 3% at floor 3 to floor 7. Such a
behaviors, and column axial force status are presented for the
large drift ratio indicates the latter two retrofitted methods are
“as-built” building with “Stage-1” retrofits only, and three
not able to meet the retrofit intent of maintaining structural
retrofitted cases with different supplemental energy
stability at BSE-2E events.
dissipation devices. The maximum results from three
nonlinear response analyses are used, as stipulated by ASCE
41-13. Two horizontal directions are evaluated separately,
while only the X-direction drift ratios and floor accelerations
are shown for discussions, and the global responses in the Y-

@Seismicisolation
@Seismicisolation
220
Figure 14. Time history of roof accelerations
Figure 12. Distributions of peak drift ratio in X-dir.
7.2 Damper responses
The maximum peak floor accelerations are examined in Fig.
In addition to the structural global responses, the peak force
13. The “as-built” case has a peak floor acceleration of 0.85g
demands of each retrofit scheme are examined. The
at roof level. FVDs are able to reduce the peak floor
hysteresis loops of one device, located in Y-direction at the
accelerations by about 30% throughout the stories, and bring
3rd floor, subjected to one ground motion are plotted for
down the peak value at roof to 0.69g. The reductions benefit
different schemes; see Fig. 15. Under a same excitation,
from the additional damping effect provided by FVDs, and
three devices have different behaviors. The FVD exhibits
the additional stiffening effect is not significant. On the other
pure viscous properties, as shown by the elliptical shape of
hand, the case installing VWDs provides limited control over
hysteresis loop. The VWD has steel tanks at the exteriors of
the peak floor accelerations: the reduction is less than 10%
the viscous material, thus having an increased load-resistant
over all the story levels, and is essential zero at roof level.
capacity. However, a BRB has a totally different energy
The third case that uses BRBs is demonstrated to have the
dissipation mechanism compared to a FVD or a VWD. A
worst control effect, where the floor accelerations are
BRB dissipates energy through the yielding of braces, and a
increased at a majority of floor levels, and the peak roof
typical hysteresis loop is represented by a bilinear curve. For
acceleration is increased to 0.96g. This counter productivity
the damper selected for investigation, the FVD dissipates
of BRBs to control floor accelerations is mainly attributed to
most input energy despite that all three devices are designed
their displacement-dependent characteristics, which would
to have a similar energy dissipation capacity. All different
increase the force demands and accelerations at each floor.
devices show a similar deformation level.
Similarly, the roof acceleration time history during free
Meanwhile, the maximum damper force demands are shown
vibration phase in Fig. 14 shows that only FVDs could
for all schemes in Fig. 16. Fairly large force demands are
contribute to a more rapid decay of vibrations among three
observed for all cases, ranging from 1200 kips to 2300 kips.
cases under investigation.
It should be noted that FVD scheme, the most effective to
suppress the peak deformations and peak floor accelerations,
turns to have the smallest peak force demands among the
three.

Figure 13. Distributions of peak floor accelerations

@Seismicisolation
@Seismicisolation
221
(a). FVD

Figure 16. Distributions of peak force of each kind of device

7.3 Column axial force status

The evaluation of the case study building reveals that


vulnerable columns have posed a great danger to the seismic
integrity of the building. In the “Stage-1” retrofit, the brittle
columns splices were fixed and thus the concerns of brittle
column splice rupture/failure were eliminated. Nevertheless,
on the compression side, the columns might be overloaded
axially. Under a combined axial forces and bending moments,
these columns are very sensitive to yielding, threatening the
(b). VWD overall stability of the building. The introduction of
supplemental energy dissipation devices could bring down
the drift ratios demands, but the total force demands in the
base columns would remain the same if the seismic input
does not change. In the case having VWDs or BRBs, the
seismic demands are increased due to their additional
stiffening effect. As such, the axial force demands on base
columns would increase, exacerbating the column conditions.

To evaluate the influence of supplemental devices on the


columns, the envelopes of axial force demand-to-capacity
(D/C) ratio are examined for one group of exterior corner
columns. The selected columns fall into Group 1, identified
in Fig. 17, which have built-up W-sections. These corner
(c). BRB columns are usually most heavily loaded under selected
Figure 15. Hysteresis loops of one damper/BRB in story 3, Y- seismic excitations. The “Stage-1” methods have
direction. strengthened the column splices, the columns tension
capacities are thus estimated based on Equation 9-8 of ASCE
41-13, i.e., Pt = Ag Fy, where Ag is the gross section of
columns, and Fy is the expected yield strength of the material.
On the compression side, the buckling of columns is
considered and the lower bound compression capacities are
calculated: Pc = Ag  Fcr, where Fcr is the material strength
considering global buckling.

The positive sign is for tension while the negative sign


signifies compression in Fig. 18. The green line is the

@Seismicisolation
@Seismicisolation
222
compression demands due to gravity force, which consumes 08. Damage and Loss Analysis
about 30% of the column compression capacities. For the
“as-built” case, the peak D/C exceeds 1.0 at floor 6-7, and The damage and loss analysis was conducted using the
there are more than half of stories having peak D/C ratios software developed by FEMA: Performance Assessment
larger than 0.5. At these levels, ASCE 41 indicates that the Calculation Tool (PACT). The PACT performs the
members should be treated to be force controlled and remain probabilistic loss calculations in the framework of
elastic. The high D/C ratios at most floor indicate a Performance Based Earthquake Engineering (PBEE). The
significant reduction of column bending capacities, which repair cost and repair time of each realization were estimated
would likely contribute to the weak column, strong beam from fragility curves of structural and non-structural
behavior observed in the results. On the other hand, tension components, and consequence functions of damaged
rupture/failure is typically not a concern with all the brittle components. Four engineering demand parameters were used
splices fixed. to predict the damage states of different components,
including the peak story drift ratios, peak floor accelerations,
For the case with FVDs, the peak D/C rations are reduced peak floor velocities and maximum residual drift ratios.
slightly at several floors on tension, though there are no Among these, the first three parameters were results from
significant reductions of compression forces. Nevertheless, nonlinear response history analyses, while the residual drift
neither VWDs nor BRBs are able to alleviate the high column ratios were estimated based on an empirical relation
axial forces. The axial D/C ratios at most floors are increased suggested by FEMA P-58 (FEMA 2012a).
instead, and widespread column failures are more likely in
both cases. Other strategies to upgrade the column capacities, The probability of the building having irreparable residual
such as filling concrete in the built-up section columns, or drifts and the probability of unsafe tagging at BSE-2E event
adding corner columns could be explored. for the “as-built” building and three fully retrofitted buildings
are summarized in Table 1. The “as-built” one is expected to
have very large residual drift ratios at BSE-2E events, making
repair work unsafe and unrealistic. This could be seen from
the high chances of irreparability and high probability of
unsafe tagging of the “as-built” case. It is most likely that a
complete tearing down and reconstruction are necessary. As
a comparison, the building inserting FVDs successfully
brings down the residual drifts, and it has only 0.6% chance
of being irreparable. A 26.9% of unsafe tagging is estimated,
which is mainly resulted from failure of Pre-Northridge
beam-to-column connections and prefabricated steel stairs.
Consistent with what have been observed from structural
analysis results, the other two retrofit methods by using either
VWDs or BRBs still exhibit large residual drifts, and are less
Figure 17. Column group designations effective to reduce the chance of tearing down the building,
nor the chance of unsafe tagging at BSE-2E.

To better assess each retrofit scheme, the cumulative


distribution curves of repair loss ratio are presented and
compared; see Fig. 19. The repair loss ratio is defined as the
repair cost divided by the building’s replacement cost
according to FEMA P-58 (2012a). The replacement cost of
this building is the reconstruction fees, and is estimated to be
$475 million based on recent market values of similar
buildings in San Francisco area (Kidder Mathews 2015).
From the cumulative distribution curves, the median values
and 90 percentile values are extracted and listed in Fig. 20.
Both the “as-built” case and the VWDs case have a median
loss ratio equal to 1.0, i.e., the building needs full
replacement after a BSE-2E event for these cases. The
retrofit scheme with BRBs brings down the median loss ratio
Figure 18. Distributions of column axial D/C ratios to about 0.084, but still hits 1.0 when a larger confidence

10

@Seismicisolation
@Seismicisolation
223
level (90 percentile) is assessed. On the contrary, the retrofit
scheme using FVDs avoids large economic losses, and
reduces the median repair loss ratios to 0.047, and 90
percentile value to 0.071.

The overall cost-benefit of each retrofit scheme needs to


include the costs associated with purchasing and
implementing supplemental energy dissipation devices in the
building. The initial costs of purchasing various
supplemental energy dissipation devices from manufactures
are estimated in Table 2. These costs are a rough estimation
based on available online data and consultation from
experienced engineers, since real data is usually confidential
and not easy to approach. BRB is the cheapest, and VWD is
the most expensive device. Additional installation fees such
as construction cost to strengthen existing structural
elements/connections, transportation cost of heavy devices to Figure 19. Cumulative distribution curve of loss ratio
higher floors, and cleaning fees to save space for
supplemental devices need to be fully accounted.
Nevertheless, it would be rather difficult to estimate this
budget considering large uncertainty of above mentioned
items. As such, a simple calibration is proposed by using an
amplification factor, which is multiplied by the initial device
costs to represent the total investment of each retrofit scheme.
An amplification factor of 10 is used considering the size of
the case study building, its importance and quantity of
devices. The cost efficiency of different retrofit methods at
BSE-2E event are compared in Table 3. The FVDs retrofit
method is considered to be the most cost-efficient to upgrade
the case study building since it has more than 50% chance to
save $452 M repair cost, and more than 90% chance to save
$441 M after a BSE-2E event. Meanwhile, the initial
investment of $84 M is relatively small compared with its
potential savings. The case using BRBs ranks the second Figure 20. Median and 90 percentile loss ratio
since it has the minimum initial investment of $17 M, and has
more than 50% chance of saving $435 M repair cost at a
BSE-2E event. Nevertheless, no benefits are expected if a Table 2. Estimated initial costs of various devices
higher confidence level of 90% is to be achieved. Lastly, Scheme Initial Cost ($M)
there is more than 50% chance that VWDs would not save
repair cost, and its initial investment is the largest. Thus FVDs 6.4
VWD is the least cost-effective method. VWDs 8.4
BRBs 1.7
Table 1. Loss estimates
Scenario Probability of Probability of unsafe
irreparability tagging Table 3. Cost benefit comparison
As-built 94.3% 98.3% Scheme Investment Benefit from savings ($M)
($M) Median 90 percentiles
FVDs 0.6% 26.9%
VWDs 66.0% 70.6% As-built 0 0 0
BRBs 45.5% 60.6% FVDs 64 452 441
VWDs 84 0 0
BRBs 17 435 0

11

@Seismicisolation
@Seismicisolation
224
09. Conclusions beams fail. The displacement-dependent BRBs are acting in-
phase with structural displacements, and increase the force
A representative Pre-Northridge high-rise steel moment demands to existing members. Besides, both VWDs and
resisting frame was selected for seismic performance BRBs provided additional stiffness, reducing the building’s
assessment. The evaluations were based on ASCE 41-13 fundamental period and increasing the seismic force demands.
procedures FEMA 351 and FEMA P-58, and identified
several major structural vulnerabilities of the case study In addition, the behaviors of dampers or BRBs are checked.
building. As such, possible retrofit methods as well as their The results indicate that fairly large devices are required in all
cost-effectiveness were explored. A “two-stage” retrofit plan schemes, while the sizes of FVDs needed are anticipated to
was proposed for the case study building. In “Stage-1”, the be the smallest despite of their best control effects among the
brittle column splices were fixed everywhere, and the exterior three schemes. To relate the structural performance to the
heavy claddings were removed. However, analysis results economic losses, a damage and loss analysis is conducted
indicated that “Stage-1” method alone was not enough to following procedures outlined in FEMA P-58. The results
meet the retrofit goal of maintaining structural stability at a are consistent with the structural analyses, indicating that
BSE-2E event. Therefore, in “Stage-2”, several supplemental FVDs are the most effective to reduce the probability of
energy dissipation devices were used in combination with having irreparable residual drifts, probability of unsafe
“Stage-1” methods to further enhance the building’s seismic tagging, and led to much reduced economic losses after a
performance. The control effect, in particularly the cost- BSE-2E event from the “as-built” case. BRBs help improve
effectiveness of each retrofit method is investigated and the structural behavior a little, but are insufficient to provide
compared in this paper. a high confidence level of 90% to reduce repair cost. On the
other hand, VWDs provide little, if any, contributions to
Three devices are investigated in this paper: FVDs, VWDs reducing the economic consequences after a BSE-2E event
and BRBs. The design started by designing FVDs. Four due to a great number of beam failure and diminished damper
perimeter frames were selected to install these devices so that effect.
the interaction of occupants and interior components could be
minimized. The total effective damping ratios were Several design considerations exist for each scheme. One
estimated to achieve the target roof displacements at each common issue among three cases is the widespread
horizontal direction. A refined damper design was proposed vulnerable columns in the building. Even after the brittle
where dampers were installed only in locations with better splices were fixed, the columns are overloaded in
control effectiveness. These locations were the same for all compression and sensitive to yielding under combined axial
retrofitted schemes using different energy dissipation devices. force and bending. This poses great threat to the seismic
In addition, the mechanical properties of three devices were integrity of the building, and additional methods to upgrade
selected based on the assumption of equal energy dissipation. columns should be explored.

