Experimental Testing of Precast Concrete Cladding For Building Facade Systems
Experimental Testing of Precast Concrete Cladding For Building Facade Systems
Experimental Testing of Precast Concrete Cladding For Building Facade Systems
Abstract: Since 1998, San Jose State University has conducted research into the seismic
behavior of building façade systems, particularly the precast concrete cladding systems. This
research has combined full-scale experimental studies and component experimental studies of
connections. In addition, the research team has been active contributors to two full-scale
experimental tests of complete building systems including precast concrete facade panels. The
testing has been combined with the development of structural analysis models to simulate
seismic behavior. The primary findings to date have been that modern precast concrete cladding
façade systems perform well when loaded to displacement levels less than the original design
displacement. New understanding of cladding performance will allow precast fabricators to
provide higher performing systems in the future.
Background. Building façade has several critical roles. It is the environmental barrier between
the interior and exterior functions of a building. The façade is often one of the primary
architectural design elements of the structure resulting in a critical aesthetic role. The façade can
have various structural roles, from providing the main structural support for gravity loads and
lateral forces to having a nonstructural role of providing gravity support only for itself and
distributing wind forces from the skin of the building to the main structure.
Over time, the interaction between the main structural system of common building
designs and the façade has changed. Traditional construction had the exterior surface of the
building as the primary if not sole vertical structural elements. With the advent of curtain wall
construction in the 1800’s, façade became a nonstructural element that was supported by the
main structural system of the building. This separation of the role of façade from the main
structure has continued, particularly as the size and shape of the vertical load carrying elements
have been reduced.
Some researchers propose the return of a larger structural role for façade system,
including the use of curtain wall type façade systems to provide the full lateral resistance of the
building. This paper reports the study of more traditional design, where the façade is considered
not to support any portion of the main structure, but to work as a nonstructural element providing
load to the main structure while successfully performing other roles. This traditional design
remains popular for multiple reasons, one of which is the clarity of each designer’s role in the
process. The architect specifies the elevation features, both shape, color, texture and window
design, the structural engineer specifies the size and shape of structural elements and the precast
façade engineer specifies the panel layout, panel structure, and panel materials.
Precast Cladding Systems: Precast concrete cladding with inset windows is one common
system for the exterior skin of commercial buildings. Cladding panels are precast at a fabrication
yard and delivered to the construction site where they are lifted into place and installed.
Typically one spandrel panel covers each perimeter floor beam. Column cover panels are then
installed in front of each column, sometimes supported by the spandrel cladding panels or
alternatively may be connected directly to the structural frame. Windows are installed to fill in
the region framed by the spandrel panels on adjoining floors and column covers on the adjoining
columns. Cladding systems are relatively similar whether installed on steel frame structures or
concrete frame structures.
The photograph
shows the
installation of
panels on the King
Library at San Jose
State University in
2002. Panels are
typically held in
place at each story
level with the
weight of the panel
supported by the
lower level beams,
similar to curtain
wall systems.
In another style of
panel layout,
spandrel panels
cover the floor
beams while
column cover
panels define and
highlight the
vertical columns of
the structural frame.
Cladding systems have changed continuously as new materials and new manufacturing
processes have resulted in technological advances. Hegel (1989) provides a typical cladding
panel and connection layout from the 1980’s. The use of spandrel beams and cantilevered
column panel arrangement and the connection configurations and locations appear similar to
current practice. Hegel explains how the arrangement of connections for precast panels has
remained relatively constant. Hegel explains that each structure-to-panel connection is intended
to have a single role: bearing connections support the weight of the panel, push-pull connections
resist the out-of-plane forces, and shear connections transfer the horizontal forces from the panel
to the building frame. Hegel suggests that the use of slotted holes or bending of steel
connections can allow the building to deflect laterally without undue interference from the
cladding system.
Precast concrete
façade systems
can be assembled
in many ways. A
common method
in California is
the use of panels
covering the
spandrel beams
and column
covers used over
the columns, as
highlighted in
red.
At the corner of
the building, a
return panel is
usually cast that
covers portions of
two of the
exterior
elevations.
To maintain a
consistent column
panel width for
the entire building
elevation, the
return panel often
has an adjoining
half-width
column cover that
is roughly half the
lineal length of
the main column
cover.
