Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Computers and Structures 88 (2010) 45–53

Contents lists available at ScienceDirect

Computers and Structures


journal homepage: www.elsevier.com/locate/compstruc

A practical method for proper modeling of structural damping in inelastic


plane structural systems
Farzin Zareian a,*, Ricardo A. Medina b
a
Department of Civil and Environmental Engineering, University of California, Irvine, CA 92697, United States
b
Department of Civil Engineering, University of New Hampshire, Durham, NH 03824, United States

a r t i c l e i n f o a b s t r a c t

Article history: This study addresses some of the pitfalls of conventional numerical modeling of Rayleigh-type damping
Received 5 March 2009 in inelastic structures. A practical modeling approach to solve these problems is proposed. Conventional
Accepted 3 August 2009 modeling of Rayleigh-type damping for inelastic structures generates responses in which unrealistic
Available online 28 August 2009
damping forces are present that results in underestimation of peak displacement demands, overestima-
tion of peak strength demands, and underestimation of buildings’ collapse potential. The approach pro-
Keywords: posed in this paper avoids these problems by modeling each structural element with an equivalent
Rayleigh damping
combination of one elastic element with stiffness-proportional damping, and two springs at its two ends
Inelastic responses
Structural damping
with no stiffness proportional damping.
Dynamic analysis Ó 2009 Elsevier Ltd. All rights reserved.
Seismic response
Seismic evaluation

1. Introduction The failure in proper modeling of structural damping is com-


pounded by (1) the lack of reliable experimental data to validate
One component that can arguably be considered overlooked in the structural damping models used to represent the energy dissi-
current methods for estimation of response of structural systems is pation characteristics of structural systems in inelastic regimes;
the appropriate modeling of structural damping. This shortcoming and (2) the decreased identification accuracy for damping ratios
becomes even more apparent when compared with the advances when compared to that of elastic natural frequencies and mode
made in modeling of the inelastic response of structural and non- shapes. Results from full-scale tests show that changes in damping
structural components during the past decades. These advance- are much greater than those for frequency over a similar amplitude
ments include computational and modeling techniques that can range [1]. In the absence of more accurate structural damping
reasonably capture material and geometric nonlinearities, as well models, linear viscous damping is usually utilized for convenience.
as the evolution of damage in a structure. In this context, the term The most common model of viscous damping used in modeling of
structural damping refers to energy dissipation mechanisms pres- multi-degree-of-freedom (MDOF) systems is the Rayleigh-type
ent in a structure due to structural and nonstructural component damping where the damping matrix, C, is composed of the super-
responses to dynamic excitation other than the energy dissipated position of a mass-proportional damping term (i.e., aM) and a stiff-
in inelastic excursions. Traditionally, linear viscous damping has ness-proportional damping term (i.e., bK) [2].
been utilized to model the energy dissipation characteristics of a
structural system exposed to dynamic excitation. In such model, C ¼ aM þ bK ð1Þ
the effect of damping is accounted for at a global scale – the energy Physically, mass-proportional damping (MPD) is equivalent to
dissipated through friction and slippage in joints for structural and having externally supported dampers attached to the dynamic
nonstructural components is represented by means of equivalent (inertial) degrees of freedom while stiffness-proportional damping
linear viscous damping. However, the use of a linear viscous damp- (KPD) implies the presence of viscous dampers (i.e., dashpots) that
ing model in many cases produces inaccurate estimates of dis- join two adjacent dynamic degrees of freedom. MPD has the effect
placements and internal forces in members. These inaccurate of having modal damping ratios that are inversely proportional to
estimates of internal forces are related to nodes in the structural the frequencies of vibration of the system, while KPD produces
model in which unrealistic damping forces are generated. modal damping ratios directly proportional to the modal frequen-
cies of the structure. In both cases, very little experimental verifi-
* Corresponding author. Tel.: +1 949 824 9866; fax: +1 949 824 2117. cation exists for such model, particularly for structural systems
E-mail address: zareian@uci.edu (F. Zareian). that undergo significant levels of inelastic deformations.

0045-7949/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compstruc.2009.08.001
46 F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53