The structural global responses, devices behaviors and In summary, among three energy dissipation devices
column axial force status are presented. The results presented investigated, FVDs have the least interaction with structural
are the maximum values from three nonlinear response members, and are able to introduce additional damping
history analyses at BSE-2E. The global responses show that without significantly increasing the structural demands on the
the FVDs are the most effective to bring down the drift vulnerable columns and beams. Therefore, they are viewed
concentrations at floor level 2 to 10, and result in a more as the most promising solution to improve the structural
uniform distribution of the peak deformations. The peak drift behavior and reduce the economic losses of a Pre-Northridge
ratio after installation of FVDs is less than 1.5%, which could high-rise steel moment resisting frame.
essentially eliminate the beam-to-column connections failure
at BSE-2E events. FVDs are also shown to be the most Acknowledgement
efficient to suppress the peak floor accelerations and
contribute to a more rapid decay of the structural vibrations. This paper is supported by Pacific Earthquake Engineering
For other two retrofitted cases using VWDs or BRBs, unique Research center (PEER) as part of its Tall Building Initiative
problems are found and neither of them is able to provide and Next-Generation Attenuation Relationship programs.
effective structural control to the building under seismic Special thanks to Dr. Jiun-Wei Lai and Dr. Matthew
excitations, and thus unable to meet the retrofit goal. Schoettler who dedicated to set up the OpenSees model, as
Specifically, the introduction of a VWD in the middle of a well as the assistance from Dr. Frank McKenna, Dr. Andreas
beam having Pre-Northridge connection details would cause Schellenberg and Prof. Dimitrois Lignos to refine the
an earlier fracture of beams, and the control effect of VWDs numerical model. The authors would also like to express
would be significantly diminished once a large number of great gratitude to Prof. Kazuhiko Kasai of Tokyo Institute of

12

@Seismicisolation
@Seismicisolation
225
Technology, Dr. Kit Miyamoto and Dr. Amir Gilani of No. 2015/14, Pacific Earthquake Engineering Research
Miyamoto International, Dr. Amarnath Kasalanati of Center, University of California, Berkeley, CA.
Dynamic Isolation Systems, Jim Malley of Degenkolb
Engineers and Rob Smith of Arup for sharing their valuable Lee D. and Taylor D. P. (2001). Viscous damper
expertise and advice. development and future trends, J. Struct. Des. Tall Buil., 10,
311-320.
Reference
Lobo, R.F., Bracci, J.M., Shen, K.L. et al., Reinhorn, A.M.,
AISC (2010). Specification for Structural Steel Buildings, and Soong, T.T. (1993). Inelastic response of reinforced
American Institute of Steel Construction, Chicago, IL. concrete structures with viscoelastic braces, Report No.
NCEER-93-0006, National Center for Earthquake
ASCE (2013) Seismic Evaluation and Retrofit of Existing Engineering Research, Buffalo, State University of New
Buildings, American Society of Civil Engineers, ASCE/SEI York at Buffalo, Buffalo, N.Y.
41-13, Reston, VA.
Makris N., and Constantinou M.C. (1990). Viscous dampers:
Constantinou M. C., Symans M. D. (1992). Experimental and testing, modeling and application in vibration and seismic
analytical investigation of seismic response of structures with isolation, Report No. NCEER-90-0028, National Center for
supplemental fluid viscous dampers, NCEER-92-0032, Earthquake Engineering Research, State University of New
National Center for Earthquake Engineering Research, York at Buffalo, Buffalo, NY.
Buffalo, NY.
Mazzoni, S., Mckenna, F., Scott, M.H., and Fenves, G.L.
FEMA (2000). Recommended Seismic Evaluation and (2009). Open system for earthquake engineering simulation:
Upgrade Criteria for Existing Welded Steel Moment-Frame User command-language manual, Pacific Earthquake
Buildings, Federal Emergency Management Agency, FEMA Engineering Research Center, University of California,
351 report, Washington, D.C. Berkeley, OpenSees version 2.0 users’ manual, retrieved
from
FEMA. (2012a): Seismic Performance Assessment of http://opensees.berkeley.edu/OpenSees/manuals/usermanual/,
Buildings, Volume 1 – Methodology. Federal Emergency August, 2016.
Management Agency, FEMA P-58-1 report, Washington,
D.C. McKenna, F., Scott, M., and Fenves, G. (2010). Nonlinear
finite-element analysis software architecture using object
FEMA (2012b). Seismic Performance Assessment of composition, J. Comput. Civil Eng., 24(1): 95-107.
Buildings, Volume 2 – Implementation Guide, Federal
Emergency Management Agency, FEMA P-58-2 report, Newell J., Love J., Sinclair M., Chen Y-N., and Kasalanati A.
Washington, D.C. (2011). Seismic design of a 15-story hospital using viscous
wall dampers, Proceedings of Structural Congress, Las
FEMA (2012c). Seismic Performance Assessment of Vegas, Nevada, U.S.
Buildings, Volume 3 – Supporting Electronic Materials and
Background Documentation, Federal Emergency Reinhorn A. M., Li C., Constantinou M. C. (1995):
Management Agency, FEMA P-58-3 report, Washington, Experimental and analytical investigation of seismic retrofit
D.C. of structures with supplemental damping, part I: fluid viscous
damping devices. NCEER-95-0001, National Center for
Fu Y., Kasai K. (1998). Comparative study of frames using Earthquake Engineering Research, State University of New
viscoelastic and viscous dampers, Structural Engineering, York at Buffalo, Buffalo, NY.
124:513-552.
Rezaeian S., Bozorgnia Y., Idriss I.M., Campbell K.W.,
Kidder Mathews (2015). San Francisco office real estate Abrahamson N.A., Silva W.J. (2012). Spectral damping scale
market review 3 rd quarter 2015, retrieved March 2016, from factors for shallow crustal earthquakes in active tectonic
http://www.kiddermathews.com/downloads/research/office- regions, PEER Report No. 2012/01, Pacific Earthquake
market-research-san-francisco-2015-3q.pdf. Engineering Research Center, University of California,
Berkeley, CA.
Lai, J.-W., Wang, S. Schoettler, M. and Mahin S. (2015). Soong T. T., Spencer B. F. (2002). Supplemental energy
Seismic performance assessment of a tall building having dissipation: state-of-the-art and state-of-the-practice. Eng.
Pre-Northridge moment-resisting connections, PEER Report Struct., 24, 243-259.

13

@Seismicisolation
@Seismicisolation
226
Takewaki I. and Yoshitomi S. (1998). Effects of support
stiffness on optimal damper placement for a planar building
frame, J. Struct. Des. Tall Buil., 7: 323-336.

Taylor, D.P., Constantinou, M.C. (1998). Development and


testing of an improved fluid damper configuration for
structures having high rigidity, Proceedings of the 69th
Shock and Vibration Symposium, St. Paul, MN.

14

@Seismicisolation
@Seismicisolation
227
CASE STUDY: 5

2017 SEAOC CONVENTION PROCEEDINGS

Practical Implementation of ASCE-41 and NLRHA Procedures


for the Design of the LLUMC Replacement Hospital

Gregory Nielsen PE, Simon Rees SE, Branden Dong PE, Kermin Chok SE,
Eaman Fatemi, Atila Zekioglu SE
Arup
Los Angeles, CA

Abstract performance-based seismic design process for this project,


including management of large data sets, nonlinear response
The Loma Linda University Medical Center Campus history driven design, and navigating the practical
Transformation Project (LLUMC CTP) is a new 17 story base- implementation of ASCE 41-06 criteria and CBC 2013
isolated 1,000,000 square foot replacement acute care hospital OSHPD amendments.
located less than 1 km from the San Jacinto Fault. The seismic
design and analysis of the structure used LS-DYNA to
efficiently perform nonlinear response history analysis
(NLRHA) with 110 individual ground motion analyses
incorporating DE, MCE, upper bound, lower bound, and
varying ground motion direction. Implementation of triple
pendulum isolators, fluid viscous dampers, buckling restrained
braces, and SidePlate moment frames in LS-DYNA will be
described. As required for OSHPD-1 facilities, the NLRHA
results demonstrated Immediate Occupancy performance at
DE and Life Safety performance at MCE using element
backbone curves and acceptance criteria from ASCE 41 as
amended by the California Building Code. Inconsistencies in
element acceptance criteria for combined lateral systems and
other code implementation challenges will be discussed. A
cloud computing and database framework, using Penguin-on-
Demand and Amazon Web Services, was developed to manage
the 8 terabytes of data generated from each set of 110 ground
motion analyses performed on each design iteration.
Automated processes enabled the team to reduce the time
between design iterations to 2 weeks for the complete suite of
NLRHA, post-processing, report generation, and design
optimization. The team’s approach to analysis data Figure 1: Structural Revit model illustrating the LLUMC CTP
management, design optimization procedures based on building.
NLRHA results, automated post-processing, and automated
report generation will be detailed. Basis of Design

Introduction The hospital building is located at a highly seismic site only 1


km from the San Jacinto Fault and 5 km from the San Andreas
The Loma Linda University Medical Center Campus Fault. A site-specific seismic hazard assessment was
Transformation Project (LLUMC CTP) is a new base isolated performed by GeoPentech and the resulting Design
1,000,000 square foot replacement hospital in Loma Linda, Earthquake and Maximum Considered Earthquake spectra are
CA. The project will bring the LLUMC campus in compliance shown in Figure 2. Due to the close proximity of the site to the
with the California Hospital Seismic Safety Act and is an causative faults, the design ground motions developed for the
OSHPD-1 acute care facility providing Level 3 Trauma project include near-fault source effects and pulse
services to San Bernardino County. An image of the structural characteristics. In addition, a vertical response spectrum was
model is shown in Figure 1. This paper describes Arup’s also produced for the site which far exceeded the code

@Seismicisolation
@Seismicisolation
228
2017 SEAOC CONVENTION PROCEEDINGS

minimum vertical seismic component (0.2SDS) as shown in base BRBF solution was rejected due to incompatibility
Figure 2. Arup’s approach to mitigating this severe vertical between the required number of brace lines and the functional
seismic component through the use of an innovative vertical program of the hospital.
isolation system (VIS) will be detailed in a future paper. A
suite of 11 tri-directional ground motions was used to be The final selected structural design uses a base isolation
consistent with the provisions of ASCE 7-16 and forms the system comprised of 126 triple friction pendulum bearings
basis for the nonlinear response history analysis (NLRHA) with +/-42” displacement capacity manufactured by
procedure used for the structural design. These were originally Earthquake Protection Systems and 104 fluid viscous dampers
developed at the MCE level and scaled by two-thirds for the with 800 kip MCE capacity manufactured by Taylor Devices.
DE level. Ultimately 110 individual ground motions were The pendulum isolators have an effective period of 4.5 seconds
required for each design iteration: 11 ground motions, 2 and the dampers have a velocity exponent of 0.7. The total
orthogonal directions, upper bound (UB) and lower bound equivalent system damping coefficient is 50% of critical
(LB) properties, DE and MCE, and an additional 45 degree damping. High damping using supplemental dampers was
oriented UB and LB analysis at MCE. selected to control the overall building displacements and
reduce reliance on the friction pendulum system for system
damping, which is affected by the changing vertical load due
to the high site-specific vertical ground motion component.
Controlling the isolator displacements to 42” instead of 84”
without supplemental damping resulted in an optimal cost
solution by controlling the isolator and damper component
costs, the costs associated with stability framing above and
below the isolators, and the costs associated with expansion
joint covers and flexible service connections.

Figure 2: Horizontal and Vertical MCEr spectra for LLUMC


Site

Arup considered a wide variety of structural steel lateral


system configurations before concluding that the optimal
structural system was a combination of base isolation with
BRB’s and SidePlate moment frames above. The choice of
lateral system was driven by the high seismicity and pulse-like
nature of the site ground motions at periods greater than 1
second (SD1 equals 1.54g and is greater than SDS) and the
overall final height and relative slenderness of the hospital
towers. At 17 stories and 230 feet tall the LLUMC CTP will
be the tallest hospital in California. Initial evaluations into a Figure 3: Representative BRBF elevation
fixed-base biaxial moment frame design proved impossible to
achieve the required 1.25% interstory drift even when using Stabilizing the top of the isolators is a moment connected grid
every column line in the building as a frame and the strongest of 60” deep steel plate girders designed for Immediate
and stiffest SidePlate moment frame beam-column Occupancy performance at MCE. The majority of these
combinations permissible by AISC 358-10. Similarly, a fixed

@Seismicisolation
@Seismicisolation
229
2017 SEAOC CONVENTION PROCEEDINGS

moment connections are bolted connections to control welding substantially less nonlinearity in the model and hence
distortion in the field during steel erection. A SidePlate special substantially less analysis time and data generation. For
moment frame system is used in the East-West (tower long LLUMC the seismic demands were much higher and
axis) direction and a BRBF system is used in the North-South optimization of the NLRHA performance was required to
(tower short axis) direction. The frames are designed to widen reduce the overall structural tonnage. NLRHA optimization
at the base in order to reduce the overall uplift demands on the could not be practically done without first creating an
base isolation system to manageable levels. A representative automated workflow which could generate, analyze, and post-
braced frame elevation is shown in Figure 3. The design uses process all of the required analyses into a reduced form
ASTM A913 Grade 65 steel sections for columns and selected suitable for design while also minimizing the feedback time to
BRBF beams in order to achieve the required IO performance fit within the aggressive project schedule. A single design
at MCE for these “overstrength” category elements. A full list iteration of 110 ground motions resulted in over 6 TB of data
of the performance requirements of various elements of the generated and over 10 design iterations were conducted
structure are listed in Table 1. While the average result was through the course of the project from Design Development to
required to meet the criteria of Table 1, no ground motion was final Permit. From the outset, the goal was to turn the time
allowed to result in elements exceeding the Collapse intensive NLRHA into a practical tool in the designer’s
Prevention (CP) limit. toolbox, similar to response spectrum analysis, so that from a
designer’s view NLRHA was not used as a final performance
Table 1: Performance requirements for NLRHA verification but rather an integral design process.
Element Performance at Performance at
DE MCE The team chose to use LS-DYNA as the analysis engine for the
FPT LRFD Design Expected Strength project. This choice was for the following reasons:
Isolator Design
FV Damper LRFD Design Expected Strength • Excellent model stability under 3d ground motions
Design with 1.5 FOS
Mat LRFD Design, Expected Strength • Reduced analysis time using explicit time domain
Foundation Settlement < 1.5” Design, Settlement < solver versus other software
6”
Level A IO, 1 șy IO, 1 șy • Staff familiarity
Isolator
Framing • Ability to use with cloud computing services
SidePlate IO, 0.25 șy IO, 0.25 șy
Columns There was a considerable learning curve in introducing LS-
SidePlate IO, 0.02 radians LS, 0.03 radians DYNA to OSHPD plan review staff. The design utilized the
Beams Collaborative Plan Review (CPR) process which allowed for
BRBF IO, 0.25 șy IO, 0.25 șy monthly meetings between the design team and the review
Columns team during the 18 month review duration which was essential
BRBF IO, 0.25 șy IO, 0.25 șy in helping OSHPD become confident in the nonstandard
Beams analysis tools. A bounding analysis study was performed in
BRBF IO, 3 ǻy LS, 10 ǻy order to select the most demanding set of bounding parameters
Braces for the soil behavior, isolator behavior, damping behavior, and
Drag LRFD Design Expected Strength BRB strengths. Figure 4 shows the result of the bounding
Connections Design analysis on a 2-dimensional frame which indicated that the
Diaphragms LRFD Design Expected Strength least favorable sets of bounding values corresponded to Lower
Design Bound of all parameters and the Upper Bound of all
parameters. The Lower Bound analysis controlled the isolator
NLRHA Workflow displacements and damper velocities while the Upper Bound
analysis controlled the superstructure drifts, frame demands,
The major analytical challenge facing the design team was and floor accelerations. Selecting a single Lower Bound and a
creating an analysis framework that could handle the large sets single Upper Bound analysis reduced the possible
of data generated by the NLRHA models which were larger permutations considerably and was essential in keeping the
than any Arup had previously had to deal with. Arup’s prior analysis set to a manageable 110 ground motions.
work on the base isolated San Francisco General Hospital
utilized an essentially elastic design on top of isolation with

@Seismicisolation
@Seismicisolation
230
2017 SEAOC CONVENTION PROCEEDINGS

SidePlate *MAT_HYSTERETIC_BEAM elements


Beam calibrated to experimental SidePlate moment
Elements frame cyclic test results.
Other *MAT_HYSTERETIC_BEAM elements
Moment calibrated to ASCE 41-06 backbone curves.
Frame
Elements
BRB *MAT_HYSTERETIC_BEAM elements
braces calibrated to experimental cyclic test results
of Nippon Steel BRBs.
Soil *MAT_SPRING_GENERAL_NONLINEAR
Springs elements calibrated to ASCE 41-06 backbone
curves.

