STRUCTURE magazine

23
practical knowledge beyond
the textbook
STRUCTURAL
PRACTICES
Nat Tocci, P.E. is the owner of Tocci
Engineering PLLC in New York
City and can be reached at
[email protected].
Sanya Levi is an engineer at Arup
New York and can be reached at
[email protected].
By Nat Tocci, P.E. and Sanya Levi
Basement Modeling in
Tall Buildings
Backstay Effect
O
ne of the least understood aspects
of modeling building structures is
dealing with at- and below-grade
components. Tis includes soil-
structure interaction, but also the question of
which below-grade structural elements should
be included in a lateral model and what is an
accurate representation of the base conditions.
Te focus of this article is what is most commonly
referred to as the backstay efect. Traditionally, lat-
eral systems have been viewed as simple cantilever
beams fxed at the base. While this analogy is
reasonable for the above-grade structure, a more
accurate analogy would also include the efects
of the below-grade structure, which behaves like
a backspan to the cantilever. In this analogy, the
lateral system is viewed as a beam overhanging
one support, where that support is created by the
at-grade diaphragm and foundation walls.
Te backstay efect is not limited to restraint at
the grade level. Backstay efects are also seen at set-
backs with changes to the lateral system, the most
common example being lower level podiums. Tey
are often very large in plan and introduce new lat-
eral elements, and are therefore signifcantly stifer
than the set-back structure above. Backstay efects
are also impacted by multiple basement levels. For
simplicity of explanation, this article will focus on
the most common example which is the efect of
the ground foor diaphragm in contributing to
backstay efects. Te concepts can be extended to
all conditions where backstay efects occur.
Backstay Effect
Backstay efects are most noticeable in buildings
with discrete lateral systems, such as shear walls,
as opposed to distributed lateral systems. Building
height is also a major factor in the magnitude of
the backstay efects. For the purposes of illustra-
tion, this article focuses on a high rise shear wall
building with a single basement.
For a typical building with one or more below
grade levels, the perimeter basement walls create
a very large and laterally
stif box. Te ground foor
diaphragm engages this
box and integrates it into
the lateral system. Sticking
with the beam analogy,
the result is an efectively larger beam section
below grade. Tis results in shedding of lateral
load from the main lateral force resisting system
(LFRS) to the basement walls. Overturning and
shear are shared between the perimeter walls and
core rather than isolated beneath the building
core. Conceptually this is fairly straightforward.
Te complexity arises in properly modeling the
change in section, and capturing an accurate dis-
tribution of internal forces and external reactions.
Te degree to which lateral loads are transferred
into the foundation perimeter is dependent on many
variables, many of which there is limited certainty
about, as they are not specifed or controlled in
a typical project. It is therefore fair to ask if it is
more conservative to simply ignore any backstay
efects and model the building core as an isolated
element. However, it can be shown that in many
cases the backstay efect will create higher demands
in some structural elements, in particular shear in
the main LFRS below grade as well as the backstay
diaphragms, and therefore cannot be ignored.
Figure 1 is a stick diagram presenting some of
the possible options for modeling the base condi-
tions of a core wall building. Te building is of
height H with a basement of height B. Te most
traditional model, a simple cantilever, is shown
in Figure 1a. It is clear that the maximum shear
is V = F. Te extreme case of the backstay efect
is shown in Figure 1b. In Figure 1b the ground
foor diaphragm and perimeter foundation are
very stif and are therefore modeled as a pin.
Statics shows that the maximum shear in the core
now occurs below grade with V = 3H/2B F. Te
Figure 1: Modeling options for base condition.
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
rig
h
t
STRUCTURE magazine June 2012
24
overall base shear has not changed, but the
backstay efect may create conditions with
much higher demands than anticipated in
certain elements. It can also be shown that
the base overturning moment in the core
has been reduced and redistributed to the
perimeter foundation walls.
Although Figure 1b shows dramatic increases
in shear, this is overly conservative for most
conditions. Te true restraint at the ground
foor is far from rigid and may range from
very stif to almost non-existent. A more real-
istic model is one in which the ground foor
restraint is modeled as a spring, producing
results somewhere between Figures 1a and
1b. Figure 1c shows this option.
Te complexity of an accurate model lies in
the fact that the spring in Figure 1c represents
the cumulative stifness of numerous elements
in the building structure and supporting soil. A
partial list of elements represented by the ground
foor spring would include: diaphragm to core
connection, diaphragm stifness, diaphragm
to basement wall connection, basement wall
stifness, foundation stifness, and passive soil
resistance against the basement wall.
Ground foor diaphragms are often thick
concrete plates with high relative stifness.
However, this stifness may be reduced by
cracking, bond slip, and discontinuities such
as large openings or slab elevation changes.
