A MATLAB-based tool for nonlinear structural analysis has been developed for vibration control simulation of seismically excited buildings. The tool coni ts of two parts: one assemble the structural models to be analyzed while the other performs time-step integration of the governing equations of motion. Both a spread plasticity model and a concentrated p.lasticity model are available to model the behavior of the structural members.
Copyright:
Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online from Scribd
Ohtori - Spencer - MATLAB Based Tool For Nonlinear Structural Analysis
A MATLAB-based tool for nonlinear structural analysis has been developed for vibration control simulation of seismically excited buildings. The tool coni ts of two parts: one assemble the structural models to be analyzed while the other performs time-step integration of the governing equations of motion. Both a spread plasticity model and a concentrated p.lasticity model are available to model the behavior of the structural members.
Original Description:
Original Title
Ohtori_Spencer_MATLAB based tool for nonlinear structural analysis
A MATLAB-based tool for nonlinear structural analysis has been developed for vibration control simulation of seismically excited buildings. The tool coni ts of two parts: one assemble the structural models to be analyzed while the other performs time-step integration of the governing equations of motion. Both a spread plasticity model and a concentrated p.lasticity model are available to model the behavior of the structural members.
Copyright:
Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online from Scribd
Download as pdf or txt
0 ratings0% found this document useful (0 votes)
519 views6 pages
Ohtori - Spencer - MATLAB Based Tool For Nonlinear Structural Analysis
A MATLAB-based tool for nonlinear structural analysis has been developed for vibration control simulation of seismically excited buildings. The tool coni ts of two parts: one assemble the structural models to be analyzed while the other performs time-step integration of the governing equations of motion. Both a spread plasticity model and a concentrated p.lasticity model are available to model the behavior of the structural members.
Copyright:
Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online from Scribd
Download as pdf or txt
You are on page 1of 6
7 Ai Buse 2900
13° A9¢E ENS. MECH DNS
CONFRGNCE, BAIT) MORE, AD
1999
A MATLAB-Based Tool for Nonlinear Structural Analysis
Y. Ohtori! and B. F. Spencer, Jr., M. ASCE
University of Notre Dame, Notre Dame, IN 46556-0767
[email protected], [email protected]
Abstract:
‘A MATLAB-based tool for nonlinear structural analysis has been developed for vibration control
simulation of seismically excited nonlinear buildings. The MATLAB environment has been chosen
because it is widely available and has very powerful analysis routines. The tool consists of two
parts: one assembles the structural models to be analyzed, while the other performs time-step inte-
gration of the governing equations of motion. Both 2 spread plasticity model and a concentrated
plasticity model are available to model the behavior of the structural members. The tool has been
implemented as a system function within SIMULINK. Finally, the accuracy of the tool is verified
against IDARC2D.
Introduction
Recently many control strategies and devices have been proposed for mitigating structural damage
and/or securing comfort during severe earthquakes. Each has certain advantages, but itis difficult
to compare and/or evaluate the efficacy of their relative merits. To do this, several benchmark prob-
lems have been developed (Spencer, et al., 1998a,b, 1999) in which the structures were assumed (0
be linear elastic. Yet itis well known that structural members exceed the elastic limit during large
earthquakes, such as the 1994 Northridge and the 1995 Kobe earthquakes. To evaluate the control
efficiency during such severe earthquakes, nonlinear effects of the members should be taken into
account.
This study reports on the development of a MATLAB-based nonlinear structural analysis tool for
simulation of seismically excited structures. The MATLAB environment is chosen because it is
widely available and has very powerful graphics and numerics routines. In the program, a spread
plasticity model and a concentrated plasticity model, which have been considered in IDARC2D
(Valles, et al., 1996), are used for expressing the nonlinear behavior of members. The analysis pro-
gram has been implemented as a system function in MATLAB's SIMULINK so that the dynamic
behavior of the structure can be effectively calculated
‘Nonlinear Structural Analysis,
In nonlinear dynamic analysis of structures the incremental equation of motions are often solved
using the Newmark-B method in conjunction with the pseudo force method, which is reviewed in
detail in Subbaraj, ct al. (1989). The incremental equation of motion for the nonlinear structural
system takes the following form
jecric Power tndustry in Japan.[a}{ a} + [C]{ Au} + [K1{Au}= - [MIP }AE, + PAP} + Coo f AP ay
where, [M), [C], and [A] are the mass, damping and stiffngss matxic6s of the building, {Au}is
the incremental response vector, A¥, is the ground acceleratipartficrement, {AF} is the incremen-
tal control force, {1°} is a load vector to the structure ] is a vector defining how the control
forces produced by attached control devices are appied to the structure, C..,, is a correc
efficient for the unbalanced forces, and {AF} is the vector of the unbalanced forces. The unbal-
anced force is the difference between the restoring force evaluated using the hysteresis model and
the restoring force assuming constant linear stiffness at time ¢ during the time interval (t~2+ Ar).