Figure 4: Results of 2d frame bounding analysis (run 1 is


with all upper bound properties, run 16 is with all lower
bound properties)

In addition, a number of verification analyses were performed


prior to analyzing the full 3d model to ensure that the modeling
methodology used in LS-DYNA appropriately captured the Figure 5: Illustration of the NLRHA Workflow
essential physical behavior and met OSHPD’s review
requirements. These elements and their LS-DYNA materials The NLRHA workflow is illustrated in Figure 5 including the
are tabulated in Table 2. various time durations for each activity. The framework for
pre-processing and visualization utilizes a MySQL database
Table 2: List of LS-DYNA materials and application with a graphical front end using Rhino and Grasshopper. The
Element LS-DYNA material model and calibration analysis uses cloud analysis nodes from Penguin-On-Demand
FPT Three single pendulum (POD) and a small number of post-processing operations such
Isolator *MAT_SEISMIC_ISOLATOR elements in as floor drifts are performed on the analysis nodes. After the
series. An additional analyses are complete, analysts can view the results remotely
*MAT_GENERAL_NONLINEAR_1DOF_ to check for any bugs or spurious results. After this quality
DISCRETE_BEAM element was added in control process, the resulting data is then transferred to
parallel to model the failure behavior of the Amazon Web Services (AWS) Redshift for the remaining bulk
isolator rim at maximum displacement. The of the post-processing and images of the visualized data are
performance of the combined element was captured in report-ready format. The report is then compiled
compared to MCEER Technical Report # 08- back on the local server where the design team can assess the
0018 for 3d movement and uplift behavior. performance of the model and make the necessary changes to
further optimize the behavior. The overall time from the
FVD *MAT_NONLINEAR_VISCOUS_DAMPE
implementation of a design change to having the full suite of
Damper R element in parallel with a
compiled results and engineering demand parameters (EDP’s)
*MAT_SPRING_GENERAL_NONLINEAR
was reduced to 1 week. This short feedback time allowed the
element to model the lock-up strength and
structural design to be finely tuned over multiple iterations and
stiffness of the viscous damper and
within the framework for design and review.
connections upon exceeding the stroke
capacity of the damper. This stiffness was
The process begins with the structural Revit model which
calibrated to physical test measurements and
defines the basic frame geometry and section sizes. These are
the strength was capped as 1.5x the MCE
pulled into the database and the structural analysis geometry is
damper capacity.
generated based on the various element types. Both the lateral

@Seismicisolation
@Seismicisolation
231
2017 SEAOC CONVENTION PROCEEDINGS

and vertical force resisting elements are imported, including nodes was limited to those processes which primarily served
the floor slabs. Rigid end offsets, element discretization, to reduce the data sets, for instance reducing a set of
element offsets, and particular material assignments are added displacement histories into drift histories or converting damper
using a set of Grasshopper components ensuring the generated relative displacements from the three axes of global
analysis geometry model is fully connected and aligns with all displacement to a single rotating along-damper axis. The
of the essential parameters validated in the LS-DYNA essential data was then transferred to AWS Redshift, a cloud
validation package signed off by OSHPD. As part of this database warehousing service, where it could be stored,
process the floor slabs are generated as cracked elastic 2d shell queried, and processed at a cheaper cost per core. Due to IT
elements with the required seismic mass and all gravity beam infrastructure limits in Arup’s LA office, the transfer from
members are included as elastic beam elements. While the POD to AWS was much faster than transferring the data from
baseline output is that all elements and nodal displacements are POD to Arup for internal processing. The AWS platform also
written by LS-DYNA at 20 Hz sampling rate, selected nodes allows for scalability and automatic backups for the large data
and elements, such as drift nodes, building separation nodes, sets. At the time that the LLUMC project was in analysis
and floor acceleration nodes, must be selected prior to analysis production, two other projects of similar scale and data
to output at a higher 200 Hz sampling rate. A 200 Hz sampling generation were being run out of the same office. Having a
rate is not used on all elements as this would generate more central off-site repository allowed the teams to focus on
data than is physically necessary to capture the relevant EDPs delivering the projects using a common set of tools rather than
and slow the entire process. Once all of these assignments are spending valuable time maintaining internal IT infrastructure.
complete, the Grasshopper components write the necessary
keyword cards for the complete LS-DYNA analysis models in Finally, the data was post-processed on AWS Redshift using
all 110 variants (ground motion, bounding properties, standard SQL queries to turn the 110 individual records into
directionality, and severity). This model preparation process is sets of EDP’s representing the average of the maximum
fully automated. An image of the LS-DYNA model is shown response from each ground motion over the ground motion
in Figure 6. suites. This final database of EDP’s was visualized in Rhino
on the same geometry wireframe model used to generate the
analysis models at the start. These final visualizations, along
with a hard drive with all of the post-processed data tables,
analysis run files, and keyword files, were submitted to
OSHPD for review. Intermediate iterations were used to
optimize the performance of the structure primarily using
conservation of energy approaches for resizing deformation-
controlled elements with target inelastic deformation limits.
The EDPs tracked are tabulated in Table 3. A sample of the
figures generated from the process is shown in Figure 7.

Table 3: List of EDPs tracked for NLRHA


Element Metadata Output
Force-Controlled Forces and inelastic
elements deformations
Deformation-Controlled Forces and inelastic
elements deformations
Figure 6: Image of the LS-DYNA analysis model. FPT Isolators Forces, displacements, and rim
inelastic displacements
The analysis models are uploaded to the Penguin-On-Demand FV Dampers Forces, stroke displacement,
(POD) cloud computing cluster which has sufficient scale to stroke velocity, lock up
allow all 110 models to be run in parallel. The longest ground element force and deformation
motion was able to run on a single compute node in 8 hours so Gravity moment Forces
the full suite of analyses could be sent off at the end of the connected elements
workday all runs would be finished by the following morning. Drag, collector, and Forces
The models are then checked for spurious results or early diaphragm truss
termination errors, which could be corrected during the elements
workday, and then some minor post-processing was Soil Springs Forces and inelastic
performed. Post-processing on the more expensive compute deformations
Building Separations Relative displacements

@Seismicisolation
@Seismicisolation
232
2017 SEAOC CONVENTION PROCEEDINGS

Interstory Drift Relative drifts at key points in Vertical Load Factor for NLRHA of Friction
building Pendulum Isolated buildings
Floor Accelerations Accelerations at key points in
building A unique feature of friction pendulum isolated buildings is that
Lateral Frames Story force time histories from the shear force in the bearings is directly proportional to the
cross-sections axial load applied to the bearing. This can lead to, and did on
Diaphragms Moment and shear envelopes the LLUMC project, a difference in interpretation regarding
derived from concurrent story the appropriate load combination to be applied to the overall
force time histories building when performing NLRHA. This issue has previously
been avoided in the designs for elastic structures on top of
isolation in which case the NLRHA was performed using an
unfactored 1.0 D combination roughly equivalent to the
seismic mass and then factoring the additional dead loads on
the superstructure elements using linear superposition.
However this is not possible for a combined NLRHA model
with nonlinearity in both the isolation system and the
superstructure lateral frame. A number of interpretations could
be considered ranging from 1.0 D + 0.25 L (ASCE 7-10
Section 16.2.3) to as much as 1.2 D + (0.2SMS W) + 0.5L or as
little as 0.9 D – (0.2SMS W) (ASCE 7-10 Section 17.8.2.5). On
LLUMC this maximum case could require the building to be
analyzed as though it was under 0.48g or 1.62g, meaning that
the same seismic force could act on the building but with 50%
or 160% of the lateral resistance in the isolators. A series of
analyses were performed to illustrate that the effect of the time
varying vertical ground motion on the lateral shear of the
isolation system was on the order of 5% variation in shear and
2% for isolator displacement. This result is primarily due to the
large amount of damping present in the isolation system
despite the very high vertical ground motions. The final load
combinations for the NLRHA lateral analysis were selected as
a compromise between the two extremes:

• Design Earthquake Cases: 1.0D + 0.2SDSD + 0.25L and


0.9D – 0.2SDSD

• Maximum Considered Earthquake Cases: 1.0D + 0.25L

The individual member design for force controlled elements


Figure 7: BRBF Frame elevation showing plastic strains from used additional load factors on the vertical loads per the
DE suite applicable sections of the code. It should be noted that the load
combination of 1.0D + 0.25L does not correspond to the
Practical Implementation Issues seismic weight of the building and therefore still results in
conservatism in lateral resistance in friction pendulum isolated
The design and plan review process of the LLUMC CTP buildings.
project illustrated a number of possible differences in
reasonable interpretation of code requirements as well as Gravity Column Deformation Compatibility for Steel
potential unintended consequences of certain code provisions. Frames
A number of these issues are summarized below and worth
further investigation by other researchers as part of future in- ASCE 7-10 Section 12.12.5 requires deformation
depth studies. compatibility to be assessed for non-seismic force-resisting
system elements due to the displacements at the design story
drift. For steel buildings designed in accordance with AISC
341-10, Section D3 applies. However since only the User Note

@Seismicisolation
@Seismicisolation
233
2017 SEAOC CONVENTION PROCEEDINGS

of Section D3 specifies that “flexible shear connections that delay onset of the flexural hinge. It should be noted that the
allow member end rotations per Section J1.2 of the magnitude of the forces in the BRBF beams was too great to
Specification should be considered to meet these make a true pin connection viable. This appears to be an
requirements.” The commentary further illustrates that only unintended consequence of the code since it does not appear
differential drift between floors leads to additional column logical to design a BRBF based on the performance of the
moment demands. Since this is not written into the body of the moment connected frame rather than rely on the energy
Seismic Provisions additional analysis was required to dissipating capacity of the BRB itself. It is unlikely that prior
illustrate that the columns did not have inelastic rotation projects utilizing deeper beam section and allowed to drift
demands when the drift histories of the building were applied. beyond 1.5% perform as required by this strict interpretation
To prove this a typical gravity column stack was modeled in of the code provisions.
LS-DYNA with ASCE 41-06 plastic hinge properties modeled
and FEMA 355 gravity shear tab inelastic hinges where the Floor Accelerations – Expected versus Actual
columns connected to gravity beams. The full suite of DE and
MCE ground motions were then applied in two orthogonal Seismic isolation is often used to reduce in-floor accelerations
directions and the plastic rotation was monitored. Only 1 for sensitive equipment. However, as Table 4 shows, the floor
ground motion at MCE resulted in hinging in the column. The acceleration results from LLUMC were in some cases higher
sizing design for the columns considered a 0.3% differential than the code prescriptive floor accelerations for nonstructural
story drift as an added moment in the LRFD column design component design in a fixed base building at the same site. The
and appears to be a reasonable design factor for this building relatively tall and flexible structure on top of the isolation
which was designed to 1.5% drift at DE. While the project plane likely leads to higher accelerations than may be
required that this be proven through rigorous methods it is a originally expected. Further, the design team found that LS-
logical result considering that the NLRHA did not indicate DYNA tends to predict higher accelerations (by up to 50%)
hinging in force controlled columns which are part of the than SAP2000. This will be investigated in the future and
lateral system and, in the case of the moment frames, are much compared to actual experimental shake table tests. It may also
stiffer than the gravity columns and hence will yield at a lower be the case that the prescriptive code floor accelerations for
rotation than the gravity columns. The use of Grade 65 steel nonstructural components in fixed base buildings may be
for the column sections also delays the onset of plastic hinging lower than they should be to ensure reliable performance. This
and is recommended. should also be investigated further and compared to actual
instrumented buildings and experimental tests.
BRBF Beam Plastic Rotation Limits Incompatible
with BRB Strain Limits Force-Controlled Frame Columns

An unintended consequence was discovered by virtue of A final lesson learned relates to force-controlled columns and
having modeled all of the lateral frame elements with plastic their treatment in the NLRHA model. Per the requirements of
hinges. ASCE 41-06 requires that beams of buckling restrained ASCE 41-06 these force-controlled columns should be
braced frames be treated similarly to the columns of the same modeled as elastic elements and the average of the maximum
frames and in some cases are force controlled. However, these forces developed in these elements from the suite of ground
low plastic rotation thresholds appear to be incompatible with motions should be compared to the capacity of the column.
the BRB strain limits of 3ǻy and 10ǻy for IO and LS However this could underestimate the response of the building
performance, respectively. The imposed deformation pattern if these elements were to exceed their capacity in any one
of the frame in the NLRHA that is required to yield the braces, ground motion. Due to this fact OSHPD requested that force
particularly for the W18 or W21 beam sections which are used controlled elements be modeled with axial load dependent
in typical designs, often results in inelastic behavior of the inelastic hinges. This then created the problem of
beams which cannot be controlled except by increasing the interpretation of results from the average of the suite of
stiffness and strength of the braces. This in turn increase the analyses. If any one of the motions resulted in plastic hinging,
column and brace forces required for capacity design. Due to no matter how slight, the average of the suite would indicate a
these reasons, the typical BRBF beam on the LLUMC project nonzero plastic rotation. This would then no longer meet the
is a stocky W14 column section since its shallow depth results requirement that there should be no yielding in force-
in delayed onset of a flexural hinge. Further the majority of the controlled elements. The only way to satisfy the requirement
BRBF beams were proportioned to have less than 0.5 P/Pcl so would be to design the column such that it did not hinge in any
as to keep them in the deformation controlled category of of the ground motions, which clearly represents overdesign
ASCE 41-06. Nonetheless this still required a plastic rotation relative to the averaging methodology used throughout ASCE
of 0.25 șy in accordance with ASCE 41-06. In a few select 41-06. This issue was finally concluded by allowing a plastic
instances the beams were even increased to Grade 65 to further rotation of up to 0.05șy in force-controlled elements. This also

@Seismicisolation
@Seismicisolation
234
2017 SEAOC CONVENTION PROCEEDINGS

highlights the need to remain consistent with the procedures of


a single standard without modification.

