The Effect of Foundation Flexibility and Structural Strength On Response Reduction Factor of RC Frame Structures

THE STRUCTURAL DESIGN OF TALL AND SPECIAL BUILDINGS
Struct. Design Tall Spec. Build. (2010)

Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/tal.654
The effect of foundation flexibility and structural strength on

response reduction factor of RC frame structures
Amir M. Halabian*,† and Masoomeh Erfani

Faculty of Civil Engineering, Isfahan University of Technology, Isfahan, Iran
SUMMARY
Current force-based design procedure adopted by most seismic design codes allows the seismic design of
building structures to be based on static or dynamic analyses of elastic models of the structure using elastic
design spectra. The codes anticipate that structures will undergo inelastic deformations under strong seismic
events; therefore, such inelastic behaviour is usually incorporated into the design by dividing the elastic
spectra by a factor, R, which reduces the spectrum from its original elastic demand level to a design level.
The most important factors determining response reduction factors are the structural ductility and over-
strength capacity. For a structure supporting on flexible foundation, as Soil Structure Interaction (SSI)
extends the elastic period and increases damping of the structure-foundation elastic system, the structural
ductility could also be affected by frequency-dependent foundation-soil compliances. For inelastic systems
supporting on flexible foundations, the inelastic spectra ordinates are greater than for elastic systems when
presented in terms of flexible-base structure’s period. This implies that the reduction factors, which are
currently not affected by the SSI effect, could be altered; therefore, the objective of this research is to
evaluate the significance of foundation flexibility on force reduction factors of RC frame structures. In this
research, by developing some generic RC frame models supporting on flexible foundations, effects of stiff-
ness and strength of the structure on force reduction factors are evaluated for different relative stiffnesses
between the structure and the supporting soil. Using a set of artificial earthquake records, repeated linear
and nonlinear analyses were performed by gradually increasing the intensity of acceleration time histories
to a level, where first yielding of steel in linear analysis and a level in which collapse of the structure in
nonlinear analysis are observed. The difference between inelastic and elastic resistance in terms of displace-
ment ductility factors has been quantified. The results indicated that the foundation flexibility could signifi-
cantly change the response reduction factors of the system and neglecting this phenomenon may lead to
erroneous conclusions in the prediction of seismic performance of flexibly supported RC frame structures.
Copyright © 2010 John Wiley & Sons, Ltd.
1. INTRODUCTION
In the current force-based design procedures adopted by most seismic design codes, the designer is
allowed to utilize the ductile capacity of the structure; and therefore, the seismic design of building
structures is based on static or dynamic analyses of elastic models of the structure using elastic design
spectra. Force-based design procedures are likely to remain as the primary seismic design method for
some time since performance-base design methods are still in the development phase. In these pro-
cedures, the elastic design strengths are substantially reduced on the provision of adequate ductility
capacity of the structure, to sustain a targeted amount of plastic deformation under a maximum cred-
ible earthquake condition. The factor, R, which reduces the spectrum from its original elastic demand
level to a design level, is related to the overall performance of different types of buildings during
actual earthquakes. Despite the fact that the reduction factor, R, which serves the same function in all
seismic codes, is called behaviour factor (q) in EC8 and response modification factor (R) in National
* Correspondence to: Amir M. Halabian, Faculty of Civil Engineering, Isfahan University of Technology, Isfahan
84156-83111, Iran
†
E-mail: mahdi@cc.iut.ac.ir
Copyright © 2010 John Wiley & Sons, Ltd.

A. M. HALABIAN AND M. ERFANI
Earthquake Hazards Reduction Program (NEHRP), it is termed the structural quality factor or the
system performance factor following recommendations of the SEAOC Committee. In the current
study, the term response reduction factor is adopted since it offers a clearer indication of the nature
of this factor, which plays a paramount role in seismic design.
Differences in the numerical values of the response reduction factors specified in different seismic
codes for the same type of structure can be quite noticeable, reflecting to this fact that the values
could be based on judgment, experience and observed ductile performance of buildings during past
earthquakes. The key role of apparent inconsistencies in the way these response reduction factors are
defined in different seismic codes has highlighted that the investigation of different parameters affect-
ing inelastic spectra and the ensuing response reduction factors are needed. As it was mentioned, it
is almost generally accepted in various seismic codes that the response reduction factor of structures
may be defined as the ratio of their elastic strength demand to their inelastic strength demand within
the selected ductility capacity limits. Accordingly, the value of the response reduction factor is known
to mainly depend on the inherent ductility of the structure, on the overstrength of individual members
and on the (effective) damping of the structure, assuming the structure is supported by fixed-base
foundation. In fact, seismic analyses of buildings are often based on the assumption that the founda-
tion flexibility has no effect on foundation–structure interacting forces. The first studies of SSI for
elastic systems showed that soil structure interaction phenomenon could influence the dynamic char-
acteristics of structures, resulting in an increase to the fundamental natural period as well as the
inherent structural damping. Veletsos and Verbic (1974) investigated the transient elastoplastic
response of a flexible-supported structure supporting on an elastic half-space, suggesting that yielding
decreases the effect of SSI on structures. Whereas in studies conducted by Avilés and Pérez-Rocha
(2003) it was shown that SSI effects for inelastic systems with different order could be as important
as for elastic systems.
The current SSI provisions based on linear structural models may not be directly applicable to
seismic design of typical buildings, expected to deform considerably beyond the yield limit
during severe earthquakes. So that, Crouse (2002) and Stewart et al. (2003) recognized that the
SSI provisions in the Applied Technology Council (ATC, 1995) and NEHRP codes have a significant
shortcoming expressing no link between the response reduction factors and the effects of foundation
flexibility. Investigations of the SSI on the response reduction factors of structures are scarce. Avilés
and Pérez-Rocha (2005) evaluated the effect of SSI on strength-reduction, Rμ, and displacement
reduction, Cμ, factors for an Single Degree of Freedom System (SDOF) elastoplastic structure sub-
jected to Michoacon earthquake (1985). The result of the research noted that these factors could be
altered by the SSI phenomenon depending on the supporting soil-structure stiffness ratio. Halabian
and Kabiri (2006) evaluated the effect of foundation flexibility on the ductility of RC stake-like
structures. The results showed that for this type of structures, the SSI could decrease the ductility
demand and consequently response reduction factor. The present study is aimed to evaluate the R
factors of RC frame structures considering the foundation flexibility effect using refined definition
of response reduction factors employed by the most modern seismic codes.
2. EVALUATION OF RESPONSE REDUCTION FACTOR
Having this fact that contributing factors affecting response reduction factors such as ductility, struc-
tural overstrength and structure’s redundancy have long been recognized and assuming no supple-
mental damping devices, an appropriate definition of the response reduction factor can be expressed
in the form of
R = Rμ Ω d RR (1)
where Rμ is the ductility reduction factor, Ωd is the overstrength factor and RR is the redundancy
factor which quantifies the improved reliability of seismic framing systems that use multiple lines
Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. (2010)
DOI: 10.1002/tal
EFFECT OF SSI ON RESPONSE REDUCTION FACTOR
of vertical seismic framing in each principle direction of a building. Since the overstrength factor
implicitly accounts for redundancy through redistribution of actions, the overstrength and redundancy
parameters can be considered as one factor and therefore, the response reduction factor is simply
given as:
R = Rμ Ω d (2)
It is understood that seismic nonlinear responses and then the ductility demand of building structures
resting on flexible foundations could be changed, due to the foundation flexibility (Halabian
and Emami, 2009). Since the concept of response reduction factor expressed in Equation (1) does
not explicitly take into account the foundation flexibility, the Equation (2) can be furthermore
modified as:
R = Rμ Ω d RS (3)
in which RS in this study is so-called as the foundation factor that depends on the foundation–structure
stiffness ratio and the structural strength as well. If a proper calibration of response reduction factors
is to be made, each of the major components contributing to it should be precisely investigated. In
the following, the ductility-dependent product of Rμ and RS is evaluated on the basis of some repeated
nonlinear analyses of flexible-base structures for selected artificial earthquake records.
There are some different methods based on force–displacement relationship resulting from inelastic
analyses of structures to evaluate the response reduction factors according to Equation (3). However,
the current study is not intended to review and discuss all these approaches. In the following, taking
into account the ground motion dependence of the response reduction factor, a more rigorous approach
for proper calibration of these factors on the basis of repeated nonlinear analyses is described. The
factor in various modern seismic codes reduces the elastic base shear (Ve), obtained from the elastic
acceleration spectrum (Sa)in in the first natural period of the structure, to the design base shear level
(Vd), evaluated from spectrum used in design (Sa)in, again corresponding to the first natural period of
the structure. Thus,
Rcode = (Sa ) (Sa )in