In the last few years, several researches have studied and iden- structure (i.e., R = 6, X = 2, Rl = 3). Rayleigh-type damping based
tified important limitations of Rayleigh-type damping as it applies on initial stiffness is used and damping ratios at the first and third
to inelastic systems. As shown by Bernal [3], Medina and Krawin- mode are set to 5%. On the same plot, the energy dissipated due to
kler [4], and Hall [5], when Rayleigh-type damping based on the hysteretic action of structural components that enter the inelastic
initial stiffness matrix is used, unrealistic damping forces are gen- regime relative to the total input energy is shown with a black line
erated at joints in which structural elements undergo abrupt with square markers. The vertical axis shows the level of ground
changes in stiffness. This is due to the tendency of degrees of free- motion intensity in terms of Sa/g (spectral acceleration at the first
dom with small inertias to undergo abrupt changes in velocity mode period of the structure). The ground motion recording used
once the stiffness of the element changes during the inelastic re- belongs to the 1994 Northridge earthquake. This building has been
sponse [3]. Thus, unrealistic damping forces develop at these de- modeled with a concentrated plasticity approach in which the hys-
grees of freedom resulting in an underestimation of the peak teretic response at the end of beams and columns exhibits moder-
displacement demands in the structure and an overestimation of ate levels of monotonic and cyclic deterioration based on the
internal forces for elements in the system that do not undergo model developed by Ibarra et al. [6].
changes in stiffness [4]. Moreover, as shown in this paper, inappro- As illustrated in Fig. 1, when the structural system begins to
priate model of structural damping may also cause underestima- experience inelastic deformations (i.e., Sa/g > 0.75), the fraction of
tion of the collapse potential of buildings. the total input energy dissipated through viscous damping is re-
The objective of this paper is to present a more appropriate and duced and the energy dissipated in structural components hyster-
easy-to-apply numerical modeling approach for implementing esis loops is increased. However, the energy dissipated through
Rayleigh-type damping in structures. This approach eliminates viscous damping has a lower-bound of approximately 27% of the
the presence of unrealistic damping forces in inelastic time history total input energy. In this case, even when cyclic deterioration
responses. The implication is that the proposed approach will pro- and relatively large levels of inelastic behavior are present (around
vide improved inelastic dynamic response predictions. An illustra- Sa/g = 4.0), the contribution of viscous damping in dissipating the
tion of the benefits of using this new approach in the context of input energy to the structural system is increased as the hysteretic
seismic performance assessment is presented. energy dissipation capacity of structural components is exhausted
due to damage and deterioration. Therefore, inappropriate model-
ing of structural damping has the potential to provide erroneous
2. Viscous damping, Rayleigh damping, and inelastic responses
demand prediction for structures exposed to strong dynamic exci-
tations. This example utilized the most commonly used linear vis-
Although damping is considered the primary energy dissipation
cous damping model, which is the Rayleigh-type damping based
mechanism for elastic structures, the focus on inelastic structures
on initial stiffness, i.e., a time invariant stiffness matrix. Muto
is warranted by the presence of unrealistic damping forces when
and Beck [7] have also highlighted the importance of appropriate
Rayleigh-type damping based on initial stiffness is assigned to
modeling of viscous damping when applying system identification
structural elements that experience inelastic responses. In addi-
techniques to estimate hysteretic structural parameters of systems
tion, for relatively large levels of inelastic behavior, the energy dis-
subjected to earthquake loading. They showed that excluding vis-
sipated via structural damping, as predicted by numerical models
cous damping from identification models will significantly modify
with Rayleigh-type damping, still constitutes a significant percent-
the value of the identified hysteretic parameters if viscous damp-
age (e.g., 25%) of the total dissipated energy. Fig. 1 shows the en-
ing is present in the structure.
ergy dissipated due to linear viscous damping relative to the
Studies conducted by Medina and Krawinkler [4] on regular
total input energy for a 4-story MDOF model, which represents a
moment-resisting frame structures exposed to far-field ground
non-ductile reinforced-concrete building, at various levels of
motions have shown that improper modeling of 5% critical struc-
ground motion intensity measure (gray line with diamond mark-
tural damping using the Rayleigh model based on initial stiffness
ers). The MDOF model has fundamental period T = 0.4 s, and the
results in the underestimation of peak-drift demands, on average,
global strength of the structure corresponds to a response modifi-
in the order of 10%. However, the overestimation in peak strength
cation factor of 6 assuming an over-strength factor of 2 for this
demands can be in the order of 30% depending upon the structural
properties and ground motion characteristics. This phenomenon is
illustrated in Fig. 2, which presents representative results
Relative Dissipated Energy to Input Energy Mean IDA curves corresponding to a numerical model of a moment-resisting frame
N=4, T1=0.6, R μ = 3.0 , ξ=0.05, Peak-Oriented model, Northridge EQ
5

60%
4 NR94cnp-5% damping ratio
NR94cnp-10% damping ratio
50% LP89cap-10% damping ratio
3
Sa/g

40%
(Mc -Mp,b)/Mp,b

2
30%

1 20%

10%
0
0 0.25 0.5 0.75 1
Relative Dissipated Energy to Input Energy 0%
0 1 2 3 4 5 6 7 8 9
Damping Energy Hysteretic Energy [Sa/g] / γ