To complete the
building
enclosure,
window units are
often installed
between column
cover panels. The
window units are
usually supported
by the façade
spandrel panel
below.
The joints between panels are critical to the success of the façade. These joints are
typically 0.75 inches wide around all edges of the panels. The joints allow for tolerance
variation in the precast assemblies. The joints also allow thermal expansion and contraction of
the façade without damage. In addition, the façade must allow for movement of the main
structure, either lateral movement due to wind or earthquake, or vertical movement due to
foundation settlement. However the joints do cause challenges. They must be water-tight and
resist air movement. The typical finish for the joint is a sealant.
Precast concrete cladding facades are primarily assemblies of large rigid concrete blocks,
rigid glass windows, and relatively flexible steel connections. One aspect of the evolution of
structural engineering is the continuous movement toward more flexible building frames,
particularly to counteract the potential damage from major earthquakes.
Bassler et al (1992) reports general design features of cladding systems and the means to
allow for structural movement. They discuss the differences between using rocking assemblies
or swaying assemblies to allow the cladding to respond to lateral movement of the building
floors. They also provide discussion about joints and sealants as well as common testing
procedures for preconstruction and quality control.
Between the
return panels and
the adjoining
column covers, a
seismic joint is
installed to allow
for the two
exterior
elevations of the
building to move
independently
during an
earthquake.
A final
installation might
look like the
adjoining picture,
where the vertical
seismic joint is
visible between
the return panel
on the left and the
half-width
column cover on
the right.
During an
earthquake in the
direction of the
building elevation
containing the
seismic joint, the
in-plane panels
move with the
supporting
structural floor,
the level below.
However, the out-
of-plane panels
tilt into the
building.
Without proper detailing and knowledge of the necessary joint width, pounding between the
panels can occur during major earthquakes. Current building code requirements typically require
a seismic joint of at least two inches to allow for suitable performance during a major
earthquake.
Past Experimental Testing of Cladding Panels: While limited published data is available
from past testing of cladding systems, some notable testing has been found. Rihal (1989, p. 124)
conducted a full-scale in-plane loading experiment on a full-story solid precast concrete panel.
This panel had push-pull connections at the top with oversized holes of 2.5 inch diameter. Wang
(1986) tested a multistory multi-bay steel frame with various types of cladding in a full-scale,
cyclic loaded test. In this study cladding systems from the United State and Japan were
compared and contrasted. Although the Japanese system appears to have performed better, the
general consensus from the United States was that the system was too complex and expensive
and that the benefit of such a high performance was not worth the added initial cost.
Since 1998, a research initiative at San Jose State has focused on experimental testing to
allow for input of structural analysis software. The primary goal of the research is the
development of input data to be used for Performance Based Earthquake Engineering design of
commercial buildings. This concept uses life-cycle cost analysis to minimize total costs of a
building. The design criteria is to match the potential higher initial investment for higher quality,
more robust new construction versus the lower expected costs for repairs after future
earthquakes. The challenge is that to implement these types of design theories, requires a large
investment of research to define the expected damage and potential repair costs that various
building components may experience.
Component testing of steel
connections used for the support of
concrete panels was the focus of
the first several years of research
at San Jose State as stated before.
Steel connections usually have
specific roles in façade systems,
push-pull connections to resist
forces in and out of the building,
bearing connections that support
the weight of the panel, and lateral
seismic connections that resist
horizontal seismic forces.
The connection shown has all three force resisting components. The leveling bolt in the
assembly supports the gravity load. The coil rod and tube provide a push-pull connection to
resist out-of-plane loading. The 25 mm plate resists in-plane shear.
Current Research Program: Building upon these past studies, three connected projects have
recently been completed to qualitatively and quantitatively measure the damage to precast
cladding systems under seismic loading. All three testing program receive primary funding from
the National Science Foundation. Additional funding has been received from the Charles
Pankow Foundation, San Jose State University and multiple industry partners. There are two
main objectives for the testing: determining the damage that will occur as a function of lateral
drift and/or acceleration and determining the force-deformation relationships for the connections
between the adjoining concrete panels and between the panels and the steel frame.