el
(4)
Collapse is normally anticipated under the effect of an earthquake with a spectrum higher than the
elastic spectrum; therefore, for a particular structure under a specific accelerogram, the following
formula evaluates an ultimate value of the response reduction factor:
Rc,dy = (Sa )c (Sa )

el in
(5)
where the subscripts ‘c’ refer to collapse and ‘dy’ refers to the yield level assumed in design. Having
this fact that structure is mainly designed for forces consistent with its yield limit state, Elnashai and
Broderick (1996) and Wafi and Elnashai (2002) used a definition that utilizes the spectral acceleration
causing actual yield in the denominator, as given in the following equation:
Rc, ay = (Sa )c (Sa )y

el el
(6)
In the above equation, the subscript ‘ay’ refers to the actual yield. Assuming the response spectra
of the yield and collapse earthquakes have constant dynamic amplification (the ratio of the peak
ground acceleration to the peak response acceleration), Equation (6) can be rewritten as:
Rc, ay = a g(collapse) a g(actual yield) (7)
DOI: 10.1002/tal
(Sa) elc
Sa
Collapse earthquake
(Sa)ely R`c,ay = R c,ay ⋅ Ω d

Yield earthquake
R c,ay = (S a ) elc /(S a ) ely
Design
Figure 1. Evaluation of the force reduction factor by definition of Rc,ay and Ωd.
In the above equation, ag(collapse) is the peak ground accelerations of the collapse and ag(actual yield)
is the Peak Ground Acceleration (PGA) corresponding to the first yielding in the system. In
fact, structures designed using modern seismic codes usually exhibit some considerable level of
overstrength that leads to significant differences between the PGA causing the first global yield
(ag(actual yield)) and the yield intensity implied by the design (ag(design yield) = design PGA/Rcode) (Figure 1).
Therefore, Equation (7) should be modified adding an overstrength factor, Ωd, as,
Rc′,ay = Rc,ay ⋅ Ωd = (a g(collapse ) a g(actual yield ) ) ⋅ Ω d (8)
Overstrength factor accounts for the reverse strength between the actual yield and design
levels and can be employed to reduce the seismic forces used in the design, hence leading to more
economical structures. The main sources of overstrength include (Humar and Rahgozar, 1996): the
difference between the actual and the design material strength; conservation of the design procedure;
load factors and multiple load cases; serviceability limit state provisions, participation of nonstructural
elements and structure’s redundancy.
Based on the approach described in this study to evaluate the R factors, incremental nonlinear
dynamic analysis, a number of response criteria could define the yield and collapse limit states. The
first yielding can be defined as the local yield when the strain in the main tensile reinforcement
exceeds the yield strain of steel or the global yield limit state which is characterized as the yield
displacement of the elasto-plastic idealization of the real system. Like yielding criteria, local and
global failure criteria can be employed to evaluate the PGA corresponding the structural collapse.
The ultimate curvature controlled by the maximum compression strain at the extreme fibre of the
confined concrete or the tension strains of reinforcement and the maximum inter-storey drift or forma-
tion of a sidesway mechanism could be referred as the member level (local) and the structure level
(global) collapse criteria. The assumptions made in developing RC generic frames used in this study
can shed light on the suitability of the local yield criterion, while it was decided to use the global
mechanism along with the ultimate curvature criterion for column members representing the struc-
tural collapse state.
To evaluate the peak ground accelerations causing collapse and first yielding, the incremental
nonlinear dynamic analysis is performed by progressively scaling and applying each of the employed
DOI: 10.1002/tal
set of earthquake records, starting from a relative low intensity, and terminating with the intensity at
which all yield and collapse definitions are achieved. This method allows evaluating the performance
of the structure at different levels of excitations. Hence, the peak ground accelerations causing yield
and collapse can be identified according to the performance criteria adopted for structural elements.
It is mentionable, since a generally applicable and precise estimation of overstrength is difficult due
to many factors contributing to it; therefore, in this study, the overstrength factor is not taken into
account in the evaluation process of response reduction factor and therefore, the effect of foundation
flexibility is then only considered on Rc,ay.
3. NUMERICAL MODELS
Since a large number of parameters generally influence the seismic demands of structures, it was
shown that generic models could be an appropriate approach for assessing the influence of these
parameters in seismic demands of RC frame buildings. In a general view, generic models adopted in
the current study consist of 2D single-bay RC moment-resisting frames reflecting different stiffnesses
and storey strengths. To take into account the structure’s height effect, three types of generic structures
representing short, medium and high-rise RC frames were developed (Figure 2). In all three types of
generic structures, using the same moments of inertia for beams and columns in every storey, it was
assumed that the storey height for all storeys is the same and half of the bay width. The stiffness of
generic structures are tuned so that the structure’s deflected shape under a given design load pattern
become a straight line (Figure 3). The lateral load pattern is selected based on the SRSS modal super-
position method. Employing UBC’97, the SRSS lateral load distribution is obtained using storey shear
5 − story 10 − story 20 − story