Fig. 1. Total damping and hysteretic energy dissipated relative to the input energy Fig. 2. Relative difference between maximum column moment (Mc) and beam
for the 4-story case study building. plastic moment (Mp,b) at the top floor of a single-bay, moment-resisting frame.
F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53 47

structure exposed to two different ground motion records. One of


the records is from the 1994 Northridge Earthquake – Canoga Park
Station (NR94cnp) and the other one from the 1989 Loma Prieta
Earthquake – Capitola Station (LP89cap). The frame was designed
such that plastic hinges have the potential to form at the end of
each beam and at the bottom of the first-story columns. All col-
umns are infinitely strong. Modal damping ratios were varied
and inelastic dynamic analyses were conducted to quantify the rel-
ative difference between the maximum column moment and the
beam plastic moment at the top floor. In concept, this relative dif-
Model A Model B
ference must be negligible. This is clearly not the case when initial
(a) (b)
stiffness proportional damping is assigned to elements whose stiff-
ness changes during the time history (i.e., inelastic beam ele- Fig. 4. Idealized SDOF structures (a) flexible beam with elastoplastic moment–
ments). As expected, the relative difference between column rotation relationship at both ends and (b) flexible beam + semi-rigid rotation
moment and beam moment is negligible in the elastic range (see springs to represent elastoplastic behavior.
Fig. 3). In addition, as both damping ratio and level of inelastic
behavior increase, the relative difference between the maximum
column moment and the beam plastic moment increases. The level Displacement Ductility Ratio, SDOF, Model A
6
of inelastic behavior is quantified by the ratio of the pseudo-spec- Mass Proportional Damping
tral acceleration at the fundamental period of the structure to the
Initial-Stiffness Proportional Damping
base shear strength coefficient, i.e., [Sa/g]/c. Such errors in overes-

Displacement Ductility Ratio


timation of demand forces can lead to over-designed structural
members that are sized using force-based methods. Consequently, 3
these elements have the potential to attract larger forces into the
joint, which may cause the structure to be more vulnerable to
damage and collapse once deterioration of structural components
is modeled. Similarly, underestimation of deformation demands
0
may lead to unsafe structural designs that use displacement-based
methods such as performance-based design.
A better understanding of the consequences of modeling vis-
cous damping based on initial stiffness can be obtained with the
model in Fig. 4a, which depicts a Single-Degree-Of-Freedom -3
0 5 10 15
(SDOF) structure in which two rigid columns are joined by a flexi-
Time (sec.)
ble beam (Model A). The beam is modeled based on a concen-
trated-plasticity approach in which elastoplastic moment– Column-end Moment Demand, SDOF, Model A
rotation relationships are assigned to the beam ends. The period 1.5
of the system is 0.4 s and the plastic moment capacity of the beam
is 3000 kip.-in. (3518.7 N m). To better show the inaccuracy of Ray- 1
leigh damping modeling based on initial stiffness, a damping ratio
of 10% is assumed, and damping in the SDOF system is modeled
Mcolumn / M p,beam

0.5
based on mass-proportional damping (Case 1) and initial stiffness
proportional damping (Case 2). Both models are subjected to a re-
0
corded ground motion that is scaled such that a displacement duc-
tility ratio of 4 (l = 4) is attained. Fig. 5 shows the results of this
analysis in terms of displacement ductility ratios and normalized -0.5
column-end moments, i.e., column-end moment divided by the
beam plastic moment. Fig. 5a demonstrates that by using Model -1
Mass Proportional Damping
Initial-Stiffness Proportional Damping
-1.5
2500 0 5 10 15
Column
2000 Moment Time (sec.)

1500 Beam
Moment Fig. 5. Comparison between response histories of the idealized SDOF system
1000 (Model A) whose 10% critical damping is modeled using ‘‘mass-proportional
Moment (kip -in.)

damping” and ‘‘initial stiffness proportional damping” subjected to a ground motion


500
recording from Northridge Earthquake: (a) displacement ductility ratio, and (b)
0 normalized column-end moment.

-500
-1000
A the displacement ductility ratio response is identical no matter
-1500
which method was used to model structural damping, i.e., Cases
-2000 1 and 2. However, Fig. 5b shows that the column-end moments
-2500 for Case 2 (i.e., initial stiffness proportional damping) exceeds the
0 5 10 15
beam plastic moment of 3000 kip.-in. (3518.7 N m). Once again,
Time (s)
this is due to the presence of unrealistic damping moments in
Fig. 3. Moment time history at the top floor of the moment-resisting frame of Fig. 2 the response. In this particular case, these unrealistic damping
exposed to ground motion LP89cap; damping ratio = 10%; [Sa/g]/c = 8. moments do not affect the displacement response significantly
48 F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53