As the timeline of Table 1 indicates, the current research is the evaluation of three large
scale experimental studies. The Pathways Project at UC Berkeley has completed static loading
of six full-scale experiments under simulated displacement-controlled seismic loading. The
advantage of static testing is that systems can be loaded to near-collapse levels of displacement
to evaluate how the system will perform under extreme overloading. The E~Defense testing and
the UC San Diego testing were both single full-scale, complete-structure specimens which were
loaded using shake table facilities that reproduce the actual recorded motion of the ground during
past earthquakes. The advantage of shake table testing is that the true acceleration and dynamic
environment can be developed in three-dimensional space. Through these coordinated test
programs, a wealth of data has been collected, including quantitative data about displacements,
accelerations and forces as well as qualitative data in video, photographic and experiential
formats.
All three test programs had similar features. The following photos show the overall form
and size of each of the test layouts. Many of the features of the test program were defined by
other aspects of the overall test program. All cladding systems were based upon current US
precast cladding design practice and were built and tested under laboratory conditions. Critical
aspects of all the cladding systems were built by industry personnel to ensure the quality and
condition matched those seen in actual commercial building construction. All panels used 5000
psi concrete and Grade 60 reinforcing steel. All steel connection components used Grade 50
steel plate and or angle. All welding was E70XX or equivalent.
However, due to the constraints defined by other aspects of each of test programs, each of
the test specimens also had unique characteristics, particularly in the detailing of the steel
connections that connect the panels to the structural frame. The Pathways project had a primary
focus on cladding and hence most of the design characteristics of the experiment were
determined by the cladding. Using the SAC 9-story LA Building as a preliminary schematic, a
cladding system using spandrel panels and supported column covers were tested. The test
program was six individual tests with each test containing three panels. The E~Defense (TIPS)
project used an existing steel frame test structure and included two panels on a frame that would
be shaken in all three directions. The UC San Diego (Structural/Non-structural) project used a
project-specific concrete frame that was completely enclosed by concrete facade on the top two
floors and was shaken in a single longitudinal direction.
Results from Testing: The primary findings to date have been that well designed and fabricated
precast panel systems perform very well during seismic loading. The only significant damage
observed in the testing has been as a result of lateral displacements far above the design
displacements. Various panels have been loaded in both static and dynamic protocols and the
damage observed at displacements below the design displacement have been minimal. All the
cladding systems represent modern design features by American fabricators in seismic zones.
Damage was observed during the static loading tests when displacements above the design
displacement were applied. The photos below show representative photographs of some
common damage patterns in the testing completed. The most common post-design-displacement
damage has been the cracking of the concrete panels due to flexure, particularly in the flat half-
width panel. This damage has coincided with cracking and loosening of the steel embeds in the
interior face of the panels, particularly at the base of the panels. The slotted connection at the top
of the panel were made with the nut on the slotted rod being placed finger tight, no wrench was
used to tighten or install this nut. The threads of the bolt were then sealed to prevent loosening
of the nut. With this finger-tight nut, the slotted connections performed as intended, with
minimal resistance to movement. This was observed in both the horizontal slotted connections
of the Pathways project and the vertically slotted connections of the E~Defense project.
The resulting model allows for both linear and nonlinear response to both static pushover
types of loading as well as the cyclic loading expected during an earthquake.
Individual connections can be
modeled in the simulation once the
force-deflection behavior is
defined. The experimental output
of the adjoining graph shows how
slotted connections respond to
cyclic experimental loading. The
connection slides very smoothly
over the length of the slot and then
resists significant force once the
slot length is exceeded.
Window Performance. Window glass was a façade material also tested during the
project. Two of the specimens at UC Berkeley were built with complete façade systems
including both the precast concrete cladding and inset window units.
Experiment 3 of the Pathways
Project contained a full window
inset between the two flat column
cover panels. During testing, the
loading moved the blue beam
horizontally, in the plane of the
window glazing.
As the lateral movement increased,
the main damage to the window
units were fracture of the F-Clip
used to connect the window frame
to the structure (yellow plate above
in the adjoining photo).
Surprisingly, none of the window
panes in any experiment every
cracked, even when story
displacements of four inches were
applied to the window unit.