T = 5.0 T = 0.1 T = 0.2
T = 7.0 T = 5.1 T = 0.3
Figure 2. Generic structures.
DOI: 10.1002/tal
Figure 3. Plastic hinge mechanism for the generic structures.
forces calculated from the SRSS combination of modal responses throughout the linear response
spectrum analysis. For RC frame generic frame structures developed for this study, the SRSS storey
shear forces distributions and corresponding lateral load patterns are shown in Figure 4(a–c). In the
generic structure, inelastic deformations are permitted only at the ends of the beam in each storey and
at the base of the columns (Figure 3). Thus, the basic plastic hinge mechanism under lateral loads
represents structures complying with the weak beam–strong column requirement. Assuming that the
overall mass matrix of the generic model is constant, a generic structure with specific fundamental
period is produced by changing all element stiffnesses in every storey, simultaneously. In this research,
the fundamental elastic period of the generic frame structures are assumed to be T = 0.5 and 0.70 s,
representing short RC frames; T = 1 and 1.5 s for medium-rise structures; and T = 2 and 3 s present-
ing high-rise RC frames (Figure 2). The cross-section dimensions for generic frames at different floors
are given in Table 1.
The seismic design of frame structures is performed according to back-calculated base shear
strength and a design lateral load distribution. The base shear strength is varied in relation to specific
purposes of the analysis. Given this quantity, the individual storey shear strengths are arranged to the
storey shear forces obtained from the design load pattern. Generally, the base shear strength Vbase is
estimated according to:
Vbase = C .W (9)
In this equation, W is the effective weight of structure and C represents the base shear strength
factor. Assuming different values for C, every generic frame includes a set of idealized models with
the same stiffness but different strengths against lateral load. In this study, to examine the strength
effect on response reduction factors, the generic frames with the same stiffness were designed for two
strength levels against lateral loading, based on ACI 318-05 code. Thus, 2D generic frames used in
this study include three groups as shown in Figure 2:
(1) Five-storey generic frames representing short RC frames having fundamental natural periods
of 0.5 and 0.7 s designed under practical base shear strengths in this range of stiffness, C = 0.1
and 0.2.
(2) Ten-storey generic frames representing medium-rise RC frames having fundamental natural
periods of 1 and 1.5 s designed under practical base shear strengths in this range of stiffness,
C = 0.05 and 0.1.
(3) Twenty-storey generic frames representing high-rise RC frames having fundamental natural
periods of 2 and 3 s designed under practical base shear strengths in this range of stiffness, C
= 0.03 and 0.08.
DOI: 10.1002/tal
SRSS Story Lateral Load Pattern
Story Level
4
3
2
1
0 0.2 0.4 0.6 0.8 1

F(i)/F(top)
(a)
10
9
8
Story Level
7
6
5
4
3
2
1
0 0.2 0.4 0.6 0.8 1
F(i)/F(top)
(b)
20
19
18
17
16
15
14
13
Story Level
12
11
10
9
8
7
6
5
4
3
2
1
0.0 0.2 0.4 0.6 0.8 1.0
F(i)/F(top)
(c)
Figure 4. The SRSS storey shear forces distributions for generic structures: (a) 5-storey;
(b) 10-storey; (c) 20-storey.
DOI: 10.1002/tal
Table 1. Cross-section dimensions for generic RC frame structures.
Dimension of beam and column sections (mm)

High-rise structures Mid-rise structures Low-rise structures
Floor T=2s T=3s T=1s T = 1.5 s T = 0.5 s T = 0.7 s
20th 394 × 394 318 × 318
19th 477 × 477 480 × 480
18th 510 × 510 403 × 403
17th 540 × 540 422 × 422
16th 553 × 553 432 × 432
15th 566 × 566 444 × 444
14th 569 × 569 450 × 450
13th 580 × 580 457 × 457
12th 586 × 586 463 × 463
11th 592 × 592 466 × 466
10th 594 × 594 470 × 470 392 × 392 326 × 326
9th 595 × 595 474 × 474 473 × 473 391 × 391
8th 598 × 598 477 × 477 514 × 514 417 × 417
7th 600 × 600 479 × 479 533 × 533 433 × 433
6th 602 × 602 481 × 481 553 × 553 447 × 447
5th 605 × 605 482 × 482 565 × 565 457 × 457 419 × 419 354 × 354
4th 608 × 608 483 × 483 577 × 577 468 × 468 511 × 511 430 × 430
3rd 610 × 610 485 × 485 578 × 578 469 × 469 564 × 561 437 × 437
2nd 612 × 612 487 × 487 580 × 580 469 × 469 562 × 562 473 × 473
1st 615 × 615 490 × 490 590 × 590 478 × 478 564 × 564 475 × 475
4. MODELLING FOUNDATION FLEXIBILITY
In order to take into account the inertial interaction of the SSI phenomenon, the substructure method
is employed for dynamic analysis of generic structures supporting on flexible foundation. The funda-
mental step in the substructure method is to evaluate the foundation impedance functions. The
dynamic stiffnesses are complex-valued functions depending on the geometry of the foundation and
the characteristic of the soil, which generally can be expressed as:
[ S (ω )] = [ K st ] . [ k ( a0 ) + ia0 c ( a0 )] (10)
in which [S(ω)] and [KST] are the dynamic and static stiffness matrix, respectively. Moreover, k and
c are dimensionless coefficients depending on Poisson’s ratio v and the dimensionless frequency
parameter a0 = ω.r/VS, where r is the equivalent radius of the footing and VS is the shear wave velocity
of soil assumed to be half-space. In this study, assuming shallow foundations of the generic structures,
the corresponding stiffness and damping constants are obtained using the theory due to Veletsos and
Wei (1971) and Veletsos and Verbic (1973). Using this method, the effect of exciting frequency,
foundation type and size, and dynamic properties of the soil layers can also be included in the estima-
tion of foundation impedance functions. The generic frames are assumed to be supported by a homo-
geneous viscoelastic soil having shear wave velocity 50 m/s ≤ Vs ≤ 500 m/s and the Poisson’s ratio
equal to 0.3.
5. NONLINEAR MODELLING OF RC FRAMES AND EQUATIONS OF MOTION
The generic models are idealized as 2D frames using beam and column elements. In this study, with
the purpose to investigate the effect of inelastic behaviour of the RC frame structural elements,
DOI: 10.1002/tal
(a) (b)
Figure 5. Nonlinear behaviour of RC structural elements: (a) backbone curve; (b) hysteretic loop.
Figure 6. Fibre element forces and deformations.
two types of model accounting different types of nonlinear behaviour were adapted. The nonlinear
behaviour of beam elements is modelled by means of moment-curvature theory; and in column
elements, multi-axial spring model based on material stress–strain relation, called ‘fiber model’,
is used to represent the interaction among axial force and bending. In the current study, the M-φ
relations in monotonic loading (backbone curve) are taken tri-linear (Figure 5(a)). The tri-linear
curve is defined by the moment and curvature corresponding to the cracking limit, to the elastic
limit (steel yielding), and finally to the section failure. The element response for cyclic loading is
governed by a set of parameters which are correlated to the strength and stiffness degradation
(Figure 5(b)).
In the Fiber Element Method for modelling nonlinear behaviour of columns, the cross-section of
each column element is subdivided into some spring elements (Figure 6). Each spring is subjected to
axial load, given by the combination of axial force and bending moment acting on the section. Given
the curvature in x and y direction (φx and φy) as well as axial strain (ε0) and assuming end displace-
ment of every spring is constrained to respect the plane section assumption, strain and stress in each
fibre element are calculated as:
DOI: 10.1002/tal
σ i = Ei ε 0 + Ei yi φ x + Ei xi φ y (11)
ε i = ε 0 + yi φ x + xi φ y (12)
where Ei is the modulus of elasticity in ith spring. Therefore, the flexibility matrix at element-end
sections can be expressed as:
⎡ f yyi f xxj f yj ⎤
[ f j ] = ⎢⎢ f xxj f xj ⎥
⎥
(13)
⎣ Sym f0j ⎦
where f yyj , f xxj and f 0j are the flexibilities of the element at the end cross-section corresponding to the
two rotational degrees of freedom and one axial deformation, respectively. By assuming the linear
distributed flexibility, the element flexural flexibility matrix is expressed as:
⎡3 f yy1 + f yy2 − f yy1 − f yy2 3 f xy1 + f xy2 − f xy1 − f xy2 −4 f y1 − 2 f y2 ⎤