because they act for relatively short periods of time (spikes in an increase in estimates of displacement demands compared with
Fig. 5a) as compared to the total strong motion duration. predictions based on initial stiffness or mass-proportional damp-
ing. However, the application of this approach may cause numeri-
cal solution instabilities once significant changes in stiffness values
3. Current approaches to overcome the limitations of Rayleigh-
occur, e.g., changes due to material strength and stiffness deterio-
type damping based on initial stiffness
ration. This approach is also computationally more expensive than
that in which the initial stiffness matrix is used.
Solutions to this problem have been proposed by Bernal [3],
Hall [5], and Charney [8]. Bernal [3] proposed a solution in which CðtÞ ¼ aM þ bKt ðtÞ ð5Þ
Caughey-type damping should be used and the damping matrix
should be assembled by restricting the exponents of the Caughey Leger and Dassault [12] proposed a solution in which Rayleigh-
series (l terms in Eq. (2)) to zero or negative. Rayleigh-type damp- type damping with variable coefficients and the tangent stiffness
ing is a special case of Caughey-type damping. matrix are used, Eq. (6). Leger and Dassault argue that this partic-
X ular model provides a more rational control of the amount of en-
C¼M al ½M1 Kl ð2Þ ergy dissipated by viscous damping in nonlinear seismic
l analyses. However, the calculation of the scalar coefficients at each
This proposed solution has the effect of not assigning stiffness- time step, while preserving modal orthogonality, is computation-
proportional damping to degrees of freedom without mass. For ally expensive. In addition, the application of this approach is ques-
example, if values of l = 0 and 1 are used in Eq. (2), the corre- tionable for structural systems that experience significant
sponding Caughey-type matrix becomes: degradation in stiffness because of material strength and stiffness
deterioration.
C ¼ a0 M þ a1 MK1 M ð3Þ
CðtÞ ¼ aðtÞM þ bðtÞKt ðtÞ ð6Þ
It is evident from Eq. (3) that for typical structural models used
in earthquake engineering practice for which masses are lumped at
floor levels, the mass matrix will be diagonal, and hence, the damp-
4. Proposed approach for proper modeling of Rayleigh-type
ing matrix will only have non-zero terms for degrees of freedom
damping in inelastic structures
associated with the inertial masses. Thus, the potential for unreal-
istic damping forces at rotational degrees of freedom is eliminated.
The approach proposed in this study deals with a formulation of
However, this approach does not avoid the presence of unrealistic
a Rayleigh-type matrix with a time invariant stiffness matrix that
damping forces at rotational degrees of freedom when masses are
is assembled by assigning zero stiffness-proportional damping to
assigned to them. The implementation of such Caughey-type
structural elements that have the potential to experience inelastic
damping compromises the numerical efficiency of the solution of
deformations. This approach requires an increase of the stiffness
the equations of motion because for l < 0 the calculation of the in-
proportional damping term to those elements that remain in the
verse of the stiffness matrix K will be required. In addition, as dem-
elastic range throughout the response to enforce damping energy
onstrated by Oliveto and Greco [9], Caughey-type damping with
conservation. The implication is that the structural model will be
l < 0 does not keep the same Caughey coefficients once a change
composed of a combination of elastic and inelastic elements, which
in the support conditions of the structure takes place.
is a common approach in current earthquake engineering simula-
An alternative could be to use only mass-proportional damping
tion studies, but Rayleigh damping is solely applied to the elastic
(l = 0) in Eq. (3), i.e., Eq. (4). Although this approach will eliminate
elements. As it will be shown in this section, this modeling ap-
spurious damping forces, the displacement response of the multi-
proach provides results that are consistent with those obtained
degree-of-freedom structure will exhibit significant higher-fre-
when Rayleigh-type damping based on the tangent stiffness of
quency content that is not present in the response of real struc-
the system is used.
tures. This is due to the presence of small damping ratios at the
The examples presented in this paper will incorporate models
higher modes once this solution is devised. Studies such as those
in which concentrated (localized) plasticity is used. The application
conducted by Otani [10] demonstrate that damping models that
of these concepts to other types of models, e.g., fiber models, is the
incorporating stiffness-proportional terms provide a better corre-
subject of current research by the authors. In the first step, we pro-
lation with experimental results.
pose an approach for proper modeling of Rayleigh damping in the
C ¼ a0 M ð4Þ form explained here for beam elements whose moment gradient is
time invariant. Next, a general approach for elements whose mo-
Hall [5] suggested the elimination of mass-proportional damp-
ment gradient is time variant will be presented.
ing contribution and the incorporation of an artificial cap (or
bound) to the stiffness proportional damping component. In the
authors’ opinion, this approach would require modifications to 4.1. Structural elements with time invariant moment gradient
the numerical solution of the equations of motion. Moreover, Char-
ney [8] proposed an extension to Bernal’s approach in which the For structural elements with time invariant moment gradient, a
stiffness-proportional component of the Rayleigh-type damping two-dimensional, prismatic beam element with six degrees of free-
matrix is based only on the diagonal terms of the initial stiffness dom (see Fig. 6) is to be replaced with a two-dimensional, pris-
matrix, i.e., terms that correspond to the dynamic degrees of free- matic beam element composed of semi-rigid rotational springs at
dom, in order to avoid assigning stiffness-proportional damping to the ends and an elastic beam element in the middle (see Fig. 6
degrees of freedom without mass. and Model B in Fig. 4b). The 6-degree-of-freedom beam element
Alternatively, one can assemble a Rayleigh-type damping ma- is referred to as the original beam element and the 8-degree-of-
trix based on the tangent stiffness of the system, i.e., the damping freedom beam element as the modified beam element. If the rota-
matrix is updated at each time step, Eq. (5), in which Kt(t) is the tional stiffness at the end of the original beam element without
tangent stiffness matrix. Petrini et al. [11], based on test results transverse loads is denoted as K0 = 6EI/L (where E is the modulus
of reinforced concrete bridge piers, showed that using tangent of elasticity, I the moment of inertia, and L the length of the beam),
stiffness proportional damping is more appropriate and results in and the ratio of the rotational spring stiffness, KS, to the elastic
F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53 49