‘This unbalanced force is handled as pseudo force at the next time step.
Because the floor slab is assumed to be rigid in the horizontal plane, the nodes associated with each
floor have the same horizontal displacements. Therefore, the dependent (slave) horizontal DOFs
‘on each floor slab can be expressed in terms of a single active horizontal DOF. Thats, the displace-
ments {Aw} can be expressed using those of the active nodes, { Ati,.¢} - This relation can be writ-
ten as following equation.
{Au} = [Te] Auge} Q)
in which [Tg] is a transformation matrix for expressing the full response vector in terms of the
active degrees of freedom. The following two expressi rBproposed by Newmark are employed to
solve Eq. (1):
= {wu u},> Aloe ii i (3
Wealunrics (hegay = tut Arta), + (00) -B) ta), + BE.) @
See rear = rear? AM}, +1, ad @
in which Af is the calculation time interval, { },4 4, and { }, are the responses at 1+ Ar and ¢,
respectively, and B, y are Newmark parameters. Usually, B=1/4 and y=1/2 are adopted. Substi-
tuting Eq. (3) and Eq. (4) into Eq. (1) and introducing the relation in Eq. (2), the following equa-
tions can be derived:
{AFp} 6)
Ez
where
Near Face
(clei + aK 6
on SEAR SMT
LINEAR PART
a Q
~tmieryax,+(Lo+ (1-1 )acr ei,
8 \2I Ls
(ap pata) es
corr AF,
3
+ (Bal+ FICT), ar PHEAPY +,UC) = Urge] ®
‘The damping matrix [C'}-tn Eq. (8) is defined based on the reduced system. Modal damping, mass/
stiffness proportional damping and Rayleigh damping can be used in the analysis. Note that the lin-
ear part [Ko] of thestiffness and the nonlinear part [AK] of the stiffness are separately handled
in the program to avoid matrix reconstruction.
Both a spread plasticity model and a concentrated plasticity model have been implemented for
modeling the material nonlinearity in the structural members. The coefficients of the spread plas-
ticity model were derived (Lobo, 1994) and then rewritten for improved numerical stability in Vall-
al. (1996). These models use a flexural formulation to derive the element stiffness of the
nonlinear elements. For the spread plasticity model, the following stiffness matrix is used for ex-
pressing the material nonlinearity
is dian tal (°4 »
ob Kon kon] On
Where o
I2EToEI Ely a
Ean = pat mw Gal” + 12EFQEIAE Ty) (10)
* I2EIgEI El, 2
Pot AU apGAzL? + 12EIgEl ,ETy) (i)
Pak
I2EIgE I.E
‘beg FU anG Arh? + 2ETEI El) (12)
Le
LEGS 00 -Fa0)+ T2EMGEI,Ela( Pant fg
fap) (13)
Pixs = 4EVEIy + (Ely— El, JEM (601, — 400%, + oF) + (Ely — Ely)EI 0 (4
(5)
a ae uel
Pqy > ~2EI EI ~ (El ~ Ely El y(204 — 064) ~ (Eg ~ Ely) El ,(205, - 024)
- 3 2
tay = SEI Ely + (Ely — El) El g0ty + (Ely - Ely) El ,(60y 4023 + 03) 16)
in which M,, and My are the moments at the ends A and B of the element, respectively; 8, and
Oy are the rotations at the ends A and B; EV, and Ely, are flexural stiffness atthe ends of the mem-
ve Ma
fcoNEATRAED
PASTICNY.
bers at A and B, respectively; Ef, is the center stiffness of the member; GA, is the shear stiffness
of the member; L is the length of the member; 0, and O, are yield penetration parameters which
are determined by using yield penetration model. Further details regarding the spread plasticity
‘model are presented in Valles, et al. (1996). With the concentrated plasticity model the members
are assumed as elastic body and yielding occurs only at the ends. This is done by adding rotational
springs at the ends of the members.
Matlab Implementation
‘The MATLAB-based tool shown in
Fig. | consists of 2 programs and 3
data sets. One program assembles
the mass, damping and stiffness ma-
trices and calculates the control pa- sm
rameters for matrix handling and so
forth. The other program, which is
implemented as a SIMULINK sys-
tem function (S-Function), performs us
the nonlinear dynamic analysis. Pare
Roan Siac
"Anup Bock
Tal Daa ree
The tool is implemented into a vi nt a
bration control simulator for seismi-
cally excited nonlinear buildings. Figure 1. Schematics of MATLAB-Based Tool.