⎢ f yy1 + 3 f yy2 − f xy1 − f xy2 f xy1 + 3 f xy2 2 f y1 + 4 f y2 ⎥
L0 ⎢ ⎥
[f]= ⎢ 3 f xx1 + f xx2 − f xx1 − f x2x −4 f x1 − 2 f x2 ⎥ (14)
12
⎢ f xx1 + 3 f xx2 2 f x1 + 4 f x2 ⎥
⎢ sym 6 f01 + 6 f02 ⎥⎦
⎣
The nonlinear governing equation of motion for the generic structure can be written as:
[ M ] {u (t )} + [C ] {u (t )} + [ K ] {u (t )} = [ M ] {I } {ug } (15)
where [M], [C] and [K] are mass and damping and stiffness matrix of soil-structure system. The
Rayliegh’s damping is used to model the structural damping and having taken into account the
damping, [Cv], coming from the foundation impedance functions, the total damping including non-
classical damping is expressed by:
[C ] = am [ M ] + a0 [ K 0 ] + ak [ K ] + [Cv ] (16)
The coefficients am, a0 and ak are damping factors proportional to mass matrix, initial stiffness and
time-varying stiffness matrix computed from the first two modes. To solve the differential equation
of motion (Equation (15)), the time step integration method is adapted in this study. Therefore, at the
time t + Δt, the equation of motion can be expressed by integration:
t + Δt t + Δt t + Δt
[M ] ∫ {du} + ( am [ M ] + a0 [ K 0 ]) ∫ {du} + ∫ ak [ K ]{du}

0 0 0
t + Δt t + Δt t + Δt
(17)
+ ∫ [CV ]{du} + ∫ [ K ]{du} = − [ M ] ∫ dug
0 0 0
A numerical integration procedure known as Newmark’ β method is used to express the differential
relationship of the time function. To satisfy the equilibrium at each time step, the iteration is performed
using the Newton–Raphson scheme within each time step. Therefore, the displacement increment is
expressed as:
DOI: 10.1002/tal
{Δu}t + Δt = Δt {u}t + ⎛ − β ⎞ .Δt 2 {u}t + β .Δt 2 {u}t + Δt

1
(18)
⎝2 ⎠
And then the acceleration and velocity are calculated as:
⎧ Δu 1 1 1
⎪⎪{ }t + Δt = βΔt 2 {Δu}t + Δt − βΔt {u}t − 2β {u}t

⎨ (19)
⎪{Δu} γ γ ⎛ γ ⎞
t + Δt = {Δu}t + Δt − {u}t − ⎜ − 1⎟ Δt {u}t
⎪⎩ βΔt βΔt ⎝ 2β ⎠
Substituting Equation (19) into Equation (17) results to,
⎡⎛ 4 + 2 a ⎞ M + 2 ( a K + C ) + K ⎤ Δu
m [ ] 0[ 0 ] [ ν ] [ ]t ⎥ { }t + Δt =
⎣⎢⎝ Δt 2 Δt ⎠ Δt ⎦
(20)
[ M ] . ⎡⎢⎛⎝ + 2am ⎞⎠ {u}t + 2 {u}t − {Δug }t + Δt ⎤⎥ + 2 (a0 [ K 0 ] + ak [ K ]t + [Cν ]t ) {ut } + {FU }t
4
⎣ Δt ⎦
where {FU}t is an unbalanced force vector; to satisfy the equilibrium at each time step, the iteration
is performed using the Newton–Raphson scheme within each time step.
6. NUMERICAL RESULTS
To get an insight on the effect of the SSI phenomenon on ductility reduction factor of RC frame
structures supporting flexible foundations, a repeated nonlinear time history analysis approach based
on refined definition of the response reduction factor described above was employed. Five artificially
generated records compatible with UBC-97 response spectrum having peak ground accelerations
approximately equal to 1.0 g (shown in Figure 7) are selected as ground excitations in nonlinear SSI
dynamic analyses.
Using assumed set of artificial earthquake records, repeated nonlinear analyses were performed by
gradually increasing the intensity of acceleration time histories to a level, where first yielding of steel
is occurred. The values of PGA for which the generic frames supporting on different flexible founda-
tions reach their yielding states of response are given in Tables 2–4 for five-storey, 10-storey,
20-storey generic frames, respectively. The yield limit state is defined when the strain in the main
tensile reinforcement exceeds the design yield strain of steel. The definition of collapse for RC frames
as a whole is quite subjective and depends on engineering judgment. In this study, the collapse limit
state is assumed corresponding to the formation of a lateral mechanism in structure. The values of
PGA for which the generic frames supporting on different flexible foundations in a process of gradu-
ally increasing the intensity of ground acceleration reaches their collapse states are also given in
Tables 5–7 for five-storey, 10-storey, 20-storey generic frames, respectively.
Using Equation (7), the ductility reduction factor can be quantified by subdividing the peak
ground accelerations of the collapse to the PGA corresponding to the first yielding in system.
The variation of ductility reduction factors with respect to soil shear wave velocities are shown in
Figure 8(a, b) for five-storey generic frames having two levels of stiffnesses at the strength level,
C, equal to 0.1 subjected to assumed ground excitations. Figure 9(a, b) shows the same information
but for the different strength level at C = 0.2. The calculated response reduction factors for the
fixed-base generic models having different levels of strengths are also shown in Figures 8 and 9.
A comparison of Figures 8 and 9 shows that the effect of SSI on response reduction factor for
DOI: 10.1002/tal
ACC1 ACC1
10 1
Acceleration,m/s2
Traveling FFT for

0.8
5
Acceleration
0.6
0
0 4 8 12 16 20 24 28 0.4
-5 0.2
0
-10
0 5 10 15 20 25
Time,sec. Frequency(Hertz)
ACC2 ACC2
1
10
Acceleration,m/s2
Traveling FFT for

0.8
Acceleration
5
0.6
0 0.4
0 4 8 12 16 20 24 28 0.2
-5
0
-10 0 5 10 15 20 25
Time,sec. Frequency(Hertz)
ACC3 ACC3
10 1
Acceleration,m/s2
Traveling FFT for