Conclusion

The LLUMC CTP project represents one of the first designs


by Arup utilizing NLRHA as an integral part of the design
development and optimization process. Key to unlocking this
potential was the development of an automated workflow for
assembling, analyzing, and post-processing complex and very
large structural analysis models in a short enough time
duration to respond to design changes. This required a
considerable investment in analysis, database, and
visualization tools but reduced the engineer time required by
an order of magnitude while increasing the throughput of
analysis models. The project generated over 50 TB of data
over 1000 individual analyses during the design and
optimization process. This optimization process was essential
in eliminating every excess ton of structural steel on the
project and mitigating the substantial budget impacts of an
extremely challenging site. This case study illustrates that
design and optimization using NLRHA is feasible and the
framework is set to apply this methodology to future projects
to increase performance reliability and reduce cost.

@Seismicisolation
@Seismicisolation
235
CASE STUDY: 6

@Seismicisolation
@Seismicisolation
236
@Seismicisolation
@Seismicisolation
237
@Seismicisolation
@Seismicisolation
238
@Seismicisolation
@Seismicisolation
239
@Seismicisolation
@Seismicisolation
240
@Seismicisolation
@Seismicisolation
241
@Seismicisolation
@Seismicisolation
242
@Seismicisolation
@Seismicisolation
243
@Seismicisolation
@Seismicisolation
244
@Seismicisolation
@Seismicisolation
245
CASE STUDY: 7

Integrated Design and Construction at the 250 West 55th Street Tower

Matt Jackson SE, PE1 and Berkay Baykal LEED AP2


1
Associate Principal, Arup, 77 Water Street, New York, NY 10005; e-mail:
[email protected]
2
Project Engineer, Turner Construction Company, 375 Hudson Street, New York, NY
10014; e-mail: [email protected]

ABSTRACT

The recently completed 40 story office tower at 250 west 55th street in Manhattan
demonstrates the best in innovative structural design, and use of 3-D coordination
tools for design and construction. This paper describes the integrated process that was
followed and some of the challenges that were met along the way, and will be of
interest to design professionals and others interested in integrated construction
processes.

The integrated 3-D process started with the use of Revit from the Schematic design
stage, and was followed through the design, procurement, and construction phases,
with all major sub contractors producing 3-D or 4-D models. These models were
carefully integrated by the general contractor, and enabled savings in schedule,
reduction of field conflicts, and reduced project risks.

The progress of the project was further complicated by a suspension of construction


for two years after completion of the foundations. This paper describes some of the
unusual steps taken to manage this process and allow for an accelerated schedule
upon restart of construction.

INTRODUCTION

250 West 55th st, is a steel framed, 600ft tall, 40 story office tower, located in
midtown Manhattan. The tower, developed by Boston Properties, and designed by
Architect Skidmore Owings and Merrill, contains approximately one million sq. ft,
and is slated for occupancy in early 2014.

The design of the project commenced in 2007, and it was decided from the beginning
of the schematic design phase, that BIM, and specifically Revit, would be used
throughout the design of the tower by the Architect, Structural, and MEP Engineers.
Although this is more common now, at the time this was very unusual, and this was
pursued with the goal of not only delivering a better integrated project, but also
developing a 3-D working process that could serve as a model for future projects.

@Seismicisolation
@Seismicisolation
246
Figure 1. Completed Tower © Kevin Chu/KCJP Figure 2. Structural Revit model

One aspect of particular focus for the coordination was the core. As is typical for a
New York City project, the core was a steel braced frame, and as such coordination of
gusset plates, with the ducts, pipes and conduits that must pass in and out of the core
is critical. This began for the typical office floors at the schematic design stage,
before the framing of the core was finalized, and included consideration of not only
the current needs of the building, but allowance for the addition of extra conduits in
the future.

Although clash detection tools exist within the software to facilitate this, this work
still requires an understanding of the critical coordination locations due to the
numerous other hits that an automated tool would produce on a model at this level of
detail. In addition it is necessary to bring forward the modeling of many items in

@Seismicisolation
@Seismicisolation
247
these critical areas, beyond the level of detail that would typical be required at this
stage.

Figure 3. Coordination of critical clearances in Navisworks

The use of the Revit model continued throughout the design into the procurement
phase with the model being shared with bidders, providing greater understanding of
the level of design and coordination. This is particularly important for fast-track
projects where early trades are bid before the design is completed, and this process
helped realize very consistent bids for the early steel package.

During the steel bid, the contract documents also required the steel detailer to provide
models and attend meetings early during the detailing phase, with the goal that the
design team would review the model, and thus expedite review of the shop
submittals, with minimal re-submittals required.

The Steel Contractor, Owen Steel, and their detailer, 4D Global Group, embraced
these ideas, and participated in several in person and on-line meetings as the model
was built and reviewed. This not only helped speed reviews, but the improved
interaction with the detailer, together with the BIM design process that was followed
enabled a substantial reduction in the number of RFI’s that were raised in the process.
In the 5 years that have passed since this phase was completed, the software available
has improved to the point where conventional shop drawings could be eliminated
entirely.

Due to the fast-track nature of the project, the detailing model, produced in Tekla,
was available to the design team, before the completion of Construction Documents
by the architect, and hence this fully detailed model was able to be overlaid with the
Architecture and MEP models and further clash checks carried out.

@Seismicisolation
@Seismicisolation
248
STRUCTURAL DESIGN CHALLENGES

A significant structural challenge arose when the wind tunnel testing was completed,
after the end of the schematic design stage. The testing revealed that the loading on
the building from the wind was less than that required by the code, and that which
had been assumed in the preliminary design. Although the loading was lower, the
accelerations that occupants would experience at the top of the tower were predicted
to be higher than desirable according to most guidelines.

Desirable

Figure 4 - Initial wind loads and accelerations

Options to reduce the accelerations to an acceptable level were studied, and this
included adding more steel to the lateral system in order to add stiffness, as well as
various options for adding damping to the structure. Because the loads were less than
originally designed for, this allowed for some steel weight to be removed, however
this would potentially increase accelerations further. If however a damping system
with sufficient damping was added, then the benefit of these reduced loads could be
taken, and overall steel tonnage reduced. The result of this was that damping was
demonstrated to be a much more cost effective solution than adding stiffness, as is
commonly the case for tall buildings.

Damping systems considered included tuned mass dampers, tuned liquid column
dampers, and a viscous damping system integrated into the structure of the building.
As the building included a mechanical room at the top of the tower, the intention was
to integrate the damping into this space. Both the tuned mass dampers and tuned
liquid column damping system would have required significant rearrangement of the

@Seismicisolation
@Seismicisolation
249
mechanical equipment to accommodate the damper, however a novel arrangement of
viscous dampers was developed that incorporated the dampers into a series of
outriggers connecting the core to the perimeter of the building.

By arranging the dampers into an outrigger configuration, and thus damping the
dominant flexural deformation rather than shear deformation, sufficient damping was
obtained with the use of just 7 dampers.

There are many challenges to designing a viscous damping system, in particular the
damping system must be analyzed as a part of the overall structure, rather than as a
separate bolt on system as can be done with a tuned mass damper.

The specification of the dampers for control of wind movements is quite different to
those used for seismic applications, where viscous dampers are more commonly used
in buildings. The dampers must operate at very small amplitudes to avoid building
movements being able to build up, and must also cycle constantly whenever the
winds reach a sufficient level. To meet these challenges, and allow for a long
maintenance free life, a damper with no conventional seals provided by Taylor
Devices, was selected.

Despite the high specification of the dampers, the overall cost of the system was less
than half the cost of a conventional tuned mass damper that would have achieved a
similar level of damping.

Figure 5 – Photo of damper and arrangement of dampers (indicated in yellow)

A more detailed description of the damping system and the challenges of integrating
this into the design is contained in the paper ‘Increasing Efficiency in tall buildings by
Damping’ (Jackson and Scott 2010).

@Seismicisolation
@Seismicisolation
250
DELAYED CONSTRUCTION

One of the most significant challenges encountered during the project was caused by
the economic crisis in 2008. Because of the changed market conditions the developer
chose to postpone construction until the market was more favorable. The team
quickly moved to put in place a plan to allow for an orderly demobilization and
efficient restart when the time was right. The foundation construction and steel
fabrication was already under way at this time, and so the team decided to complete
the structure up to the grade level to both stabilize the perimeter walls and allow the
site to be more easily waterproofed and protected. The fabrication continued for the
remainder of the steel, and this was stored in a yard in South Carolina, close to
fabricator Owen Steel. The team developed a plan to ensure that the site would be
safe and agreed a monitoring program with the department of buildings to make
certain the site would continue to be secure.

When time came to restart construction, the team was concerned about the corrosion
that had developed on the surfaces that had been prepared for slip-critical
connections; in particular, those that had been blast cleaned to achieve a Class B
surface. The RCSC bolting specification suggests that some corrosion for up to a year
should be acceptable, but no further data was available to specify exactly what level
of corrosion over what period of time would still perform in a satisfactory way.

Figure 6 – Steel storage yard

We decided to test a sample of representative connections from the stored pieces and
verify the coefficient of friction directly. The resulting tests showed that the stored

@Seismicisolation
@Seismicisolation
251
steel exceeded the required 0.5 coefficient of friction, and all of the samples showed
higher levels of friction than some freshly blast cleaned control samples, some of
which did not quite reach the required friction level.

CONSTRUCTION RESTART

Once the developer made the choice to restart construction, several unusual
opportunities and challenges arose. The availability of the already fabricated steel
meant that steel erection could start almost immediately, and the previously
completed and waterproofed cellar levels meant that some of the critical MEP spaces
were already available for installation, however none of the MEP subcontractors had
yet to start their work. This meant that engineering, coordination and fabrication for
MEP trades, was now on the critical path for construction, and any reduction in this
time would directly improve the schedule.

One of these installation areas, where construction could begin, was the switchgear
room, which was on the critical path of the construction schedule. Another challenge
during the cellar level coordination was the fact that because no electrical contractor
coordination was carried out in the first phase of construction, no conduit had been
embedded into the slab; this resulted in substantial amount of electrical conduits to be
accounted for in coordination.

A number of things helped in overcoming these MEP coordination challenges: First,


the building was already designed in Revit and the architect has turned over this
model to be used by contractors. Secondly, a structural steel fabrication model was
readily available in Tekla. By combining these models, an accurate background
model was established; and 3D MEP coordination started very quickly.

Navisworks was used for 3D coordination; and every trade used their own choice of
software specialized in their field. Within 20 sessions, coordination of the cellar
levels was completed. Thanks to BIM, subcontractors were able to go into fabrication
from the coordination model almost immediately. Typically, on similar projects,
subcontractors would submit shop drawings before and after coordination for
engineer’s review and approval. In this project, the MEP design engineer, Cosentini,
participated in the 3D coordination, and the final model was submitted for their
review and approval. As there was no need to wait for shop drawing review and
approval, fabrication started almost immediately after coordination.

Participation of engineer in the 3D coordination process also saved tremendous time


and reduced the number of RFIs and potential future issues. During the entire
coordination of 250 West 55th street, there were only 180 RFIs submitted for MEP
trades. Due to the unique nature of each project, it is not easy to compare this number
to a benchmark. However, it is safe to say that this was far fewer than the number of

@Seismicisolation
@Seismicisolation
252
RFIs that would have been submitted if the team did not utilize BIM and engage the
engineer during this process. Most of the RFIs were being addressed on the fly and in
person by the engineer. Virtual meetings were used when the engineer was not
available to attend all day meetings.

Figure 7 - C2 Level Coordination

With the help of strategies outlined above, MEP installation started in cellar levels
earlier than scheduled. One of the key areas, as mentioned before, was the switchgear
room. Early start of this area eventually resulted in early delivery of permanent
power, which was one of the major construction milestones scheduled for December
of 2012. Instead, this milestone was delivered in May of 2012. Needless to say,
getting permanent power earlier than scheduled benefited the overall construction
schedule.

As coordination on cellar levels was ongoing, coordination of the risers concurrently


progressed. Through quickly identifying any issues with structural openings and large
penetrations, such as requirements for modifications and the need for adding more
openings, it was possible to modify the steel as it was being erected. This as a result
eliminated the need for comeback fixes. Coordination of beam penetrations in
miscellaneous areas of the building was another focus area. Even though the design
engineer coordinated major penetrations and steel was fabricated accordingly, due to
the nature of tight coordination, additional beam penetrations were required. With

@Seismicisolation
@Seismicisolation
253
BIM it was possible to identify these penetrations early on and with great accuracy
allowing steel subcontractor to apply penetrations without any field measurements.