Ordinary beam element

i j 2 5
3 6
1 4
Modeling Degrees of freedom

Equivalent elastic beam element with end springs


i j
7 3 2 5 6 8
1 4
is js
Modeling Degrees of freedom

Fig. 6. Beam element and equivalent model that consists of an elastic beam element with springs at both ends.

beam stiffness, Ke, of the modified beam element is defined as mass-proportional damping. Model B in Fig. 4b is an SDOF system
n = KS/Ke then: modeled using the proposed approach and Fig. 7 shows the re-
  sponse of this system exposed to the same ground motion used
KSKe n in Fig. 5. It can be seen that when the normalized column-end mo-
K0 ¼ ¼ Ke ð7Þ
KS þ Ke nþ1 ment is plotted, unrealistic damping moments are no longer pres-
In this approach, when stiffness-proportional damping is used,
zero stiffness-proportional damping is assigned to the semi-rigid Displacement Ductility Ratio, SDOF, Model B
springs and the stiffness-proportional damping multiplier (see 6
Eq. (1)) of the modified elastic beam element is varied from b to
0 0
b . The value of b is calculated by equating the damping work done
by the elastic beam of the modified element plus the damping
Displacement Ductility Ratio

work done by the rotational springs with the damping work done 3
by the original elastic beam, i.e., Eq. (8). In this equation, WD is the
total damping work done by the element; W De is the damping work
done by the elastic beam; W Ds is the damping work done by the
rotational springs; M is the bending moment at the end of the ele-
ment; he is the rotation at the end of the elastic beam; hs is the 0
spring rotation; h_ e is the rotational velocity at the end of the elastic
beam; and h_ s is the spring rotational velocity. Given that hs = 1/nhe
0 Mass Proportional Damping
and h_ s ¼ 1=nh_ e one can utilize Eq. (9) for calculating b . Initial-Stiffness Proportional Damping (Proposed Approach)
Tangent-Stiffness Proportional Damping
      -3
1 1 1
W D ¼ 2W De þ 2W Ds ¼ 2 Mhe þ 2 Mhs ¼ 2 bK e he h_ e 0 5 10 15
2 2 2 Time (sec.)
 
1 _
þ 2 bK s hs hs ð8Þ Column-end Moment Demand, SDOF, Model B
2 1.5
b0 ¼ ½ð1 þ nÞ=nb ð9Þ
1
It is important to note that in this approach the semi-rigid
spring of the modified beam element has a post-yield stiffness
Mcolumn / M p,beam

tuned to provide the target hysteretic response at the end of the 0.5
beam. Evidently, one may advocate the use of a fully rigid spring
0
at the end of the element, which will make the estimation of b a 0
0
moot issue, i.e., b = b . However, this solution is not recommended
in order to avoid numerical instabilities in the response, especially -0.5
when piece-wise linear hysteretic models are used.
The requirement that the moment gradient on the beam ele- -1
ment be time invariant guarantees that the behavior of the modi- Mass Proportional Damping
Initial-Stiffness Proportional Damping (Proposed Approach)
fied beam element is identical to the behavior of the original beam Tangent-Stiffness Proportional Damping
-1.5
element. Another implication of this approach is that the computer 0 5 10 15
program used to conduct the numerical studies should have the Time (sec.)
capability of assigning a stiffness-proportional damping multiplier
to individual structural elements. This is a capability common to Fig. 7. Comparison between response histories of the idealized SDOF system
many computer programs currently available. (Model B) whose 10% critical damping is modeled using ‘‘mass-proportional
damping”; ‘‘initial stiffness proportional damping” based on the proposed approach
By using the approach proposed hereby for modeling Rayleigh in this paper; and ‘‘tangent stiffness proportional damping” subjected to a ground
damping of elements, the displacement response is significantly motion recording from Northridge Earthquake: (a) displacement ductility ratio and
different from that of initial stiffness proportional damping or (b) normalized column-end moment.
50 F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53

ent in responses corresponding to the proposed approach and beam element with the two end springs. The upper portion of Fig. 6
those corresponding to damping based on the tangent stiffness. shows the prismatic beam element with 6 degrees of freedom whose
The conclusion is that a conventional formulation with initial stiff- stiffness matrix, K0, can be expressed as shown in Eq. (10). In this
ness proportional damping in elements that undergo changes of equation, A is the cross sectional area, and L is the length of the beam
stiffness throughout the response overestimates the amount of element. The gray numbers around the stiffness matrix show the
damping when the response is inelastic. This constitutes a major associated degrees of freedom in Fig. 6.
drawback of the current implementation of Rayleigh-type damping
based on initial stiffness. This conclusion is consistent with the
observations made by Petrini et al. [11], who suggest that numer-
ical models based on initial stiffness proportional damping tend to
underestimate inelastic displacement demands obtained from test
conducted with reinforced-concrete piers.