0.8
5
Acceleration
0.6
0
0.4
0 4 8 12 16 20 24 28
-5 0.2
-10 0
Time,sec. 0 5 10 15 20 25
Frequency(Hertz)
ACC4 ACC4
10 1
Acceleration,m/s2
Traveling FFT for
0.8
5
Acceleration
0.6
0
0.4
0 4 8 12 16 20 24 28
-5 0.2
-10 0
Time,sec. 0 5 10 15 20 25
Frequency(Hertz)
ACC5
10 ACC5
1
Acceleration,m/s2
Traveling FFT for
5 0.8
Acceleration
0.6
0
0 4 8 12 16 20 24 28
0.4
-5 0.2
-10 0
Time,sec. 0 5 10 15 20 25
Frequency(Hertz)
Figure 7. Acceleration time history and Fourier Amplitude Spectrum for artificial records.
DOI: 10.1002/tal
Table 2. Ground acceleration at yield limit state for generic frames (five-storey frames).
ag (actual yield)
Artificial records Fixed base Vs = 500 Vs = 300 Vs = 200 Vs = 100 Vs = 50
T = 0.5 s, C = 0.1
ACC1 0.37 0.36 0.34 0.33 0.36 0.35
ACC2 0.39 0.38 0.37 0.37 0.35 0.33
ACC3 0.35 0.37 0.4 0.39 0.37 0.3
ACC4 0.34 0.34 0.34 0.34 0.29 0.26
ACC5 0.41 0.4 0.36 0.35 0.3 0.3
T = 0.5 s, C = 0.2
ACC1 0.59 0.59 0.56 0.56 0.56 0.55
ACC2 0.54 0.56 0.54 0.55 0.52 0.53
ACC3 0.49 0.5 0.51 0.51 0.51 0.41
ACC4 0.52 0.52 0.53 0.52 0.52 0.5
ACC5 0.59 0.58 0.59 0.58 0.59 0.5
T = 0.7 s, C = 0.1
ACC1 0.38 0.39 0.38 0.36 0.32 0.2
ACC2 0.37 0.38 0.36 0.34 0.31 0.22
ACC3 0.37 0.33 0.31 0.27 0.26 0.27
ACC4 0.35 0.33 0.28 0.27 0.26 0.2
ACC5 0.34 0.34 0.33 0.32 0.27 0.24
T = 0.7 s, C = 0.2
ACC1 0.47 0.47 0.48 0.48 0.46 0.6
ACC2 0.47 0.47 0.47 0.48 0.5 0.6
ACC3 0.4 0.4 0.4 0.4 0.4 0.55
ACC4 0.48 0.49 0.49 0.49 0.48 0.54
ACC5 0.49 0.5 0.5 0.53 0.54 0.55
Table 3. Ground acceleration at yield limit state for generic frames (10-storey frames).
ag (actual yield)
T = 1 s, C = 0.05
ACC1 0.36 0.36 0.36 0.37 0.38 0.38
ACC2 0.36 0.36 0.36 0.36 0.39 0.4
ACC3 0.39 0.39 0.38 0.37 0.4 0.45
ACC4 0.32 0.32 0.32 0.31 0.3 0.45
ACC5 0.4 0.4 0.38 0.4 0.36 0.43
T = 1 s, C = 0.1
ACC1 0.48 0.48 0.48 0.47 0.44 0.5
ACC2 0.52 0.52 0.52 0.52 0.53 0.5
ACC3 0.47 0.47 0.48 0.46 0.45 0.47
ACC4 0.48 0.48 0.48 0.46 0.44 0.55
ACC5 0.46 0.46 0.47 0.45 0.5 0.45
T = 1.5 s, C = 0.05
ACC1 0.34 0.35 0.35 0.34 0.35 0.3
ACC2 0.35 0.35 0.34 0.34 0.33 0.32
ACC3 0.32 0.32 0.32 0.32 0.32 0.34
ACC4 0.34 0.34 0.34 0.34 0.36 0.34
ACC5 0.36 0.36 0.36 0.36 0.3 0.34
T = 1.5 s, C = 0.1
ACC1 0.41 0.41 0.41 0.42 0.43 0.44
ACC2 0.38 0.37 0.37 0.37 0.37 0.38
ACC3 0.36 0.36 0.36 0.36 0.36 0.36
ACC4 0.4 0.4 0.4 0.4 0.41 0.46
ACC5 0.36 0.36 0.36 0.36 0.36 0.38
DOI: 10.1002/tal
Table 4. Ground acceleration at yield limit state for generic frames (20-storey frames).
ag (actual yield)
T = 2 s, C = 0.03
ACC1 0.3 0.3 0.3 0.32 0.32 0.3
ACC2 0.38 0.39 0.39 0.36 0.37 0.48
ACC3 0.37 0.37 0.37 0.38 0.36 0.32
ACC4 0.39 0.39 0.38 0.37 0.33 0.33
ACC5 0.36 0.36 0.35 0.34 0.33 0.35
T = 2 s, C = 0.08
ACC1 0.58 0.58 0.58 0.59 0.63 0.42
ACC2 0.56 0.57 0.56 0.56 0.57 0.47
ACC3 0.53 0.52 0.52 0.52 0.53 0.49
ACC4 0.67 0.67 0.66 0.64 0.68 0.42
ACC5 0.64 0.63 0.63 0.61 0.57 0.41
T = 3 s, C = 0.03
ACC1 0.38 0.39 0.39 0.38 0.8 0.43
ACC2 0.54 0.49 0.5 0.5 0.53 0.49
ACC3 0.49 0.49 0.49 0.49 0.49 0.49
ACC4 0.36 0.34 0.34 0.34 0.34 0.35
ACC5 0.46 0.45 0.45 0.45 0.45 0.49
T = 3 s, C = 0.08
ACC1 0.64 0.64 0.63 0.62 0.59 0.69
ACC2 0.9 0.9 0.9 0.91 0.92 1.13
ACC3 0.94 0.93 0.93 0.95 0.85 1.01
ACC4 0.95 0.95 0.96 0.85 0.86 1
ACC5 0.63 0.63 0.63 0.61 0.57 0.93
Table 5. Ground acceleration at collapse limit state for generic frames (five-storey frames).
ag (actual yield)
T = 0.5 s, C = 0.1
ACC1 1.98 1.95 1.93 1.89 1.9 0.95
ACC2 2.09 2.01 1.91 1.9 1.87 1.05
ACC3 1.87 1.96 1.95 1.87 1.9 1.1
ACC4 2.32 2.31 2.33 2.28 2.19 0.95
ACC5 1.93 1.9 1.92 1.93 2.17 1.1
T = 0.5 s, C = 0.2
ACC1 2.2 2.3 2.19 2.1 2.2 1.3
ACC2 2.1 2.15 2.02 2 1.75 1.7
ACC3 1.95 1.95 1.95 1.91 1.87 1.13
ACC4 2.2 2.19 2.18 2.05 1.97 1.3
ACC5 2.2 2.12 2.11 2.03 2.01 1.35
T = 0.7 s, C = 0.1
ACC1 1.7 1.68 1.65 1.6 1.6 0.55
ACC2 1.67 1.65 1.67 1.66 1.67 0.65
ACC3 1.51 1.51 1.5 1.49 1.52 0.7
ACC4 2.01 1.98 1.85 1.92 1.78 0.6
ACC5 1.39 1.4 1.4 1.49 1.42 0.65
T = 0.7 s, C = 0.2
ACC1 1.48 1.47 1.49 1.44 1.32 1.23
ACC2 1.8 1.79 1.78 1.79 1.71 1.43
ACC3 1.75 1.74 1.73 1.71 1.7 1.16
ACC4 2 1.99 2.01 2 1.9 1.23
ACC5 1.36 1.36 1.35 1.31 1.24 1.13
DOI: 10.1002/tal
Table 6. Ground acceleration at collapse limit state for generic frames (10-storey frames).
ag (actual yield)
T = 1 s, C = 0.05
ACC1 1.34 1.35 1.35 1.36 1.37 1.16
ACC2 1.48 1.47 1.48 1.44 1.36 1.2
ACC3 1.39 1.39 1.36 1.32 1.44 1.49
ACC4 1.94 1.95 1.99 1.99 2 1.69
ACC5 1.91 1.97 1.83 2.05 1.96 1.83
T = 1 s, C = 0.1
ACC1 1.79 1.79 1.78 1.91 1.9 1.43
ACC2 2.19 2.19 2.1 2 1.69 1.49
ACC3 1.8 1.8 1.79 1.76 1.72 1.1
ACC4 1.79 1.78 1.76 1.7 1.74 1.58
ACC5 1.64 1.65 1.66 1.67 1.6 1.3
T = 1.5 s, C = 0.05
ACC1 1.25 1.36 1.37 1.35 1.27 1.04
ACC2 1.3 1.29 1.28 1.3 1.22 1.06
ACC3 0.99 0.99 0.95 0.97 0.93 0.92
ACC4 1.61 1.63 1.63 1.63 1.62 1.39
ACC5 1.9 1.92 1.91 1.93 1.85 1.48
T = 1.5 s, C = 0.1
ACC1 1.38 1.38 1.37 1.39 1.4 1.2
ACC2 1.81 1.76 1.75 1.75 1.76 1.37
ACC3 1.55 1.51 1.51 1.51 1.4 1.35
ACC4 1.43 1.43 1.42 1.46 1.38 1.27
ACC5 1.49 1.48 1.48 1.48 1.43 1.22
Table 7. Ground acceleration at collapse limit state for generic frames (20-storey frames).
ag(actual yield)
T= 2 s, C = 0.03
ACC1 1.3 1.31 1.32 1.46 1.5 1.38
ACC2 2.1 2.16 2.17 2 2.05 1.73
ACC3 1.53 1.53 1.54 1.55 1.54 1.28
ACC4 1.68 1.71 1.68 1.67 1.74 1.27
ACC5 1.56 1.56 1.52 1.51 1.5 1.48
T= 2 s, C = 0.08
ACC1 1.99 2.04 2.05 2.05 1.89 0.93
ACC2 2.49 2.51 2.46 2.45 2.36 1.5
ACC3 1.99 1.95 1.97 1.95 1.92 1.02
ACC4 2.27 2.3 2.23 2.05 2.08 1.13
ACC5 1.76 1.73 1.72 1.66 1.55 0.79
T= 3 s, C = 0.03
ACC1 1.24 1.33 1.33 1.28 1.24 1.26
ACC2 1.5 1.88 1.88 1.86 1.5 1.43
ACC3 1.37 1.5 1.5 1.5 1.4 1.1
ACC4 1.5 1.53 1.53 1.52 1.46 1.38
ACC5 1.8 1.79 1.79 1.78 1.72 1.76
T= 3 s, C = 0.08
ACC1 2.31 2.31 2.31 2.25 2.01 1.76
ACC2 1.69 1.69 1.69 1.82 1.74 1.57
ACC3 2.17 2.16 2.15 2.18 2.02 1.69
ACC4 1.48 1.48 1.47 1.44 1.43 1.59
ACC5 2.16 2.17 2.16 2.07 2 2.55
DOI: 10.1002/tal
T=0.5sec , C=0.1
8
7
(PGA)collapse/(PGA)yield
1
ACC1 ACC2 ACC3 ACC4 ACC5
Fixed Base Vs=500m/s Vs=300m/s Vs=200m/s Vs=100m/s Vs=50m/s