One of the most challenging areas of 250 West 55th Street project was the
construction of 39th floor mechanical room. The construction team has coordinated
the installation sequence of this area in great level of detail with the help of BIM.
Specifically, 4D-scheduling was used to identify underutilized and overcrowded
areas, leading to an efficient staging and sequence of installation. The team also
heavily relied on the 3D-model to identify necessary structural steel leave-outs.

Figure 8 - Roof Level Coordination

Two particular examples of obtaining tangible results from using BIM were boilers
and the generator. These were the largest and heaviest pieces of equipment on this
floor, originally intended to be set during steel erection on temporary posts and
dropped to their permanent positions after floor was poured. With the staging

@Seismicisolation
@Seismicisolation
254
coordination efforts outlined above, the construction team was able to continue to
pour concrete and set both of these equipment without the use of temporary posts.
This way, the 39th floor was delivered to MEP trades earlier and substantial schedule
savings were realized. Another important benefit of this strategy was the reduced risk
due to not double-handling this sensitive equipment.

Another example as a result of structural steel coordination on the 39th floor was the
issue with cooling tower dunnage steel. This steel was fabricated prior to the approval
of cooling towers due to the fabrication schedule and stalled nature of this project.
Early enough in the project the discrepancy between cooling towers and dunnage
steel was identified, allowing the team to make adjustments on cooling towers and
giving the structural engineer enough time to analyze and make sure that the offset
nature of the new layout was structurally acceptable.

Figure 9 - Construction at the roof level

Once the tower cranes were erected, and steel erection restarted, the steel progressed
very quickly with topping out achieved in approximately 9 months. With the MEP
trades able to follow closely behind through the expedited coordination and
accelerated schedules.

@Seismicisolation
@Seismicisolation
255
CONCLUSIONS

The design and construction of the 250 West 55th St tower utilized a tightly integrated
approach in order to maximize the benefit to the client, and minimize construction
issues and cost. This included extensive use of BIM during design and construction,
as well as a closely integrated team approach to dealing with project challenges.

Overall, the full building TCO was received in May of 2013, compared to the initial
scheduled date of September of 2013 when the project was restarted. From cost
perspective no construction contingency was used throughout the project. Although
there are many other factors in these great achievements, the carefully integrated
approach to design and construction using the best available BIM tools, was a key
factor.

REFERENCES

1. Jackson, M. and Scott, D. (2010) “Increasing Efficiency in Tall Buildings by


Damping” Structures Congress 2010: pp. 3132-3142

@Seismicisolation
@Seismicisolation
256
CASE STUDY: 8

71st Annual Structural Engineers Association of California (SEAOC) Convention Santa Barbara, California, September 2002

Design of an Essential Facility with Steel


Moment Frames and Viscous Dampers Using
2000 NEHRP
Kit Miyamoto, M.S., S.E., President & CEO
Marr Shaffer & Miyamoto, Inc.
West Sacramento, CA

Robert D. Hanson, PhD., P.E.


University of Michigan
Walnut Creek, CA
Abstract This paper presents an earthquake design procedure and
a case study of the Vacaville Police Headquarters. The
This new 2-story, 40,000ft2 police headquarters earthquake design “goal” of this essential facility is to
becomes the first building in the United States to apply provide an immediate occupancy performance for a
2000 NEHRP procedure to design an essential facility 475-year return seismic event. However, the project
with Fluid Viscous Dampers (FVDs). The structure is requirement is to keep the construction cost within
located in Vacaville, California, which is in a region of typical code conformed buildings. The combination of
high seismic activity and classified as zone 4 per 1997 Special Moment Resisting Frames (SMRF) and Fluid
Uniform Building Code. The lateral force resisting Viscous Dampers (FVDs) are used as the lateral force
system (LFRS) consists of special steel moment frames resistance system. This system as described by Gimmel,
with FVDs. In accordance with 2000 NEHRP, the Lindorfer, and Miyamoto, (2002) results in cost
LFRS is sized and designed with strength requirements efficiency and superior seismic performance. The 2000
of the code level force. FVDs are provided to control NEHRP (FEMA, 2000) guideline was used to design
displacement of the structure. This design philosophy the project, since it is considered to be a state-of-the-art
leads to a low frequency structure with low procedure for seismic damping devices. This project
acceleration. FVDs reduce the displacement level to becomes the first structure in the United States to use
less than 0.01 story drift ratio. Earthquake performance this advanced procedure.
and cost effectiveness are the primary concerns in
designing this building. Site specific response spectra Building Description
and time histories are synthesized for a 500-year and a
2,500-year return event. Performance Based Design The project is located in Vacaville, California, which is
using both linear and nonlinear time history analyses is within a region of many active faults and high seismic
conducted to ensure “immediate occupancy” activity. The structure is a 2-story, 40,000ft2 (3,716m2)
performance. A cost study shows that much of the steel framed structure. The roof is composed of metal
FVD’s cost is offset by reducing the weight of the deck and WF beams, and the floor is composed of 21/2-
LFRS while providing a far superior performance than inch (6.4cm) lightweight concrete over 3-inch (7.6cm)
the “code-compliant” structures. metal deck and steel composite beams. The exterior
finish is lightweight architectural finish over
nonstructural metal stud walls. See figure 1 for
Introduction architectural rendering. See figure 2 for the second floor
structural plan. A perimeter SMRF is provided along
Marr Shaffer & Miyamoto, Inc.
1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
257
the longitudinal direction. For the transverse direction, plan. A 24-inch (61.0cm) deep pad foundation is
one-bay SMRF is provided at each column line. The provided at WF columns.
location and quantity of SMRF is the same for the roof

Figure 1: Architectural Rendering (Courtesy of Indigo Architecture)

Light Weight Concrete


Over Metal Deck

Indicates Indicates
Composite Moment
WF Beam Frames

Figure 2: Second Floor Structural Plan

Seismic Risk The site soil consists of 65feet (19.8m) of alluvium


overlaying siltstone. The site is considered to be Sd soil
per 1997 UBC. The Vaca fault, which is not considered
Marr Shaffer & Miyamoto, Inc.
1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
258
to be active, is closest to the site at a distance of 0.2 km. Great Valley Fault. The 1889 Antioch (M 6.3)
The next closest faults are segments 4 and 5 of the earthquake is attributed to the Greenville fault (Singh,
Great Valley Seismic Source Zone located at distance 2002). The 1997 Uniform Building Code (ICBO, 1997)
of 6.6 and 9.8 km, respectively. These faults are ignores the near fault effects from blind thrust faults
considered to be blind thrust faults. The closest fault such as the Great Valley Source, therefore, the site
considered capable of surface rupture is the Green specific response spectra were created for this project
Valley-Concord fault located at 18 km from the site. (Singh, 2002). See figure 3 for a 475-year return and
The significant nearby earthquake was the 1892 figure 4 for a 2,500-year return response spectra.
Vacaville/Winters (M 6.5), which was attributed to the

Average
Fault Normal
1.400 Fault Parallel
Acceleration (g)

1.200
1.000
0.800
0.600
0.400
0.200
0.000
0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 4.000

Period (sec)

Figure 3: 475-year Return Response Spectra (Singh, 2002)


Average
Fault Normal
2.500 Fault Parallel
Acceleration (g)

2.000

1.500

1.000

0.500

0.000
0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 4.000

(Period) (sec)

Figure 4: 2,500-year Return Response Spectra (Singh, 2002)

Conventional Structural Design The structure was first designed as a conventional


SMRF to provide a benchmark for cost and seismic
performance comparison with a high-tech system. The

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
259
2000 NEHRP was used to design SMRF. The following discussion, only the longitudinal frame is considered.
are design parameters for the Equivalent Lateral Force Tributary weight of the roof is 380kip (1,690KN) and of
Procedure. the floor is 924kip (4,110KN). Table 1 shows the results
Seismic Use Group III I = 1.5 of the modal analysis.
SMS = 1.95g at 0.3sec. (site specific)
SML = 1.05g at 1.0sec. (site specific) Period (sec) Mass Participation %
SDS = 2/3 x 1.95g = 1.3g Mode 1 0.69 84
SDL = 2/3 x 1.05g = 0.7g Mode 2 0.27 16
Seismic Design Category D Table 1: Results of Modal Analysis
SMRF: R = 8, Cd = 5.5 (Conventional Design)
Cs = SRDS I = 0.24g
Nonlinear static pushover was conducted to gauge an
Cs = SDL I = 0.32g earthquake performance of this frame. Figure 6 shows
RTa
capacity/demand spectra with a site-specific 475-year
Ta = 0.4 return event. Please note that figure 6 is for a single
0.24g should be used for seismic shear. degree of freedom system.
The above value is compared with 1997 UBC. The following are results of the pushover for a 475-year
Ca = 0.44xNa = 0.44 return event. The results are converted to the multi
Cv = 0.64xNv = 0.64 degree of freedom system.
Near field factors are 1.0, since blind thrust faults are Maximum roof displacement = 5.6 inch
ignored by 1997 UBC. (14.2cm)
R = 8.5, I = 1.25 Base shear = 0.80g
2.5 CaI Effective period = 0.71sec.
V= R = 0.16g Effective damping = 8.4%
Cv I Max drift ratio = 0.016
V = RT = 0.24g Some yielding events were observed at the bottom of
0.16g should be used for seismic shear. This value is the first floor columns and second floor beams. The
lower than 2000 NEHRP value. It is affected by the drift ratio is reasonable, but the base shear of 0.8g may
magnitude of R, I, and near field factors. The 2000 cause nonstructural damage to the second floor
NEHRP allows 75% of seismic shear to be used for the equipment and roof HVAC units. This is the limitation
damped frame if the total effective damping is 14% or of the conventional design. This fairly strong SMRF
greater. Therefore, 0.75x0.24g = 0.18g. SMRF is provides near elastic response. However this system
designed for both strength and drift criteria using 0.18g also produces high roof and floor accelerations. For this
base shear. The 0.18g value is larger than the 0.16g ground motion, the high frequency system such as shear
value required by the 1997 UBC; therefore it will be walls and steel brace systems would produce an even
used to design this frame to compare with the damped higher acceleration and increase seismic demands on
frame described later. The drift criteria is the controlling nonstructural components. The base isolation may be an
criteria of the design rather than the strength criteria. ideal solution for this case, yet, the cost increase was
Allowable story drift ratio is 0.015 and computed not allowed by the project requirement.
maximum drift is multiplied by Cd .
I
Figure 5 shows the longitude frame elevation. For the
transverse direction, SMRFs and dampers are provided
to approximate equivalent stiffness and strength as the
longitudinal direction. Therefore, for the following

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
260
30'-0" (TYP) W21X50 @ ROOF
(TYP)
(9.14 m)

(4.88 m)
16'-0"
W24X68 @ FLOOR
(TYP)

(4.27 m)
14'-0"
W14X159 COLUMN PAD FOOTING 3'-0"x3'-0" (0.92 m x 0.92 m)
(TYP) (TYP) GRADE BEAM (TYP)

Figure 5: Longitudinal Frame Elevation

1.4

5% Damped Spectra
1.2 Demand Spectra
Capacity Spectra
Spectra Acceleration (g)

0.8

0.6

0.4

0.2

0
0 1 2 3 4 5 6 7 8

Spectra Displacement (in)


Figure 6: Capacity/Demand Spectra for Conventional Frame

Hi-tech Systems Design elevation. The difference from figure 5 is the “pinned”
foundation condition and roof beam sizes. See table 2
The structure was then redesigned using SMRF with for FVD properties.
FVDs per 2000 NEHRP. The base shear of 0.18g as Damping Constant ‘C’ Per a FVD
described above was used to resize the frame members. Unit
The 2000 NEHRP describes that the frame members are 1st Floor FVDs 60kip-sec2/in (105kN-sec2/cm)
sized with strength requirements of the code level 2nd Floor FVDs 30kip-sec 2/in (52.5kN-sec2/cm)
(0.18g), and FVDs are provided to control displacement Table 2: FVD Property
of the structure. See figure 7 for the new frame

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
261
The damping force is defined as
F = CV α
V = Velocity
α = 0.6 Nonlinear time history analysis showed that all SMRF
These damping properties were selected based on an elements remained elastic. The maximum roof
optimal displacement reduction and FVD force output. displacement is reduced by 55% from the conventional
Table 3 shows the results of the modal analysis. The design; the base shear is reduced by 65%; and the
results show that the predominant period shifted from maximum story drift ratio is reduced by 38%. The
0.69sec of the conventional SMRF to 1.2sec. This maximum FVD force per unit is 206kip (916kN) at the
frequency shift effectively brings the dynamic response first level. These results show that the structural damage
to a lower acceleration range in the site-specific is eliminated and nonstructural damage is significantly
response spectra. reduced by adding FVD. Figure 8 shows FVD force vs.
FVD displacement for one of the first floor FVD units.
Period (sec) Mass Participation (%) It shows the effect of the damping exponent 0.6. The
Mode 1 1.20 96.5 shape of the hysteresis loop is between the oval
Mode 2 0.08 03.5 (α = 1.0) and the rectangular (α < 0.1). Figure 9 shows
Table 3: Results of Modal Analysis FVD force vs. 2nd floor velocity for one of the first floor
units. It shows nonlinear response of FVD unit. Figure
The nonlinear computer model with discrete damping 10 shows FVD force at one of the first floor FVD units
elements were created using ETABS 7. Three sets of and first floor column bending moment of the first 20
time history ground motions compatible to a 475-year seconds of this 40-second record. It shows maximum
return event were synthesized by Singh (2002). FVD force is out-of-phase from maximum bending
Nonlinear time history analyses using step-by-step moment. Figure 11 shows the base shear of SMRF with
linear acceleration procedure were conducted. Table 4 FVD and the conventional SMRF for the first 20
compares the results of this analysis with the results of seconds. It shows a substantial reduction of the base
the push over analysis of conventional SMRF. shear. A linear time history analysis was conducted on
the elastic frame of the conventional SMRF. The elastic
frame was used since the push over results show near
Conventional
SMRF w/ FVD elastic response of the conventional frame. Results of
SMRF
linear time history and pushover analyses are slightly
Max. roof
2.5in (6.4cm) 5.6in (14.2cm) varied. Figure 12 shows the roof displacement of SMRF
displacement
with FVD and the conventional SMRF for the first 20
Max. base shear 0.29g 0.8g seconds. It shows a substantial reduction of the
Max. story drift displacement. Figure 13 shows energy balance. FVD
0.010 0.016
ratio energy dissipates the majority of the input energy.
Table 4: Performance Comparison for 475-year
Record