In addition to the stifness of the diaphragm
itself, the connections at each end must be
considered for their ability to transfer the
backstay shears. Te same can be said for the
basement walls which will have varying stif-
ness dependent on the same factors.
Te overall stifness of the diaphragm and
basement wall system is also afected by the
supporting foundation elements. Diferences
in relative stifness between core and perim-
eter wall soil support conditions may magnify
or lessen backstay efects.
Te passive resistance provided by the soil on
the basement wall face in the direction of force
should also be considered. Tis component is
typically small relative to the other elements
and may possibly be neglected in many cases.
In addition, this force is present only in the
compression cycle of loading and should be
modeled as such.
Clearly there are many parameters to consider.
In most cases, the best that can be done is to
model all contributing elements and make an
educated estimate of the element stifnesses.
Te number of possibilities is too numerous
for a prescriptive approach that will work for
all buildings, which is perhaps why there is
little literature on the subject. Most building
codes provide requirements for loading and
design of structural elements, but rarely provide
detailed guidance on modeling procedures. A
very good resource for an in depth discussion
of the backstay efect and recommendations for
modeling is Modeling and Acceptance Criteria
for Seismic Design and Analysis of Tall Buildings,
PEER/ATC 72-1, which is available as a free
download from the PEER (Pacifc Earthquake
Engineering Institute) website.
Te backstay concept is more familiar to
engineers working in high seismic regions
and has had less attention in other regions.
Te concepts, however, are applicable for both
wind and seismic loading.
Modeling
A reasonable frst step may be to assess whether
the backstay efect is a consideration for the
building under investigation. A quick study of
the parameters that create the backstay efect
may quickly rule out the need for a more in
depth analysis. Te building system or con-
fguration may also determine the potential
for backstay efects.
For buildings where backstay efects need
to be considered, it will most likely be nec-
essary to consider multiple scenarios. Both
an overestimation and underestimation of
backstay efects can produce underestimates
of demand. For example, overestimating
backstay restraint may underestimate the
overturning demand at the base of the
main LFRS. Te common approach is
to consider reasonable extremes for both
conditions and design each element for
the bounding condition. Tis is typically
referred to as bracketing.
Te backstay diaphragms must be mod-
eled as semi-rigid elements. Semi-rigid
elements have stifness taken from the
material and geometric properties of
the slab. Any large discontinuities in the
slabs should be modeled, and a mesh size
should be chosen that produces accurate
results. To account for cracking, bond slip,
interface slip, and other unknowns, the stifness
of the slab should be reduced for both shear
(GAv) and fexural (EI) stifness. Similar mod-
eling guidelines and stifness reductions should
also be applied to basement wall elements.
Soil stifness should also be bracketed, typi-
cally starting with recommendations provided
by the project geotechnical engineer. Te sup-
porting stifness under all elements should be
taken at an upper and lower bound, and passive
resistance provided against the perpendicular
wall should also be bracketed if it is modeled.
PEER/ATC 72-1 Table A-2 and Table A-3
provide recommended upper and lower
bounds for bracketing the stifness of the
above elements. PEER/ATC 72-1 also recom-
mends that elements outside of the backstay
infuence (primarily tower elements) need not
be bracketed and should be modeled with
the same assumptions used for their design.
Since these recommendations are intended for
buildings in high seismic regions, it may be
appropriate to adjust the recommendations
for wind controlled design to account for
primarily elastic behavior.
Due to the complexity of capturing backstay
efects in the analysis, it may be desired to elimi-
nate the phenomenon in the actual building.
Tis can be accomplished by isolating the LFRS
from the foundation elements by providing lat-
eral joint at the backstay diaphragms. Typically
this is done by providing a corbel or similar
detail at the diaphragm to shear wall interface.
Conclusion
Ignoring the contribution of at- and below-
grade structural elements in lateral models may
underestimate demands in key elements. A
quick initial study may be enough to determine
if a more in-depth model, which includes back-
stay elements, is justifed. If backstay efects
are included in the model, current practice
is to bracket stifness parameters and design
for a bounding solution. Unfortunately, this
approach results in overdesign of at least some
members. As knowledge of the topic increases,
bracketing parameters will be refned and
increase the efciency of designs.
Typical concrete core building confguration.
The easiest to use software for calculating
wind, seismic, snow and other loadings for
IBC, ASCE7, and all state codes based on
these codes ($195.00).
Tilt-up Concrete Wall Panels ($95.00).
Floor Vibration for Steel Beams and Joists
($100.00).
Concrete beams with torsion ($45.00).
Demos at: www.struware.com
A
D
V
E
R
T
I
S
E
M
E
N
T

-
F
o
r

A
d
v
e
r
t
i
s
e
r

I
n
f
o
r
m
a
t
i
o
n
,

v
i
s
i
t

w
w
w
.
S
T
R
U
C
T
U
R
E
m
a
g
.
o
r
g
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
rig
h
t