(a)
T=0.7sec , C=0.1
8
7
1 ACC1 ACC2 ACC3 ACC4 ACC5

(b)
Figure 8. Variation of response reduction factor with supporting soil shear velocity for low-rise
frames having low strength.
DOI: 10.1002/tal
T=0.5sec , C=0.2
5
4
1
0
(a)
T=0.7sec , C=0.2
5
1
0
(b)
Figure 9. Variation of response reduction factor with supporting soil shear velocity for low-rise
frames having high strength.
RC frames supporting on very soft soils could be detrimental. In this case, excessive flexible base
could probably act as a figurative soft storey causing noticeable decrease (up to 50%) in response
reduction factor. The results also demonstrate that for typical stiff soils with 300 < Vs ≤ 500 m/s,
the effect of foundation flexibility on ductility reduction factors could be beneficial. However, for
short period structures, this result should be used with caution. In the case of typical soft soils with
100 ≤ Vs ≤ 300 m/s, it can be noted that depending on the strength level of the structure, the effect
of the foundation flexibility on ductility reduction factor of RC frames could be beneficial or det-
rimental. The results emphasize that five-storey frames with lower strength show up to 50% increase
in ductility reduction factor; while five-storey frames with higher strength outcomes show a decrease
of about 15% in the response reduction factor. In fact, it is indicated that the frames with the same
stiffness but various strengths could demand different ductility when considering SSI effect. The
variation of mean and the mean plus standard deviation response reduction factors of the five-storey
frames under five artificial records, referred to as (mean) and (mean + standard deviation), with
soil shear wave velocity are shown in Figure 10(a, b) for two levels of stiffnesses and two different
strength levels.
Similarly, the variation of strength dependent ductility reduction factors for 10-storey generic
frames subjected to artificial excitations used in this study are shown in Figures 11(a, b) and 12(a,
b). Each figure illustrates the variation of ductility reduction factors covering a range of flexibly
DOI: 10.1002/tal
T= 0.5sec
8
7
3
T=0.5sec , C=0.1 (mean+standard deviation)
2 T=0.5sec , C=0.1 (mean)
1
T=0.5sec , C=0.2 (mean)
0
(a)
T= 0.7sec
8
7
3
2 T=0.7sec , C=0.1 (mean)
1
T=0.7sec , C=0.2 (mean)
0
(b)
Figure 10. Variation of mean and mean plus one standard deviation factor response reduction
factor with supporting soil shear velocity for low-rise structures.
DOI: 10.1002/tal
T=1sec, C=0.05
8
7
0
(a)
T=1.5sec, C= 0.05
8
7
0
(b)
Figure 11. Variation of response reduction factor with supporting soil shear velocity for mid-rise
DOI: 10.1002/tal
T=1 , C=0.1
5
4
1
0
(a)
T= 1.5sec , C= 0.1
5
1

(b)
Figure 12. Variation of response reduction factor with supporting soil shear velocity for mid-rise
supported to fixed-supported structures at a particular strength level. This variation is shown for
structures with two different fundamental natural periods, i.e. (a) T = 1 s and (b) T = 1.5 s. The
results show that like short-rise RC frames, the foundation flexibility could significantly decrease
the force reduction factors for mid-rise structures supporting on very soft soils. However, in the
case of typical stiff soils with 300 < Vs ≤ 500 m/s, the effect of foundation flexibility is approxi-
mately negligible. For 10-storey frame structure resting on soils with wave velocity 100 ≤ Vs ≤
300 m/s, the effects of SSI on reduction factors vary with strength levels of generic frames; so that
results demonstrated 15% increase or decrease in force reduction factors depending on excitation
frequency content. In general, it can be concluded that the effect of foundation flexibility on ductil-
ity reduction factor of mid-rise RC frame structure is of lesser degree compared with the short-rise
structures. The variation of mean and the mean plus standard deviation response reduction factors
of the 10-storey frames under five artificial records, referred to as (mean) and (mean + standard
deviation), with soil shear wave velocity are shown in Figure 13(a, b) for two levels of stiffnesses
and two different strength levels.
Figures 14(a, b) and 15(a, b) present ductility reduction factors for 20-storey generic frames
subjected to each particular acceleration time history versus different supporting soil shear wave
velocities for a practical range of structure’s strength level. In order to evaluate the effect of
the structure period, the variation of the ductility reduction factors with the strength level and the
DOI: 10.1002/tal
T= 1sec
8
7
3
T=1sec , C=0.05 (mean+standard deviation)
2
T=1sec , C=0.05 (mean)
1 T=1sec , C=0.1 (mean+standard deviation)