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
262
30'-0" (TYP) W18X40 @ ROOF W24X68 @ FLOOR
(9.14 m) (TYP) (TYP)

(4.88 m)
16'-0"
(4.27 m)
14'-0"
W14X159 COLUMN PAD FOOTING INDICATES FVD
(TYP) (TYP) (TYP)

Figure 7: Longitudinal Frame Elevation with FVDs

250

200

150

100

50
Forces (k)

-50

-100

-150

-200

-250
-1.5 -1 -0.5 0 0.5 1 1.5

Displacement (in)

Figure 8: FVD Force vs. FVD Displacement for 475-year Record

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
263
250

200

150

100

50
Force (k)

-50

-100

-150

-200

-250
-15 -10 -5 0 5 10 15

Velocity (in/sec)

Figure 9: FVD Force vs. 2nd Floor Velocity for 475-year Record

600

500

400
FVD Force
Moment (k-in) or Force (k)

Column Bending Moment


300

200

100

-100

-200

-300

-400

-500
0 2 4 6 8 10 12 14 16 18 20

Time (sec)
Figure 10: FVD Force vs. Column Moment for 475-year Record

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
264
1500

Conventional SMRF
1000
SMRF with FVD
Base Shear (k)

500

-500

-1000

-1500
0 2 4 6 8 10 12 14 16 18 20

Time (sec)

Figure 11: Base Shear of SMRF with FVD vs. Conventional SMRF for 475-year Record

Conventional SMRF
6

SMRF with FVD


Roof Displacement (in)

-2

-4

-6
0 2 4 6 8 10 12 14 16 18 20

Time (sec)
Figure 12: Conventional SMRF for a 475-year Record vs. Roof Displacement of SMRF with FVD

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
265
5000

4500

4000

3500

3000

2500
Input Kinetic
Potential Inherent Damp
FVD
2000

1500

1000

500

0
0 2 4 6 8 10 12 14 16 18 20

Figure 13: Energy Balance for 475-year Record

Discussion and Conclusions References

The results of this case study show that the 2000 Gimmel, Lindorfar, and Miyamoto, “Design of a New
NEHRP procedure is a very effective way to design a Moment Frame Building Incorporating Viscous
damped structure. The elastic frequency of the structure Dampers Following the Guidelines of the 1999 SEAOC
is shifted to a low frequency that generates lower floor Blue Book,” Proceedings of ATC17-2, May 2002,
and roof accelerations while the story displacements are Redwood City, California.
controlled by dampers. The maximum story drift was
limited to less than 1.0%, while all members remained NEHRP, 2000, “Recommended Guidelines for the
elastic. The maximum base shear is 0.29g for a 475- Seismic Design of Buildings and Other Structures,”
year return event. The final study will include results of FEMA 368, Washington, DC
a 2,500-year return. These parameters indicate that
structural and non-structural damages are significantly ICBO, 1997, Uniform Building Code, International
reduced when compared to conventional lateral system. Council of Building Officials, Whittier, California.
The cost of FVDs are effectively offset by the reduction
in costs of the foundation system and the structural J.P. Singh and Associates, 2002, “Geotechnical and
steel of the roof beams. Seismological Findings and Recommendations,”
Richmond, California.

Marr Shaffer & Miyamoto, Inc.


1450 Halyard Dr., Suite One
West Sacramento, CA 95691

@Seismicisolation
@Seismicisolation
266
CASE STUDY: 9

Seismic Retrofit of the Tower of Hope – Preservation of a


Masterwork of Mid-Century Modernism

Bryan Seamer, S.E., Director


Daniel Wang, S.E., Director
LPA, Inc.
Irvine, CA

Abstract Background

Richard Neutra’s iconic Tower of Hope on the Christ Built in 1968, the Tower of Hope was the final piece of the
Cathedral (formerly “Crystal Cathedral”) campus in Garden four-building campus that formed the original home to
Grove, California has been an important Orange County Reverend Robert H. Schuller’s growing Reformed Church of
landmark since it was built in 1968. The thirteen-story tower America congregation in Garden Grove, California. Designed
– the tallest building in Orange County when it was built – has by famed international architect Richard Neutra, the Tower of
been called an “overlooked masterwork in Neutra’s oeuvre” by Hope joined Neutra’s Arboretum worship hall and the Large
architectural historians. and Small Galleries to create an enclosed garden courtyard at
the heart of the campus. The Tower was originally planned to
Like many concrete buildings built prior to the 1971 Sylmar be a low-lying companion to the other low-profile buildings on
Earthquake in California, the Tower of Hope’s concrete frames the site but was ultimately reconceived as a slender vertical
lack the ductility needed to safely dissipate seismic energy. tower with 28,000 square feet of offices and classrooms in
After acquiring the Crystal Cathedral campus in 2012 the thirteen stories. The Tower of Hope along with the other three
Roman Catholic Diocese of Orange undertook a Neutra-designed buildings on the Christ Cathedral campus are
comprehensive renovation and seismic retrofit project to recognized by architectural historians as important examples
provide 21st century seismic resilience to the historic tower. of mid-century modernism as well as works in Neutra’s
This challenging seismic retrofit and renovation project was celebrated portfolio.
completed in 2015. The retrofit work included the installation
of fluid viscous dampers on the second through fifth floors of Reverend Schuller’s ministry grew dramatically during the
the tower in combination with fiber-reinforced polymer 1970s and 1980s as his televised “Hour of Power” became
strengthening of targeted concrete columns and walls. synonymous with televangelism and his campus grew to
include Philip Johnson’s landmark glass-and-steel clad Crystal
This paper focuses on two challenges unique to the Tower of Cathedral directly to the north of the Tower of Hope.
Hope. First, it was imperative that the retrofit design respect
the historically significant mid-century modernist architecture, In 2012, the Roman Catholic Diocese of Orange purchased the
preserving those features that were emblematic of that period former Crystal Cathedral campus including the Tower of Hope
of significance. Seismic retrofit construction was limited to from Reverend Schuller to serve as its long-planned diocesan
areas that didn’t affect Neutra’s open floor plate design cathedral. The Diocese immediately begin a program of
aesthetic or lessen the inside-outside connectivity of each of modernization and renovation of all of the buildings on the
the spaces. This openness was particularly challenging to newly re-named Christ Cathedral campus. From the beginning
preserve in the glass-walled first floor lobby where seismic the fate of the Tower of Hope was in doubt. During the
forces are at their most intense. The second unique challenge Diocese’s acquisition due diligence process in the Fall of 2011,
was the large damper connection forces that had to be a seismic assessment suggested that the Tower of Hope was
developed into the existing cast-in-place concrete frames the most vulnerable building on the Cathedral campus. While
without damaging the existing steel rebar. The strategies the Diocese recognized the Tower’s architectural and cultural
described by the authors are generally applicable to other significance, it decided that the safety of its large parish
historic buildings from the mid-century modernist movement population must ultimately take precedence. It was at that time
and to the use of fluid viscous dampers to retrofit concrete that contingency plans were made to demolish the Tower of
frames. Hope and replace it with a modern office building in case a
viable seismic retrofit solution could not be devised.

@Seismicisolation
@Seismicisolation
267
Building Description

The Tower of Hope was originally built to house offices and


classrooms supporting Reverend Schuller’s growing
Reformed Church of America. The tower consists of thirteen
occupied stories with a double-high volume thirteenth story
that houses the famed Chapel in the Sky and its panoramic
views of Orange County. The building footprint is very small
by modern office building standards, 58’-4” in the longitudinal
direction and 32’-2” in the transverse direction. With less than
1,900 gross square feet of space per floor the building
configuration is more akin to a series of stacked rooms than a
traditionally subdivided office building. The floor to floor
height is typically 11’4” with the 13’8” first floor and the 25’3”
high thirteenth floor chapel serving as exceptions. The roof
height is approximately 162’above grade. An interstitial floor
above the thirteenth floor houses a machine room above the
building’s two elevators on the west side of the tower.

Figure 1 – The Tower of Hope’s 13th floor Chapel in Figure 2 – Tower of Hope South Elevation.
the Sky offers 360o views of Orange County.

The structural system of the tower is entirely cast-in-place


concrete. Floor construction typically consists of 4 ½”
concrete slabs spanning between 27” deep concrete beams
spaced at 11’-0” on center. The beams are not isotropic along
their length, the width varying from 20” at each end to 12” at
mid-span. These main beams span north to south across the
32’-2” dimension providing completely open column-free
floors. Perimeter concrete columns are rectangular 16” by 28”
on the south side and trapezoidal 16” by 42” on the north side
of the tower. Column reinforcing consists of tightly confined
circular cores with more widely tied perimeter longitudinal
bars forming the finished shapes of the columns.

@Seismicisolation
@Seismicisolation
268
Figure 4 – Typical floor plan showing locations of
concrete moment frames (in blue) and concrete
shear walls (in red).

Several building elements extend out from the tower footprint.


An enclosed interior stair tower extends from the northwestern
corner of the main tower. With 12” thick concrete walls on
two sides the cast-in-place concrete stair ties into each main
floor diaphragm via an 8’4” wide 6” thick concrete slab. At
the southeast corner of the main building a second exterior stair
cantilevers thrillingly from the tower. This column-free stair
is cantilevered almost 12’ from the main floor slabs with post-
tensioned tendons. Thin architectural steel cables at 5” on
center enclose two sides of the vertigo-inducing stair but do
not serve as structural hangers for the far landing. One of the
most prominent features of the building is illuminated steel
cross that soars to 88’ above the roof of the tower.

The lateral force resisting system of the tower consists of


concrete moment frames formed by the perimeter columns and
beams and the main transverse interior beams. Supplemental
lateral force resistance is provided by the two 12” thick
Figure 3 – Erection of the steel cross on the roof of concrete walls that are on the north and west sides of the
the Tower of Hope in 1968. (Richard and Dion Neutra northwest stair tower. The location of these shear walls, far
Paper Archives, 1925-1970). from the main tower’s center of mass, and the lack of a
similarly rigid element on the other side of the tower result in
The foundation of the building is a Raymond step-tapered pile a torsional response of the tower when loaded in the north-
system with pile caps tied to each other with reinforced south direction.
concrete grade beams. The Raymond pile system, developed
by Alfred E. Raymond in 1893, consists of a series of helically-
corrugated cylindrical steel shells driven into the soil in 8’ to
16’ lengths. Each subsequent shell is wider than the shell
below it so the diameter of the pile is tapered along its length.
The steel shells are filled with concrete as the shell-driving
process advances into the soil to form a steel-concrete
composite pile.

@Seismicisolation
@Seismicisolation
269
Seismic Vulnerabilities
The initial seismic assessment of the Tower of Hope was based
When the Diocese of Orange purchased the Christ Cathedral on a “Tier 1 Screening” as described in ASCE Standard 31-03
campus in 2012 the due diligence phase of the real estate “Seismic Evaluation of Existing Buildings.” This process
transaction identified the Tower of Hope as a building of consists of a series of quick checks to identify potential
elevated seismic risk due to its age and concrete moment frame vulnerabilities that warrant more detailed study. The
construction. The seismic vulnerabilities associated with non- prescriptive checklist of potential vulnerabilities was
ductile concrete frames and the risk they pose to buildings of supplemented by a detailed review of the construction
this age and construction type are well-known. As part of a documents by experienced structural engineers to identify
campus wide modernization and renovation program the potentially brittle concrete details and other system-wide
Diocese solicited the services of several structural engineers, vulnerabilities.
including Irvine-based integrated design firm LPA, Inc., to
perform a detailed evaluation of the Tower of Hope. This Several serious deficiencies were identified during the
initial assessment comprised three basic steps: data collection, screening phase, most of them related to non-ductile detailing
seismic screening, and identification of potential deficiencies. of the concrete frames:

A common challenge with assessment and retrofit of buildings  Inadequate confinement of column reinforcing. The
of this age is that original construction documents are not often central core of each columns is confined with a tight
available. This potential challenge is amplified in a concrete spiral of #4 bars at a 2” on center. However, the
building because direct observation of the steel reinforcing is remainder of the vertical bars, including those at the
impossible and non-destructive testing methods are time- perimeter of the column that are potentially most
consuming and not always accurate. This challenge was effective in resisting flexural forces is confined with
largely bypassed on the Tower of Hope retrofit project, #3 ties spaced at 12” on center.
however, because the building is an important piece of
architectural history and the design team had incredible access  Short splices in column vertical bars. Typical column
to original sources of information on the design and splices are 30 bar diameters.
construction of the building.
 Vertical column bars are not fully developed into the
For the Tower of Hope project these issues were avoided foundation.
entirely due to the careful preservation of Richard Neutra’s
records by architectural historians at the “Richard and Dion  Frame beam longitudinal bars not fully developed
Neutra papers, 1925-1970” archive at the Charles E. Young into frame columns. In multiple locations not all of
Research Library on the campus of UCLA. This archive holds the longitudinal bars are fully developed into the
nearly comprehensive documentation on the design and columns due to 90-degree hooks that don’t extend far
construction of the Tower of Hope. Complete construction enough into the columns or bottom bars that don’t
drawings by Richard and Dion Neutra, Architects and have hooks at all.
Associates and J. Kinoshita & Associates Consulting
Structural Engineers dated May 15, 1966 proved to be  Torsional irregularity. The stair tower at the
instrumental in understanding the construction of the Tower. northeast corner is enclosed with 12” thick concrete
In addition to original construction documents, the design team walls while the much larger main tower is a moment-
was able to review meeting minutes, correspondences, resisting space frame. The difference in lateral
construction RFIs, submittals, and crucially, inspection and stiffness of these two systems leads to a torsional
testing reports. The availability of original concrete testing response and induces amplified seismic forces in the
reports was important for two reasons. First, it gave the design outer frames and the relatively narrow portion of floor
team confidence in the as-built compressive strength of the slab that ties the stair tower to the rest of the structure.
concrete. This confidence is directly applied analytically in
the form of a knowledge factor, , that is a part of the seismic Taken in total these deficiencies – particularly those related to
retrofit provisions of ASCE Standard 41-06. The most non-ductile concrete detailing – represent a serious risk to the
important discovery during the design team’s review of building despite a seismic-force resisting system that is
Neutra’s project records was the fact that during construction otherwise relatively well-proportioned and redundant for a
the contractor decided to use 4,000 psi concrete in lieu of the building of this size.
3,000 psi concrete called for in the structural. This change is
not insignificant in relation to the seismic performance of the
building as is discussed further below.