ð10Þ
4.2. Structural elements with time variant moment gradient

An extension to the approach for proper modeling of viscous


damping using the Rayleigh model that can be applied to beam/col-
umn elements whose moment gradient is time variant is proposed.
The proposed approach is useful in modeling elements that experi-
ence transverse loads, and elements for which the location of the
point of inflection is expected to change considerably during their The lower portion of Fig. 6 shows the assembly consisting of an
inelastic dynamic responses. The varying moment gradient results elastic beam element and two end springs (i.e., 8 degrees of free-
in varying elastic stiffness of the structural element making the stiff- dom). The stiffness matrix of the assembly can be expressed by Ka:
ness ratio of the end springs to elastic element, n, a variable. The solu-  
Kbb Kbc
tion presented for structural components with time invariant Ka ¼ ð11Þ
Kcb Kcc
moment gradient can be interpreted as a special case of this more
general one that is applicable to structural elements for which In Eq. (11), Kbb is a 2  2 stiffness matrix that corresponds to the
changes in the location of the point of inflection are not expected. degrees of freedom #3 and #6 (i.e., to be eliminated through static
The approach for proper modeling of Rayleigh-type damping in condensation), Kcc is a 6  6 stiffness matrix that corresponds to de-
structural elements with time variant moment gradient involves grees of freedom to be kept after condensation, and Kbc and Kcb are
modifying the stiffness matrix of the elastic internal beam element 2  6 and 6  2 stiffness matrices generated through partitioning
explained previously such that the effect of fixed stiffness of the of Ka. Eqs. (12)–(14) express the components of Ka as a function of
springs at its two ends is compensated. The modified stiffness matrix stiffness coefficients Sii, Sjj, and Sij; and the elastic element’s moment
of the internal elastic beam element is obtained by equating the stiff- of inertia Ie. The parameters A and E are those defined for the pris-
ness matrix of a general prismatic beam element with the condensed matic beam element in Eq. (10). The gray numbers around the stiff-
form of the stiffness matrix of an assembly that consists of an elastic ness matrix show the associated degrees of freedom in Fig. 6.

ð12Þ

ð13Þ

ð14Þ
F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53 51

The objective is to find the values of Sii, Sjj, and Sij such that the elastic beam element with moment of inertia obtained from Eq.
condensed form of the stiffness matrix of the assembly, K ^ cc , shown (17), stiffness coefficients obtained from Eqs. (18) and (19), and two
in Eq. (15), is equal to K0. end springs with initial stiffness obtained from Eq. (16). The variation
of the stiffness coefficients Sii and Sij with respect to n is plotted in
^ cc ¼ Kcc  Kcb  K1  Kbc
K ð15Þ Fig. 8. Values of Sii and Sij asymptotically reach 4.0 and 2.0 for large
bb
values of n. The equivalent stiffness proportional damping coefficient,
By assuming predefined values for Ks, and Ie shown in Eqs. (16) 0
b , for the elastic beam element is found by using Eq. (9).
and (17), respectively, as a function of n, one can find the values of
The aforementioned modification of stiffness coefficients Sii, Sjj,
Sii, Sjj, and Sij identified in Eqs. (18) and (19).
and Sij for the elastic beam element in the assembly guarantees the
6EIe response of the assembly is identical to the elastic response of its
Ks ¼ n ð16Þ equivalent prismatic beam. The inelastic parameters of the end
L
nþ1 springs in the assembly are tuned such that the responses of the
Ie ¼ I ð17Þ two equivalent systems are identical once the springs in the
n
6ð1 þ nÞ assembly go inelastic. The backbone characteristics of the end
Sij ¼ Sji ¼ ð18Þ springs are shown in Fig. 9. The constitutive models considered
2 þ 3n
1 þ 2n for these nonlinear springs not only include strength and stiffness
Sii ¼ Sjj ¼ Sij ð19Þ degradation (represented by hp, hpc/hp, and Mc/My in Fig. 9) but also
1þn
gradual deterioration of strength and stiffness under cyclic loading
This implies that, in the elastic range of response, the original beam (represented by the parameter k), considering a peak-oriented hys-
element could be modeled with an assembly that consists of an teretic model, based on the energy dissipated in each cycle [13]. By
tuning the values of hp, hpc/hp, and k of end spring elements in order
to obtain the target inelastic behavior of the assembly, one can ob-
Stiffness coefficients for equivalent elastic beam tain proper modeling of inelastic behavior with proper modeling of
springs at both ends of elastic element damping in the component.
4
The general approach proposed in this paper for modeling Ray-
leigh damping of elements has the advantage that it can be applied
to beam/column elements whose moment gradient varies with
3.5
time. In addition, this approach includes the use of a constant
stiffness coefficient

S ii = S jj

S ij = S ji damping matrix, which avoids the additional computational effort


required to calculate damping forces based on the tangential stiff-
3
ness of inelastic members.