0
(a)
T=1.5sec
8
7
3
2 T=1.5sec , C=0.05 (mean)
1
T=1.5sec , C=0.1 (mean)
0
(b)
factor with supporting soil shear velocity for mid-rise structures.
DOI: 10.1002/tal
T= 2sec , C= 0.03
8
7
0
(a)
T=3sec , C=0.03
8
7
(PGA)collapse/(PGA)yiel
d
0
(b)
Figure 14. Variation of response reduction factor with supporting soil shear velocity for high-rise
DOI: 10.1002/tal
T=2sec ,C=0.08
5
3
1
0
(a)
T= 3sec , C=0.08
5
4
1
0
(b)
Figure 15. Variation of response reduction factor with supporting soil shear velocity for high-rise
supporting soil shear wave velocity is presented for selected period values T = 2.0 and 3.0 s. The
calculated response reduction factors for the fixed-base generic models having different levels of
strengths are also shown in Figures 14 and 15. It can be noted that whenever the ductile behaviour
of structures is considered in design, the ductility reduction factor of the structure is sensitive to
the stiffness and strength of the structure as well as the foundation flexibility. It can be noticed
that the results for soil shear wave velocities less than about 150 m/s showed that the additional
flexibility due to the foundation compliance to the flexible structures reduces the ductility of the
system with lesser degree compared with short-rise and mid-rise RC frames, an apparently detri-
mental effect of SSI. In the case of typical stiff soils with 200 < Vs ≤ 500 m/s, the effect of foun-
dation flexibility is approximately negligible. In general, the results illustrated that the foundation
flexibility effect on response reduction factor is of the lesser order than the stiffness and strength
of the structure. The variation of mean and the mean plus standard deviation response reduction
factors of the 20-storey frames subjected to five artificial records, referred to as (mean) and (mean
+ standard deviation), with soil shear wave velocity are shown in Figure 16(a, b) for two levels of
stiffnesses and two different strength levels.
It is interesting to study the effect of different parameters affecting response reduction factors
of RC frame structures. Accordingly, the results of more than 1200 inelastic time–history analyses
presented in Figures 8–16 were employed to perform regression analyses obtaining response reduc-
tion factor envelope for examined RC generic frames supporting on different flexible foundations.
DOI: 10.1002/tal
T=2sec
8
7 T=2sec , C=0.03 (mean)
6
5
0
(a)
T=3sec
8
T=3sec ,C=0.03 (mean+standard deviation)
7
T=3sec ,C=0.03 (mean)
6 T=3sec ,C=0.08 (mean+standard deviation)

T=3sec ,C=0.08 (mean)
5
0
(b)
factor with supporting soil shear velocity for high-rise structures.
Figure 17(a,b) depict the mean plus standard deviation response reduction factor points of generic
frames subjected to the five artificial seismic actions considered in this study, considering the
strength level of the structure. The generic RC frames are classified again according to their strength
level, small strength and large strength levels, taking also their stiffness into account as follows:
⎧C ≤ 0.1 for 0.5sec ≤ T < 1 sec

⎪
Small Strength → ⎨C ≤ 0.05 for 1sec ≤ T < 2 sec
⎪C ≤ 0.03 for 2sec ≤ T ≤ 3 sec
⎩
⎧C > 0.1 for 0.5sec ≤ T < 1 sec

⎪
Large Strength → ⎨C > 0.05 for 1sec ≤ T < 2 sec
⎪C > 0.03 for 2sec ≤ T ≤ 3 sec
⎩
The response reduction factors are shown for the fixed-base and flexibly based structures. Again,
Figure 17(a,b) demonstrate that force reduction factor could be related to the period of the structure.
Furthermore, the strength of the structure could be of the important parameters affecting the force
reduction factors which also alters the trend of the variation of force reduction factors with period of
DOI: 10.1002/tal
Small Strength
8
{(PGA)collapse/(PGA)yield}mean
7
+standard deviation
5
2
fixed base Vs=500m/s Vs=300m/s
1 Vs=200m/s Vs=100m/s Vs=50m/s
0
0 0.5 1 1.5 2 2.5 3
T(sec)
(a)
Large Strength
5
{(PGA)collapse/(PGA)yield} mean
4
+ standard deviation
fixed base Vs=500m/s Vs=300m/s

1
Vs=200m/s Vs=100m/s Vs=50m/s
0
0 0.5 1 1.5 2 2.5 3
T(sec)
(b)
Figure 17. Variation of mean plus one standard deviation factor response reduction factor with
supporting soil shear velocity for: (a) low strength structures; b) high strength structures.
structures. Based on the resulting data shown in Figures 17 and 18 on various kinds of structures in
terms of stiffness and strength supporting on different types of soil conditions, the following expres-
sions may be suggested for the force reduction factors:
⎧ R = −2.074 − 1.1 + 7.4 + 3.965T − 0.815T 2 − 0.327 Log (Vs) for Small Strength
⎪⎪ T2 T T
⎨ (21)
⎪ R = 3.7 + T − 0.35T + 1776 34. 4
2
− for Large Strength
⎪⎩ Vs2 Vs
The fitted regression equations of response reduction factors suggested in Equation (21) along
with the mean plus standard deviation of response reduction factors of the frames examined in this
study are presented in Figures 18(a,b) and 19(a,b) for two levels of structural strength in terms of the
structural period representing structural stiffness and soil shear wave velocity representing supporting
soil flexibility as well. The variation coefficient, r2, for the fitted regression equations are 0.95 and
0.87 for higher and lower strength levels, respectively; where,
DOI: 10.1002/tal
(a)
(b)
Figure 18. Regression lines for response reduction factor of small strength structures in terms of:
(a) structural period; (b) supporting soil shear wave velocity.
SSR
r2 = (22)
SSR + SSD