@Seismicisolation
@Seismicisolation
270
Figure 5 – Typical column reinforcing detail.

Figure 7 – Original construction drawing showing


column to pile cap connection, column bar lap
length and column tie configuration

Figure 6 – Representative beam to column joint


detail.

@Seismicisolation
@Seismicisolation
271
Seismic Retrofit Constraints

The Roman Catholic Diocese of Orange approached the


renovation and retrofit project with the safety of its large
congregation and staff as the main priority but knew that
preserving the Tower of Hope would be challenging for
several reasons. As a historically significant work by one of
the masters of the modern architecture movement preservation
of Neutra’s original design aesthetic through the retrofit
process was essential. To ensure that the historic architecture
was maintained and celebrated through the renovation project
Barbara Lamprecht, an architectural historian and expert in the
works of Richard Neutra joined the design team. With her
guidance, LPA’s team of architects and structural engineers
designed renovation and retrofit measures that were true to the
architectural period of significance. Specifically, the inside- Figure 9 – Southern view from lobby to courtyard and
outside connectivity between the glass-walled lobby and the fountain.
fountain and garden to the south and entrance to the east
needed to be maintained. This dictum essentially precluded Another factor limiting the placement of new seismic load
the addition of new lateral-force resisting elements at the resisting elements is the very small floor plates. At only 32
perimeter of the first floor - the location where such elements feet wide by 58 feet long the usable square footage of each
would be most effective. floor is very valuable. Any new walls or braced frames in the
interior of the building would make space planning a major
The Tower along with the former Crystal Cathedral campus is challenge and make it nearly impossible to design fully open
also an important part of the history of Orange County, CA. floors. Further, because the Diocese’s program included a
The Diocese chose to preserve the top floor Chapel-in-the-Sky combination of entirely open floorplan classrooms, subdivided
and Rev. Robert Schuller’s private office on the twelfth floor office floors, a television studio, a chapel and the open-plan
in recognition of the building’s regional historical significance. lobby there were no common walls at the interior of the
building that stacked from floor to floor.

Another major constraint to designing an effective seismic


retrofit strategy was cost-related. Due to site soil conditions the
Tower of Hope is supported by deep pile foundations. A
traditional retrofit design that adds significant lateral stiffness
to the tower would lead to increased foundation loads and the
need for new piles. Construction of any new piles, while
technically possible beneath the existing building, proved to
prohibitively expensive. The exterior glazing and cladding of
the tower also needed to be preserved both because of the
historic aesthetic and because of the cost needed to replace it.
Any seismic retrofit design would need to be able to be built
without replacing or damaging the glass.

These constraints when taken together form a set of seismic


Figure 8 – Glass-walled lobby looking west. design criteria that severely limited the possible retrofit
options.

@Seismicisolation
@Seismicisolation
272
Seismic Retrofit Design shear, flexure, and axial behavior and rebar slip deformations
per the requirements of ASCE 41-06.
Based on the aesthetic, historic, economic, and practical
constraints the seismic retrofit strategy for the Tower of Hope The seismic analysis of the building was performed using
was designed to meet the following objectives: ETABs Version 9.7 structural analysis and design software
published by Computers and Structures, Inc. The finite
 Respect the period of architectural significance and element model was subjected to seven pairs of site-specific
historical context of the Tower by not adding response spectra-scaled time histories for each of the two
structural elements to the first floor, twelfth floor earthquake hazard levels. The time histories were constructed
offices of Reverend Schuller or thirteenth floor by Leighton Consulting, Inc. geotechnical engineers based on
Chapel-in-the-Sky earthquake records having magnitudes between 6.0 and 7.0 at
 Limit new seismic-force resisting elements to the distances ranging from 10 to 20 kilometers and geologic and
perimeter column lines to maximize usable interior seismic/tectonic environments compatible to the site of the
space and allow for future flexibility. Tower of Hope. Leighton Consulting, Inc. built these site
 Avoid the need for adding new foundation elements specific acceleration time histories to meet the requirements of
in order to minimize construction costs. Section 1.6.2.2 of ASCE 41-06. Because seven sets of time
histories were considered in the analysis the average value of
LPA, Inc. structural engineers worked closely with the Diocese each of the maximum response parameters from each time
of Orange to establish the structural performance objectives for history was used for assessing the acceptability of each
the seismic retrofit. In accordance with the voluntary seismic structural element. Multi-directional seismic effects were
retrofit provisions of the 2013 California Building Code and taken into account by using 100% of the response parameter
American Society of Civil Engineers Standard 41-06, “Seismic with a given time history applied in the X-direction combined
Rehabilitation of Existing Buildings” (ASCE 41-06) the with 30% of the response parameter with the time history
following structural performance objectives were selected: applied in the Y-direction. This resulted in fourteen time
history analyses for each of the two earthquake hazard levels.
 Life Safety performance during an earthquake having
a 10% probability of exceedance in 50 years (BSE-1) The results of the time history analysis were exported to
 Collapse Prevention performance during an Microsoft Excel and post-processed using proprietary LPA,
earthquake having a 2% probability of exceedance in Inc. spreadsheets and Visual Basic macros. Each existing
50 years (BSE-2). concrete beam, column and shear wall that resists seismic
forces was checked against the ASCE 41-06, Supplement No.1
In order to satisfy both the practical and analytical project acceptability criteria for both Life Safety and Collapse
objectives two specialized structural components were used in Prevention performance. These acceptance criteria are based
tandem. First, supplemental damping was added to the on multiple force-controlled or deformation controlled actions
building in the form of diagonally-oriented fluid viscous for each element and explicitly include consideration of stress
dampers. This served to reduce the seismic demand on the level, rebar splice length, confining reinforcement, and
existing concrete frames without adding significant foundation development of rebar into beam-column joints. Because of the
forces. Second, fiber-reinforced polymer (FRP) was added to lack of ductile rebar detailing at the Tower of the Hope the
select concrete columns and walls for increased strength. acceptance criteria of many of the concrete beams and columns
necessitates nearly elastic behavior.
The structural design for the seismic retrofit followed the
linear dynamic procedure of ASCE 41-06 using site specific
time histories. The linear time history procedure was chosen
for two reasons. First, the fluid viscous dampers are velocity-
dependent so a time-history analysis was needed to model the
effect of this supplemental damping. Because of the limited
ductility of the existing concrete beams and columns these
elements had very little post-elastic capacity so any effective
retrofit design would necessarily result in nearly linear behaver
of these elements. Because of this practical reality and for
computational efficiency a linear time history analysis was
performed. The concrete elements were modelled with
reduced effective stiffness parameters between 30% and 70%
of EcIg for flexure and 40% of EcAg for shear to account for

@Seismicisolation
@Seismicisolation
273
resulted in dampers with a maximum of 260 kips of axial force
and 4” of stroke during the suite of BSE-2 time histories. A
factor of safety of 2.0 against yielding and 2.5 against ultimate
failure was used in the design of the damper components.
Taylor Devices of North Tonawanda, NY designed and
fabricated the 32 dampers for the Tower of Hope.

During the preliminary design phase of the project a review of


the original Tower of Hope construction files uncovered
concrete submittals and field inspection and testing reports that
revealed that the as-designed 3,000 psi concrete for the
superstructure was replaced with 4,000 psi concrete during
construction. The design team hypothesized that this change
was made to accelerate erection time as each subsequent level
had to reach adequate strength before the next level up could
be poured. This construction change had an initially
counterintuitive effect on the retrofit design. While the high
strength concrete did have more capacity than originally
thought it also resulted in a more rigid structure. This added
stiffness led to higher seismic forces and reduced the
effectiveness of the fluid viscous dampers which generate
damping to the structure proportionally to their drift-induced
stroke velocity.

In its original condition, without any supplemental damping


approximately one third of the existing concrete elements did
not meet the seismic performance objectives required by
ASCE 41-06. With damping added on levels two through five
only about 6% were still deficient. These remaining deficient
elements were strengthened with fiber-reinforced polymer
(FRP) wrap to add flexural capacity, confinement or both.

Figure 10 – ETABS model of Tower of Hope in


original condition.

The final retrofit design for the Tower of Hope balanced the
addition of fluid viscous dampers and FRP-strengthening of
columns and walls with the goal of minimizing total
construction cost. To that end, the retrofit design process was
iterative with supplemental damping increased until the
addition of more damping had only incremental effect on the
acceptability of the existing concrete frames and shear walls.
The final result was a design that included dampers added in a Figure 11 – Fluid viscous dampers on the north side
two-story X configuration on each of the perimeter column of the Tower of Hope.
lines on stories two through five. After analytically
experimenting with several different combinations of damper
properties it was determined that dampers with a damping
constant, C, of 120 kip-sec/in and a velocity exponent, , of
0.5 was most effective for this building. These properties

@Seismicisolation
@Seismicisolation
274
Figure 12 – West elevation of the Tower of Hope, fluid
viscous dampers visible at second through fifth
floors.

The Raymond Step-Tapered pile system was also evaluated


using the acceptance criteria detailed in ASCE 41-06.
Ultimate pile uplift, compression, and lateral capacities were
derived from information found in the Raymond Pile design
guide and soil properties provided by the project geotechnical
engineer. This analysis demonstrated that the existing pile
system met the seismic performance objectives without need
for retrofit.

Figure 13 – Structural model of the Tower of Hope


with dampers on second through fifth floors.

@Seismicisolation
@Seismicisolation
275
Seismic Retrofit Detailing and Construction

The seismic retrofit of this historically significant 1968 non-


ductile concrete frame building included unique detailing
challenges both structurally and architecturally.

The most challenging part of the structural design was


developing the large damper forces into the existing concrete
frame. The vertical component of the damper reaction was
resisted by Grade 105 bolts through the existing concrete
frame beams near the diagonal damper driver connection point
at beam mid-span or at the beam-to-column joint. Horizontal
forces were resisted with a series of Hilti KBTZ expansion
anchors along the entire length of the concrete frame beams.
This resulted in steel connection plates that ran the length of
the floor at the perimeter of the building between columns
which practically maximized the ability for the concrete frame
to transfer forces into the dampers without failure of these
connections.

The installation of the critical damper to concrete frame


connection bolts and expansion anchors proved equally
challenging. Because the essence of the retrofit strategy was
to take advantage of the capacity of the existing concrete
frames it was imperative to preserve as much of that inherent
capacity as possible during installation of the dampers. This
meant that any connections between new steel elements and
existing concrete beams and columns needed to avoid
Figure 14 – Tower of Hope entrance with FRP visible damaging existing steel rebar to the extent that was possible.
on exterior face of concrete column. While the record drawings showed sizes, quantities, and
configuration of all of the column and beam reinforcing the
precise location could not be determined without extensive
testing in the field. Ground penetrating radar (GPR) was used
extensively during construction to precisely locate the existing
rebar. The design allowed for small adjustments of connection
bolt locations in the field to avoid existing rebar although in
several locations connections had to be reconfigured and
reanalyzed during construction to allow through-bolts to snake
through the existing beams without damaging their
reinforcing. This resulted in highly customized connections at
many of the 64 damper and driver connection locations. In
order to keep the project on schedule much of the steel
connection fabrication took place on site with bolt holes being
drilled in plates and plates welded to braces in the field only
after a viable bolt pattern had been identified and analyzed on
a case-by-case basis. The contracting team of MATT
construction and Saunders Commercial Seismic implemented
a 24-hour work day in three 8-hr shifts, six days a week during
the four-week long damper installation phase of the
construction. In the end the design and construction team
worked hand-in-hand to successfully install over 250 anchors
and bolts.
Figure 15 – Column confinement FRP installation.

10

@Seismicisolation
@Seismicisolation
276
Figure 16 – Damper connection to existing concrete
beam-column joint detail.

At the west side of the building the concrete frame beams are
narrower than on the other three sides because of the proximity
of the building’s elevator shaft. This condition reduced the Figure 17 – Damper connection to beam and column
effectiveness of expansion anchors installed on the top surface at southern column line.
of the beam due to reduced concrete edge distance. Because of
this Grade 105 through-bolts were added through the two Maintaining and preserving the mid-century modernist
columns on the west side of the building. The through-bolts at aesthetic and the specific elements of Richard Neutra’s design
both column and beam were designed to resist the combined was as important as the structural engineering requirements of
effects of tension and shear. Because the column ties were the project. While much of the FRP wrap occurred within the
spaced too closely to avoid damaging them when installing the curtain-wall envelope of the Tower there were locations where
through-bolts the retrofit design included the addition of FRP was needed on the exterior of the building as well. In
confinement FRP between the connection plates and the order to hide the FRP yet maintain an aesthetic true to the
concrete columns on this side of the tower. The sequencing original period of architectural significance the design team
and coordination between the GPR testing company, structural worked closely with the team’s architectural historian to
engineer, FRP sub-contractor and steel sub-contractor had to develop fluted 10-gage metal cladding that was differentiated
be carefully orchestrated to ensure that the dampers and FRP from but consistent with the original Neutra design.
could be installed without damaging the longitudinal column
and beam reinforcing.

11

@Seismicisolation
@Seismicisolation
277
 The addition of supplemental damping to an existing
structure is an effective way to significantly reduce
seismic demands on the building. Fluid viscous
dampers are most effective near the base of the
building but need not extend through the first floor to
the foundation to improve overall seismic
performance of a structure.

 Seismic retrofit in a concrete building may be limited


by the ability to transfer seismic forces between the
existing concrete structure and strengthening
elements. This limitation is particularly important
when a small number of bracing elements are added
because the connection capacity between new steel
elements and existing concrete is relatively small.