2.5
5. Implications for seismic performance evaluation

Improper modeling of structural damping has significant impli-


2 cations in terms of the reliability of seismic design and assessment
0 5 10 15 20 25 procedures. Inadequate estimates of force demands in current de-
n sign methods as a result of improper modeling of structural viscous
damping can lead to inefficient designs. Similarly, proper estima-
Fig. 8. Variation of stiffness coefficients Sii and Sij with n for equivalent elastic beam tion of deformation and acceleration demands in a building due
element.
to seismic excitation is fundamental to improve the reliability of

M Post-Yielding
Elastic Pre-Capping Post-Capping
Basic Parameters

Ke Initial Stiffness
Mc
My Yield Moment Capping
My Mc Point
Mc Capping moment
Yielding My
My ratio
Point
θp Plastic Hinge
Rotation Capacity
θ pc Post-Capping Rotation Ke θp θ pc
θp Capacity Ratio θ
θy θc θu
Derived Parameters

My Mc
θy = Yield Rotation Mc = My Capping Moment
Ke My

θc = θ y + θp Capping Rotation θu = θc + θpc

Fig. 9. Component backbone curve and its parameters.


52 F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53

building loss estimates that are part of performance-based assess- context of seismic collapse assessment by estimating collapse fra-
ment methods [14,15]. Appropriate estimation of the collapse po- gility curves [16]. A collapse fragility curve expresses the probabil-
tential of buildings is another important component of current ity of collapse as a function of Sa. In Fig. 11, the data points show
performance-based assessment methods. In this section, inconsis- the smallest value of Sa at which the nonlinear response history
tencies in estimation of seismic demands and collapse potential of solution of the building subjected to a given ground motion has
structures with conventional modeling of structural damping are converged, i.e., collapse capacity. The solid squares show this value
evaluated and compared to values obtained from the viscous for a building whose damping is modeled using the modeling ap-
damping modeling methodology proposed in this paper. proach proposed in this paper whereas the diamonds show the col-
An evaluation of the displacement response of a MDOF system lapse capacities obtained using a conventional Rayleigh-type
with Rayleigh-type damping based on the initial stiffness modeled damping model based on initial stiffness. The cumulative distribu-
using the approach proposed in this paper, as well as conventional tion function, assuming a lognormal distribution, of these spectral
Rayleigh-type damping, provides a more comprehensive illustra- acceleration values that correspond to structural collapse is de-
tion of the differences between both approaches. Fig. 10 depicts fined as the ‘‘collapse fragility curve” and is shown with heavy lines
mean values of 1st story drift ratio response for the 4-story rein- in Fig. 11. It can be seen that in this case conventional modeling of
forced-concrete frame used in Fig. 1 exposed to a set of 40 ground Rayleigh damping based on the initial stiffness of the system can
motion recordings. Modal damping ratios of 5% are applied to the lead to underestimation of the median of collapse capacity by 30%.
first and third modes. Each record was scaled and the peak 1st
story drift ratio responses were plotted as a function of the inten- 6. Summary and conclusions
sity of the records, Sa/g. Scaling was conducted until the limit state
of collapse was imminent. It is evident from this figure that the This study demonstrates that conventional modeling of linear
structural damping model significantly influences the estimates viscous damping via the implementation of a Rayleigh-damping
of peak drift ratio demands. matrix with initial stiffness proportional damping results in inelas-
The influence of the proposed modeling approach for Rayleigh- tic dynamic responses that exhibit unrealistic damping forces. The
type damping based on initial stiffness can also be evaluated in the presence of these unrealistic damping forces is more prevalent
when both the damping ratio and the level of inelastic behavior
of the structural system increase. For relatively small levels of
1st Story Drift Mean IDA curves inelastic behavior, deformation demands are not significantly af-
N=4, T1=0.6, Rμ = 3.0 , ξ=0.05, Peak-Oriented model fected by conventional modeling of Rayleigh-type damping based
6
on initial stiffness. This is not the case for strength demands, as
peak column moment demands may increase considerably for
5
the case of moment-resisting frames.
4
For levels of inelastic behavior consistent with structural sys-
tems approaching the limit state of collapse, the collapse capacity
Sa/g