in which SSR = Σ(T i − T )2 and SSD = Σ(Ti − T i)2. T , Ti and T i are the mean of variables in the regres-
sion process, the variable from the set of data contributing in the regression process and the corre-
sponding variable measured using the fitted regression equations, respectively.
7. CONCLUSIONS
In this study, to determine the SSI effect on force reduction factor of RC frame structures, 2D generic
frames were developed. The incremental time history analysis using five artificial earthquake records
was employed to assess response reduction factor. The results showed that whenever the ductile
behaviour of RC frame structures is considered in design, the ductility reduction factor of the structure
is sensitive to the stiffness of the structure, strength of the structure and the foundation flexibility.
The main findings drawn from this study can be summarized as:
(1) For RC frame structures resting on very soft soil with Vs = 50 m/s, effect of foundation flexibil-
ity considerably decreases force reduction factor. In this case, excessive flexible base may
DOI: 10.1002/tal
(a)
(b)
Figure 19. Regression lines for response reduction factor of large strength structures in terms of:
(a) structural period; (b) supporting soil shear wave velocity.
probably act as a figurative soft storey which causes noticeable decrease in force reduction
factor.
(2) For typical stiff soils with 300 < Vs ≤ 500 m/s, the effect of foundation flexibility for RC frame
structures is approximately negligible.
(3) The foundation flexibility could alter force reduction factor for typical soft soils with
100 ≤ Vs ≤ 300 m/s, depending on the structural period and the level structural strength. For
short-rise frames with low strength level, the response reduction factor increases to maximum
about 50%, while for similar frame with higher strength, force reduction factor decreases
to maximum about 15%. For medium-rise frames under the same soil condition, force reduc-
tion factor increased or decreased to maximum about 15% depending on the frequency content
of excitation. These changes for high-rise frames are lesser to the order of maximum about
10%.
(4) The effect of foundation flexibility on force reduction factor decreases while the period of
frame systems increases.
(5) The results also demonstrated that as the strength of frame structures increases, the effect of
foundation flexibility on the response reduction factor decreases.
DOI: 10.1002/tal
REFERENCES
Avilés J, Pérez-Rocha JL. 2005. Influence of foundation flexibility on Rμ and Cμ factors. Structural Engineering 131(2):
221–230.
Avilés J, Pérez-Rocha LE. 2003. Soil-structure interaction in yielding systems. Earthquake Engineering and Structural Dynam-
ics 32: 749–1771.
Crouse CB. 2002. Commentary on soil-structure interaction in U.S. seismic provisions. In Proceeding of the 7th U.S. National
Conference on Earthquake Engineering, Boston, USA.
Elnashai AS, Broderick BM. 1996. Seismic response of composite frames-II. Calculation of behaviour factors. Engineering
Structures 18(9): 707–723.
Halabian AM, Kabiri S. 2006. Effect of soil-structure interaction on system ductility of R/C stake-like structures. First Euro-
pean Conference on Earthquake Engineering and Seismology, Geneva, Switzerland; 540.
Halabian AM, Emami AR. 2008. Effect of foundation flexibility on inelastic response of 3D R/C frame structures. The Fourth
National Conference on Civil Engineering, Tehran, Iran.
Humar JL, Rahgozar MA. 1996. Concept of overstrength in seismic design. 11th World Conference on Earthquake Engineer-
ing, Oxford, England; 639.
Stewart JP, Kim S, Bielak J, Dobry R, Power MS. 2003. Revisions to soil-structure interaction procedures in NEHRP design
provisions. Earthquake Spectra 19: 677–696.
Veletsos AS, Verbic B. 1973. Vibration of viscoelastic foundations. Earthquake Engineering and Structural Dynamics 2:
87–102.
Veletsos AS, Verbic B. 1974. Dynamics of elastic and yielding structure–foundation systems. Proceedings of the 5th World
Conference on Earthquake Engineering, Italy; 2610–2613.
Veletsos AS, Wei YT. 1971. Lateral and rocking vibration of footing. Journal of the Soil Mechanics and Foundations Division,
ASCE 97: 1227–1248.
Wafi M, Elnashai AS. 2002. Calibration of force reduction factors of RC structures. Journal of Earthquake Engineering 6(2):
239–273.
DOI: 10.1002/tal

The Effect of Foundation Flexibility and Structural Strength On Response Reduction Factor of RC Frame Structures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Effect of Foundation Flexibility and Structural Strength On Response Reduction Factor of RC Frame Structures

Uploaded by

Copyright:

Available Formats

THE STRUCTURAL DESIGN OF TALL AND SPECIAL BUILDINGS

Struct. Design Tall Spec. Build. (2010)

The effect of foundation flexibility and structural strength on

Amir M. Halabian*,† and Masoomeh Erfani

Copyright © 2010 John Wiley & Sons, Ltd.

2. EVALUATION OF RESPONSE REDUCTION FACTOR

Rcode = (Sa ) (Sa )in

Rc,dy = (Sa )c (Sa )

Rc, ay = (Sa )c (Sa )y

Rc, ay = a g(collapse) a g(actual yield) (7)

(Sa)ely R`c,ay = R c,ay ⋅ Ω d

R c,ay = (S a ) elc /(S a ) ely

Rc′,ay = Rc,ay ⋅ Ωd = (a g(collapse ) a g(actual yield ) ) ⋅ Ω d (8)

5 − story 10 − story 20 − story

Figure 2. Generic structures.

Figure 3. Plastic hinge mechanism for the generic structures.

SRSS Story Lateral Load Pattern

0 0.2 0.4 0.6 0.8 1

SRSS Story Lateral Load Pattern

0.0 0.2 0.4 0.6 0.8 1.0

Table 1. Cross-section dimensions for generic RC frame structures.

Dimension of beam and column sections (mm)

4. MODELLING FOUNDATION FLEXIBILITY

5. NONLINEAR MODELLING OF RC FRAMES AND EQUATIONS OF MOTION

Figure 6. Fibre element forces and deformations.

⎡3 f yy1 + f yy2 − f yy1 − f yy2 3 f xy1 + f xy2 − f xy1 − f xy2 −4 f y1 − 2 f y2 ⎤

[ M ] {u (t )} + [C ] {u (t )} + [ K ] {u (t )} = [ M ] {I } {ug } (15)

[M ] ∫ {du} + ( am [ M ] + a0 [ K 0 ]) ∫ {du} + ∫ ak [ K ]{du}

{Δu}t + Δt = Δt {u}t + ⎛ − β ⎞ .Δt 2 {u}t + β .Δt 2 {u}t + Δt

And then the acceleration and velocity are calculated as:

Substituting Equation (19) into Equation (17) results to,

Traveling FFT for

Traveling FFT for

Traveling FFT for

Traveling FFT for

Traveling FFT for

Fixed Base Vs=500m/s Vs=300m/s Vs=200m/s Vs=100m/s Vs=50m/s

1 ACC1 ACC2 ACC3 ACC4 ACC5

Fixed Base Vs=500m/s Vs=300m/s Vs=200m/s Vs=100m/s Vs=50m/s

1 ACC1 ACC2 ACC3 ACC4 ACC5

1 ACC1 ACC2 ACC3 ACC4 ACC5

Fixed Base Vs=500m/s Vs=300m/s Vs=200m/s Vs=100m/s Vs=50m/s

1 T=1sec , C=0.1 (mean+standard deviation)

1 ACC1 ACC2 ACC3 ACC4 ACC5

1 ACC1 ACC2 ACC3 ACC4 ACC5

T=3sec ,C=0.03 (mean)

6 T=3sec ,C=0.08 (mean+standard deviation)

⎧C ≤ 0.1 for 0.5sec ≤ T < 1 sec

⎧C > 0.1 for 0.5sec ≤ T < 1 sec

fixed base Vs=500m/s Vs=300m/s

Vs=200m/s Vs=100m/s Vs=50m/s

You might also like