Instability of the entire window unit after severe loading was a concern. During the original
testing, the frames often accommodated lateral building drift by fracture of the F-Clip connecting
the window unit to the building. Since the static loading was in the plane of the window, the
frame remained standing, due to cantilever action of the vertical aluminum mullions from the
base of the unit. During an earthquake, out-of-plane acceleration after these F-Clips fracture
may cause the entire assembly to fall from the building. This potential is a concern, but it was
not seen during the in-plane loading tests.
Adaptive Reuse. A payload research task is the study of potential adaptation of damaged
cladding panels for alternative use. Panels can be unusable for three common reasons, either
they are mis-manufactured originally or they are damaged during extreme loading or they are
removed from a structure due to renovation and changes to the main structure. For any of these
reasons, the owner is presented with a significant cost of demolition and disposal. With current
requirements about the disposal of concrete with embedded steel rebar, the cost of discarding
these damaged panels has become significant. Recycling of concrete via crushing has become a
common alternative, particularly for panels that have been miscast and are fabrication waste.
However, recycling does have limits due to energy requirements and the need for matching
aggregate with building design limitations.
To explore alternatives, one case study was implemented as to a potential reuse of the panels
in their entirety. The case study was to take panels damaged during the experimental testing,
perform moderate trimming of heavily damaged portions, and installing the panels at a local non-
profit undergoing renovation work. The goal was to replace the planned slab-on-grade hardscape
with a layout of panels closely spaced.
The concept for the Adaptive
Reuse case study was that if
panels were damaged during an
earthquake or blast event, the
damaged items would be
‘harvested’ by removing the
panels from the building.
Work Forthcoming: As the research initiative continues, work will focus primarily on four
areas: data reduction, computer modeling, component testing, and research dissemination. Data
reduction is expected to be a significant task as data has been collected in both quantitative and
qualitative formats.
Building upon the full scale system tests, San Jose State is expanding upon the
connection component tests. The current work is to expand the knowledge about the effect of
lateral load on the coil rod push-pull connections. The full scale experiments have shown that
correct detailing of these connections shows high potential to use their ductile capacity to use
flexural yielding as a means of reducing the size of the seismic joint. Hence a series of
experiments are using various levels of displacement to define the fracture limit state of the
connections.
The test setup is used to apply lateral force to the coil rod by extending and retracting the
actuator. By applying constant displacement cycles of loading, a predictive formula is to be
developed to allow precast fabrication engineers the ability to predict suitable performance for
the rods.
Fracture consistently occurs at a region of concentrated yielding at one end of the rod. While a
fracture of the rod would result in potential collapse of a panel and is thus unacceptable,
predicting the fracture limit state would allow the precast fabricator to make suitable decisions
about the proper detailing of connections.
As experimental data is processed and combined with analytical studies, dissemination of
research findings is continual. Project webpages and online repositories of data allow for online
access and rapid dispersal. Webinars are in development as well as design procedure documents
for fabricator engineering staff.
Conclusions. The primary conclusion has been that current precast concrete facade systems
designed for seismic motion perform very well when displaced up to the level that was expected
during design. Additional conclusions are:
1. When displaced significantly beyond the design displacement, the prevalent form of
damage seen was cracking of the concrete, both due to flexural of the panel and around
connection embeds.
2. Modern window systems performed very well during testing. Damage was seen to the
aluminum connection between the structure and the window frame. Glass panes slid in
the gaskets and frames distorted to a level that left visible gaps in the assembly, but glass
breakage only occurred after extremely large levels of story drift.
3. Damage of panels seems closely related to the size of the seismic joint. Modern designs
usually contain very large width of joints. Concern about the successful performance of
older systems with narrow joints does appear to be a concern.
4. Detailing of cladding connections to allow ductile yielding during major earthquakes has
the potential to develop suitable performance of commercial buildings with narrower
seismic joints.
5. Computer simulation of nonlinear behavior of cladding is suitable for flexible rod
systems where yielding controls the behavior but modeling slotted connections using
traditional commercially available software is limited in capability.