Figure 2 shows the vibration control
simulator, in which a nonlinear anal-
ysis block accepts earthquake record
a FS a]
oe on
ico emeitegen
ae een
Py Ease Fee ere cae
‘ teats
a esar tsa
ee
temsniertneare
rine, Ha sow renlel
Ty] Tea
+ <4 ones EL
ie ea
or
Figure 2. SIMULINK Block Diagram for Vibration Control Simulator.data and control forces produced by control device model. The block evaluates the nonlinear dy-
namic responses using this information and the structural data. The response is marched in time
using the Newmark-B.. It then passes the structural responses for evaluation, as observed data at
sensor positions and as raw data for connection points of passive and semi-active devices,
Example Calculation
‘The 2-bay 2-story building model shown in Fig.
3 is used for example calculations and verifica-
tion. The structure is assumed to be a concrete
frame, so the spread plasticity model is em-
ployed. The flexural stiffness (EI) of the col-
umns and beams are 20E8 and 1,0E8
KN-cm?, respectively. The axial stiffness of
the columns is 4,.0B8 KN. The yielding curva-
ture of the members is 5.0B-4 cm-l. The post
yielding stiffness of the columns and beams is
10% of the initial stiffness. The N-S component
(Fig. 4) of the El Centro earthquake strong mo-
tion recorded at the Imperial Valley Irrigation
District substation on May 18th, 1949, is used.
for the example calculations. The magnitude of
the earthquake record is adjusted to 1500 gal to
ensure inelastic deformation in the members.
Mass proportional damping of 2% is used for
this example, The time interval is set at 0.001 :
sec. The comparison results are calculated by
using IDARC2D under the same calculation
conditions.
Figure 4. El Centro Record.
The first 2 natural periods of the structure evaluated by the MATLAB-based tool are 0.8842 sec.
and 0.2353 see. The differences with the results of the IDARC2D are negligible. Figure 5 shows
the time history of the response acceleration and displacement. Excellent comparison is seen. The
hysteresis loops of the beams and columns are shown in Fig. 6. Small differences can be seen in
both of the hysteresis loops. Minor differences in how the programs calculate the hysteretic forces
and moments cause this discrepancy. There effect on the response of the building is minimal.
Conclusion
‘The MATLAB-based nonlinear structural analysis tool simulation of the seismic response of build-
ing structures has been developed and implemented as a SIMULINK system function. The accu-
racy of the tool has been verified with the IDARC2D. Any control algorithms and control devices
for active, passive and semi-active control methods for the seismically excited nonlinear structure
can be implemented into the vibration control simulator, allowing for the effectiveness of various
control method to be evaluated and comparedAsslinsiick —B
a cma
Figure 5. Responses at the Top of the Building.
()Colums No.1 (Bouom)
Figure 6. Comparison of Hys-
teresis Loops with IDARC2D.
Acknowledgement
The authors gratefully acknowledge the partial support of this research by the National Science
Foundation under grants No. CMS 95-00301 and CMS 95-28083 (Dr. S. C. Liu, Program Director)
and the Central Research Institute of Electric Power Industry in Japan. We also wish to thank A.
M. Reinhorn, S. H. Kunnath and M. V. Sivaselvan for their assistance during this study.
Reference
Lobo, R.F. (1994), “Inelastic Dynamic Analysis of Reinforced Concrete Structures in Three Di-
mensions.” Ph.D. Dissertation, Dept. of Civil Engrg., State Univ. of New York at Butfalo.
Spencer Jt, B.E, Dyke, S.J. and Deoskar, H.S. (1998a,b). “Benchmark Problems in Structural Con-
trol — Part I: Active Mass Driver System; Part I: Active Tendon System.” Earthquake Engrg.
and Struct. Dyn. , Nol. 27, No. 11, pp-1127-2247.
Spencer Jr., B.F., Christenson, RE, and Dyke, S.J. (1999), “Next Generation Benchmark Control
Problems for Seismically Excited Buildings.” Proc., 2nd World Conf. on Structural Control ,
(T. Kobori, et al., eds., Wiley), Vol2, pp.1135-1360.
‘Subbaraj, K. and Dokainish, M.A. (1989), “A Survey of Direct Time-Integration Methods in Con
putational Structural Dynamics - Il. Implicit Method.” Comp. and Struct., Vol.32, No. 6.
pp.1387-1401.
Valles, R.E., Reinhorn, A.M.,.Kunnath, S.K., Li, C. and Madan, A. (1996). “IDARC2D Version4.0:
‘A Computer Program for the Inelastic Damage Analysis of the Buildings.” Technical Report
NCEER-96-0010, Nat. Ctr. for Earthquake Engrg. Res., Buffalo, New York.
MATLAB (1997), The Math Works, Inc., Natick, Massachusetts