3 of a system with Rayleigh-type damping based on initial sitffness is


overestimated, i.e., the probability of collapse for a given ground
2 motion intensity is underestimated. This is due to an overestima-
tion of the energy dissipation capacity of the system once conven-
1 tional modeling of Rayleigh-type damping based on initial stiffness
is implemented. This latter issue is significant because contrary to
0 the common notion that damping energy dissipation in structural
0 5 10 15 20 25 30 models may be de-emphasized in the inelastic range, the numeri-
1st story drift (inch) cal results from this study showed that even when cyclic deterio-
Conventional damping model Improved damping model ration and relatively large levels of inelastic behavior are present,
the contribution of viscous damping in dissipating the input energy
Fig. 10. Mean 1st story drift IDA curves for a 4-story moment-resisting frame
to the structural system is significant (on the order of 25%) as the
modeled with conventional and improved Rayleigh stiffness proportional damping.
hysteretic energy dissipation capacity of structural components is
exhausted due to damage and deterioration. Therefore, inappropri-
ate modeling of structural damping has the potential to provide
Effect of Damping Model on Collapse Fragility Curves
N = 4, T1 = 0.6, Rμ = 3.0, ξ = 0.05, Peak Oriented Model
erroneous demand prediction for structures exposed to strong dy-
0.75 namic excitations.
Collapse fragility curve from Conventional damping model: This study proposed a modeling approach developed by the
datapoints
analytical fragility curve authors to mitigate these effects and obtain more reasonable esti-
Probability of Collapse

Collapse fragility curve from Improved damping model:


datapoints
mates of seismic demands and seismic collapse capacities for plane
0.5
analytical fragility curve structural systems modeled with localized (concentrated) plastic-
ity approaches, as well as Rayleigh-type damping based on initial
stiffness. The proposed approach proved to be viable even for cases
in which cyclic strength and stiffness deterioration was present in
the response.
0.25

Acknowledgements

This study originated from discussion among the authors while


0
0 1 2 3 4 5 the first two authors of the paper were doctoral students at Stan-
Sa(T1 )/g ford University working under the supervision of Prof. Helmut Kra-
winkler. His guidance and helpful suggestions are appreciated. The
Fig. 11. Effects of damping model on collapse fragility curves. financial support provided by the National Science Foundation
F. Zareian, R.A. Medina / Computers and Structures 88 (2010) 45–53 53

through the Pacific Earthquake Engineering Research (PEER) Center [8] Charney FA. Unintended consequences of modeling damping in structures.
ASCE J Struct Eng 2008;134(4):581–92.
is gratefully acknowledged. The authors would also like to give
[9] Oliveto G, Greco A. Some observations on the characterization of structural
special thanks to Dr. Luis Ibarra for sharing his ideas and providing damping. J Sound Vibr 2002;256(3):391–415.
much valuable input at the beginning stages of this study. [10] Otani S. Nonlinear dynamic analysis of reinforced concrete building structures.
Canadian J Civil Eng 1980;7(2):333–44.
[11] Petrini L, Maggi C, Priestley N, Calvi M. Experimental verification of viscous
References damping modeling for inelastic time history analyses. J Earthquake Eng
2008;12(1):125–45.
[1] Jeary AP, Wong J. The treatment of damping for design purposes. In: [12] Leger P, Dussault S. Seismic-energy dissipation in MDOF structures. ASCE J
Proceedings, international conference on tall buildings and Urban Habitat, Struct Eng 1992;118(5):1251–69.
Kuala Lumpur; 1999. [13] Zareian F, Krawinkler H. Simplified performance-based earthquake
[2] Chopra AK. Dynamics of structures. 2nd ed. Prentice Hall: New Jersey; 2001. engineering. Report No. 169, John A. Blume Earthquake Engineering Center,
[3] Bernal D. Viscous damping in inelastic structural response. ASCE J Struct Eng Department of Civil and Environmental Engineering, Stanford University,
1994;120(4):1240–54. Stanford, CA.
[4] Medina RA, Krawinkler H. Seismic demands for nondeteriorating frame [14] Krawinkler H, editor. Van Nuys hotel building tested report: exercising seismic
structures and their dependence on ground motions. PEER Report 2003/15. performance assessment. Report No. PEER 2005/11, Pacific Earthquake
Berkeley (CA): Pacific Earthquake Engineering Research Center; 2004. Engineering Research Center. Berkeley, California: University of California at
[5] Hall JF. Problems encountered from the use (or misuse) of Rayleigh damping. Berkeley; 2005.
Earthquake Eng Struct Dynam 2005;35(5):525–45. [15] ATC-58 35% draft guidelines for seismic performance assessment of buildings.
[6] Ibarra L, Medina RA, Krawinkler H. Hysteretic models that incorporate strength Redwood City, CA: Applied Technology Council; 2005.
and stiffness deterioration. Earthquake Eng Struct Dynam 2006;34(12):1489–511. [16] Zareian F, Krawinkler F. Assessment of probability of collapse and design for
[7] Muto M, Beck JL. Bayesian updating and model class selection for hysteretic collapse safety. Earthquake Eng. Struct. Dynam. 2007;36(13):1901–14.
structural models using stochastic simulation. J Vibr Control 2008;14(1–2):7–34.

You might also like