6. Adaptive reuse shows potential for replacing the tradition options of disposal or crushing
for recycling. However the case study showed that using the panels in lieu of a slab-on-
grade hardscape is challenging due to fit up and placement issues. Higher potential would
be the use of panels as independent entities, where a large structural component is
needed, such as a tank foundation, a set if spaced pavers, or riprap for erosion protection.
Acknowledgements: This material is based upon work supported by the National Science
Foundation under Grant Nos. CMMI-0619157, CMMI-1113275, and CMMI-0936505.
Additional funding for testing has been provided by the Charles Pankow Foundation and the
Precast Concrete Institute. Industry engineering personnel contributed many hours of work to
complete the design, construction, experimental program and interpretation of the experimental
results. Any opinions, findings, and conclusions or recommendations expressed in this material
are those of the authors and do not necessarily reflect the views of the National Science
Foundation, other sponsors, or assisting industry groups.
Related References:
Full-Scale Structural and Nonstructural Building System Performance during Earthquakes &
Post-Earthquake Fire. http://bncs.ucsd.edu/index.html. Last accessed: April 11, 2012.
Bachman, R. E., Hamburger, R.O., Comartin, C. D., Rojahn, C., and Whittaker, A. S., 2003,
“ATC-58 Framework for Performance-Based Design of Nonstructural Components”,
Proceedings of ATC-29-2 Seminar on Seismic Design, Performance, and Retrofit of
Nonstructural Components in Critical Facilities, ATC-29-2 Report, Applied Technology
Council, Redwood City, California, pages 49-61.
McMullin, Kurt M. (2013). “Defining fracture limit states of coil rods for precast concrete
cladding panels.” Proceedings, 2013 PCI Convention and National Bridge Conference, Precast
Concrete Institute. Grapevine, TX. Sept. 21-24.
McMullin, Kurt M. (2012). “Full-scale dynamic testing of precast concrete cladding panels.”
Proceedings, 2012 PCI Convention and National Bridge Conference, Precast Concrete Institute.
Nashville, TN. September 29-October 1.
McMullin, Kurt M. and Rai, Kathi (2012). “Adaptive reuse of precast concrete building
cladding panels.” Proceedings, 2012 PCI Convention and National Bridge Conference, Precast
Concrete Institute. Nashville, TN. September 29-October 1.
McMullin, Kurt M., Ortiz, Maggie, Patel, Lokesh, Yarra, Siddaiah, Kishimoto, Tatsuo, Stewart,
Caleb, and Steed, Bob (2012). “Response of Exterior Precast Concrete Cladding Panels in
NEES-TIPS/NEES-GC/E~Defense Test on a Full-Scale 5-story Building.” Proceedings, 2012
Structures Congress, ASCE. Chicago, IL. March 29-31.
McMullin, Kurt M. and Ortiz, Maggie (2011). “Experimental testing of precast concrete
cladding for building facade systems.” Proceedings, 2011 PCI Convention and National Bridge
Conference, Precast Concrete Institute. Salt Lake City, UT. October 22-25.
McMullin, K. M., Wong, Y., and Kwong, A. (2003). “Damage thresholds and performance
levels of precast cladding systems.” Proceedings of Seminar on Seismic Design, Repair and
Retrofit of Nonstructural Components on Critical Facilities, Applied Technology Council
Project 29-2, Los Angeles, April.
McMullin, K. M., Choi, C. Chan, K., and Kwong, A. (2003). “Performance engineering of
precast cladding systems.” 2003 Structures Congress and Exposition. Seattle, WA.
McMullin, K. M., Wong, Y., Choi, C. Chan, K., and Kwong, A. (2003). “Performance
Thresholds of Precast Cladding Systems.” ATC 29-2 Seminar on Seismic Design, Performance
and Retrofit of Non-structural Components in Critical Facilities. Santa Ana, CA. Oct.
Merrick, D., McMullin, K. M., Hildebrand, M. and Kwong, A. (2003). “In-Situ Vibration
Characteristics of Precast Cladding Panels.” 2003 Structures Congress and Exposition. Seattle,
WA.
Rihal, Satwant S. (1989). “Earthquake resistance and behavior of architectural precast cladding
and connections.” Proceedings: Architectural Precast Concrete Cladding – It’s Contribution to
Lateral Resistance of Buildings. PCI, November 8-9, Chicago. Pages 110 to 140.