 The Tower of Hope seismic retrofit project was


successful in part because of the relatively small floor
plates of the building. Each floor comprises less than
1,900 square feet of space so perimeter dampers
proved effective. The supplemental damping needed
to effectively retrofit the Tower of Hope resulted in
approximately one damper per 500 square feet of
floor area in each direction. To be more generally
applicable for typical non-ductile concrete moment
frame retrofit in buildings with larger floor plates it
should be understood by the design team that more
dampers than is architecturally practical may be
needed.

Figure 18 – Architectural metal paneling to hide FRP  Locating and avoiding existing steel reinforcing in a
at column adjacent to building entrance. Reference large concrete frame is very challenging. While the
Figure 14. quantity of bars may be understood by reviewing
record drawings and other as-built information the
Conclusions exact location in the field may vary by several inches
in any direction. Careful detailing to allow for
This paper presented a case study in performance-based flexibility during construction and thoughtful
seismic retrofit for a historically significant non-ductile coordination and planning between the structural
concrete frame building using a combination of fluid viscous engineer, contractor, testing company and project
dampers and fiber reinforced polymer (FRP). Several of the inspector is essential when new steel elements are
lessons learned by the project team on this project may be being added to an existing concrete building.
broadly applicable to other projects with similar project goals
and features. These general conclusions include the following: Acknowledgements

The authors would like to thank the Roman Catholic Diocese


 Respect for the features of the building that make it
of Orange County for their commitment to preserving this
historically significant are important. The project
iconic building and their willingness to embrace an innovative
team for the Tower of Hope included an architectural
performance-based design retrofit strategy. The Diocese team
historian who is an expert in the works of Richard
was led by Rob Neal, Managing Partner of Hager Pacific
Neutra to ensure that the seismic retrofit and
Properties, and Tracy Bejotte, Chief Operating Officer of the
associated renovation work was differentiated from
Christ Catholic Cathedral Corporation.
yet compatible with the building’s period of
architectural significance.

12

@Seismicisolation
@Seismicisolation
278
The authors would also like to acknowledge the contributions
of the LPA, Inc. integrated architecture and interior design
team led by Jim Wirick, AIA, and Maria Louie.

Finally, the dedicated construction team led by Kevin Pitzer of


MATT Construction and Amber Zamora of Saunders
Commercial Seismic deserves recognition for their
commitment to high quality work on this very demanding
construction project.

References

ASCE, 2003, ASCE Standard 31-03 Seismic Evaluation of


Existing Buildings, 2003, American Society of Civil
Engineers, Reston, Virginia.

ASCE, 2006 ASCE Standard 41-06 Seismic Rehabilitation of


Existing Buildings w/ Supplement No. 1, 2006, American
Society of Civil Engineers, Reston, Virginia.

ACI, 2011, Building Code Requirements for Structural


Concrete (ACI 318-11), American Concrete Institute,
Farmington Hills, Michigan.

Barbara Lamprecht, lamprecht, archiTEXTudal, personal


communication, various dates.

CBSC, 2013, California Building Code, California Code of


Regulations, Title 24, Part 2, Volume 2, California Building
Standards Commission, Sacramento, California.

Computers and Structures, Inc., 2011, ETABS version 9.7


[Computer software], 2011, Berkeley, California.

Leighton Consulting, Inc., 2012, Geotechnical Exploration for


Proposed Seismic Upgrade of Tower of Hope and Arboretum
Building Existing Crystal Cathedral Campus 12141 South
Lewis Street Garden Grove, California, Project No. 10021.001

Raymond International Inc., Raymond Step-Taper Piles, 1977,


Houston, Texas.

Richard and Dion Neutra papers archive, 1925-1970, Charles


E. Young Research Library, University of California, Los
Angeles.

13

@Seismicisolation
@Seismicisolation
279
CASE STUDY: 10

Strong
Renovation and Retrofit

Medicine
By Douglas R. Wilson, P.E.,
Russell D. Kent, P.E.,
Stephen Stanek, P.E., and
David B. Swanson, P.E., S.E.

N
aval Hospital Bremerton, in This effort provided the US Navy and the
Bremerton, WA, serves 60,000 design team with a “big picture”
A much-needed military families in the Puget overview of the building stock and seis-
seismic retrofit is just Sound area. Located near
Seattle, it is only one of two
mic hazards, as well as an initial relative
ranking of seismic risk among the facili-
what the doctor major hospitals on Washington’s Kitsap ties. It was an exceptional tool to priori-
Peninsula. In the aftermath of a serious tize further work for investigating in
ordered for this aging natural disaster, like a large-scale earth- more detail those facilities with the high-
quake, the hospital could be called on to est risk.
naval hospital immediately serve more than 250,000 One of the high-risk buildings was the
people. late 1960s-era main hospital building—a
serving thousands nine story, 250,000 sq ft. structure with a
on Washington’s First Step structural steel moment frame, compos-
The hospital complex includes more ite concrete on metal deck floors, precast
Puget Sound. than 20 buildings, some of which were concrete cladding, and concrete stair
constructed as early as the 1930s. The US towers. The main hospital building did
Navy wanted to know the seismic risk of not have the worst hazard score. How-
the Bremerton medical facility, and how ever, the structure is significantly larger
best to go about mitigating that risk. than any of the other medical buildings
Starting in 1999, structural engineers from in the complex and is home to the most
Reid Middleton embarked on a series of essential medical functions.
seismic screenings and evaluations of the
various naval hospital facilities to system- Recommendation
atically determine seismic deficiencies. A detailed seismic evaluation of the
The first step was to understand the hospital using performance-based engi-
extent and type of seismic structural haz- neering standards (FEMA 310 – Hand-
ards and evaluate the risk based on book for the Seismic Evaluation of Buildings–
building type, use, and occupancy. A Prestandard and FEMA 356 – Pre-
FEMA 154 – Rapid Visual Screening of standard and Commentary for the Seismic
Buildings for Potential Seismic Hazards: A Rehabilitation of Buildings) was performed
Handbook techniques were employed to to gain a better understanding of the
screen and document initial findings. potential seismic deficiencies.

Modern Steel Construction • February 2005

@Seismicisolation
@Seismicisolation
280
The building’s lateral force resisting
Nisqually Earthquake Scaled to 10% in 50 Years
system (LFRS) is comprised of a highly
redundant steel moment frames system ∆ = 12.1”
at all beam-to-column connections. With (0.74% Drift)
Roof
more than 1,200 moment connections, the Floor 7
LFRS has a large amount of redundancy.
Floor 6
However, it is too flexible, resulting in
∆ = 3.8” Floor 5
excessive drift, large torsional response (0.47% Drift)
of the narrow tower, high expected rota- Mechanical

tional demands, and forces on the “pre- Floor 4

Northridge” 1960s-era moment frame Floor 3

connections. Floor 2
Additionally, there was incompatibil-
ity between the flexible LFRS and the Floor 1

rigid concrete stair. The cladding panel


Basement
connections were not designed to accom-
modate the expected drifts from a Transverse Direction - Pre-Retrofit
design-level earthquake and presented a
In its pre-retrofit state, the hospitalʼs story drifts could be significant.
potential falling hazard for the building
exit paths. Excessive drifts also caused
some of the building columns to be over-
stressed in axial load, combined with Nisqually Earthquake Scaled to 10% in 50 Years
biaxial bending.
This detailed seismic evaluation was ∆ = 7.2”
(0.44% Drift)
completed in late 2000 and recom- Roof

mended a comprehensive seismic retrofit Floor 7

of the hospital. Floor 6

∆ = 2.4” Floor 5
Mother Nature Steps In (0.30% Drift)
Mechanical
In February 2001, the magnitude 6.8
Floor 4
Nisqually Earthquake shook the Puget
Sound area. Shaking at the hospital was Floor 3

modest because of the earthquake’s Floor 2

depth, and because its epicenter was


Floor 1
located approximately 30 miles away. A
seismograph in the hospital recorded a Basement

horizontal peak ground acceleration of


0.11g at the basement level and a peak Transverse Direction - Post Retrofit
roof acceleration of 0.47g. The hospital The seismic retrofit significantly reduces story drifts. It will also reduce floor accelerations by
structure experienced significant lateral about 1/3 for all floors.
drifts during the small, “less than design-
level” earthquake, particularly on the complete detailed inspections. It took cant cracks or fractures) related to the
upper floors of the tower. Calculated several months for all of the earthquake modest earthquake shaking was
peak roof displacements from this mod- repairs to be made and for the building observed.
est earthquake shaking were over 6” (a to be restored to full use.
drift ratio of 0.5%). The movement Because the main hospital was con- Performance-Based Design
caused a significant amount of damage to structed in the late 1960s with pre-North- Because a traditional seismic retrofit
non-structural features and finishes in ridge steel moment frame connections, a by strengthening and stiffening the
the hospital—especially at upper floors. detailed inspection and testing program moment frame LFRS would have been
It could have been much worse. If a in accordance with the FEMA 350 series costly and disruptive, alternative retrofit
design level earthquake had struck Puget standards was undertaken to investigate design schemes were evaluated. The use
Sound, roof displacements of the nine- whether the earthquake had damaged of supplemental seismic damping
story hospital could have been several the moment frame connections. Thirty of proved to be the best design scheme to
feet, resulting in much more significant the approximately 1,550 moment connec- improve the seismic performance of the
damage and loss of hospital function. tions were exposed and visually 30-plus-year-old building while mini-
Approximately six hours after the inspected for damage. Several of these mizing disruption during construction.
earthquake, the facilities staff had per- connections were also inspected using The purpose of supplemental passive
formed rapid inspections of the hospital, non-destructive ultrasonic and magnetic damping is to reduce lateral displace-
and enough information had been col- particle test procedures to increase the ment of the building through benign dis-
lected and analyzed to allow the hospital likelihood of finding damage not identi- sipation of the earthquake’s energy
be reopened for further operations. It fied by visual inspection. Of the 30 con- through heat created in the damper sys-
took several days for Reid Middleton to nections inspected, no damage (signifi- tem. One of the challenges of seismic

February 2005 • Modern Steel Construction

@Seismicisolation
@Seismicisolation
281
ity to minimize the impact of construc-
tion on essential hospital operations. The
project will breathe life back into the
building for a cost in line with other
upgrades to the facility. ★

Douglas Wilson is a project engineer for


Reid Middleton, Inc.’s Structural Engineer-
ing Group. Russell Kent is facility manager
for Naval Hospital Bremerton. Stephen
Mock-ups of the dampers were used at vari- Stanek is a project lead with Engineering
ous locations in the hospital to gauge the
aesthetic and operational impact of damper Field Activity Northwest. David Swanson is
locations. director of Reid Middleton, Inc.’s Structural
Engineering Group.
retrofit projects is “not making things was able to “tune” his site-specific
worse.” Retrofits can potentially alter the ground response study to the characteris-
original load paths, causing members to tics of the site and building. This Owner
carry loads they were not originally response record provided the design US Navy
designed for. team with a benchmark response, and
Architect
The passive damping mechanism the owner with a clear understanding of
CollinsWoerman, Bellevue, WA
supplements the existing LFRS. It is a future building performance.
permanent installation that remains in Slated for construction in 2005, a total Structural Engineer/Prime
place and is functional for the life of the of 88 seismic dampers will be installed at Consultant
structure, even following a seismic event. 44 select locations in the building. Each Reid Middleton, Inc., Everett, WA
Through energy dissipation, dampers damper has a 200 to 300 kip capacity with
significantly reduce the seismic forces, a ±3” stroke. Some of the dampers will be Engineering Software
displacements, and floor accelerations in left exposed to the view of patients and ETABS
the structure, thus reducing or eliminat- staff, while others will be concealed in
ing earthquake damage to the building’s existing wall cavities and by new fin-
primary structural system and many of ishes. The installation of seismic dampers
the non-structural systems. is an effective means to seismically retro-
Supplemental seismic damping was fit essential facilities like Naval Hospital
designed to be installed at discreet loca- Bremerton to improve their earthquake
tions throughout the main hospital build- performance and post-earthquake func-
ing. Because the seismic damping system tionality.
is passive and supplemental, it does not
appreciably change the fundamental Cost
LFRS response for wind loading in the The estimated replacement cost of the
building. main hospital is $60 million, while the
One benefit of this damping system is estimated cost of a seismic retrofit is $4
the ability to shift damper locations from million. Because the existing building is
floor to floor to avoid extremely sensitive expected to support the hospital’s mis-
areas of the hospital. Also, shifting sion for at least 30 years, investment in
damper locations reduces the demands earthquake performance improvements
placed on the existing building columns, just made sense.
compared to a traditional retrofit where
the lateral elements are stacked from Results
floor to floor. The retrofit will have significant
impact on the building’s behavior:
Analysis • Story drifts and floor accelerations
State-of-the-art 3D nonlinear finite reduced by approximately 30% at all
element analyses were used in the design floors
of the hospital seismic retrofit. Target per- • Damped response for a 2,500-year
formance levels are “Immediate Occu- event reduced to the same magnitude
pancy” for the 10%/50 year Design Basis as the response of the un-retrofitted
Earthquake (DBE) and “Collapse Preven- building to a 500-year event
tion” for the 2%/50 year Maximum Con- • Diaphragm rotations reduced by 30%
sidered Earthquake (MCE). to 70% at all levels
Because the design team had an actual The use of fluid viscous dampers,
record of the building’s roof, fifth floor, along with performance-based design,
and basement responses to the Nisqually provides a solution that is cost effective
Earthquake, the geotechnical engineer and, just as importantly, has the flexibil-

Modern Steel Construction • February 2005

@Seismicisolation
@Seismicisolation
282
@Seismicisolation
@Seismicisolation
283
@Seismicisolation
@Seismicisolation
284
@Seismicisolation
@Seismicisolation
285
90 Taylor Drive
North Tonawanda, NY 14120 -0748
Phone: (716) 694-0800 | Fax: (716) 695-6015

www.taylordevices.com

© 2019 Taylor Devices Inc.


@Seismicisolation
@Seismicisolation
All rights reserved throughout the world. Edition: 05-19
286

You might also like