Optimum Design of Steel Moment Resisting Frames Subjected To Seismic Excitation

You might also like

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

Optimum Design of Steel Moment Resisting Frames Subjected to Seismic Excitation

Abstract

The main aim of this work is to develop interior penalty function based algorithm to multi-storey steel frames for minimum weight of frame members. The frames are intended for resisting lateral sway along with gravity forces. Many structural systems are used for resisting seismic (lateral) forces; however steel moment resisting frames (MRF) are considered for the present work. The design methodology incorporates codal provisions of IS 800-2007, thereby obtains the frames with optimum weight for in-plane moments with lateral support of beam elements. Strength and buckling criteria are considered as behavior constraints along with side constraints in formulating optimization problem. A computer program is developed that uses IPF for weight minimization of two-dimensional moment resisting steel framed structures. The program, written in the MATLAB, performs search, and structural design in an iterative procedure. The design examples have shown that the proposed algorithm provides an efficient tool for the practicing designers in designing steel framed structures. The program is applied to 3, 6,9,12 and 15 story (6 bay) steel moment resisting frames. The program demonstrated its capability of optimizing the weight of five medium size frames in a reasonable amount of time. The structural weights for the five frames are reduced by an average of 23.4% from their conventionally designed weight. Keywords: Interior penalty function, Steel frame, Optimum design, Minimum weight.

1.

Introduction

In the late 19th and early 20th centuries, the structural steel construction has started[]. Structural steel moment frames often are used as part of the seismic force-resisting systems in buildings to resist earthquakes. Structures designed to resist moderate and frequently occurring earthquakes must have sufficient stiffness and strength to control deflection and to prevent any possible damage. Recent specifications,codes and standards
[2,3,4]

also consider limit state of

serviceability in the design. The steel frame design requires selection of appropriate standard sections for the make up members , safety and economical consideration.

To compensate for external forces, designers will opt lateral force resisting frames which absorb and transfer these external forces to the foundation of the structure. Steel moment resisting frames (MRF) were used to resist these lateral forces, thereby prevent undesirable horizontal sway behavior under earthquake loading. These frames consisting steel columns and beams welded at ends so as to give rigidity to resist mainly moments due seismic loading.

Mostly, MRFs are analyzed using the equivalent lateral force (ELF) procedure as specified by code provisions[2-6] .The following sections present structural model generation, ELF procedure and optimization technique respectively for the minimization of objective function.

2.

Model Building Frames were of

The model building moment resisting frames were developed

rectangular symmetric unbraced. The three, six, nine, twelve and fifteen storey frames considered; have the same typical floor size of 24.00mx42.00m. The typical building floor plan and frame elevations are shown in Figure1. The

model buildings are proposed at Hyderabad (seismic zone II). The assumptions made are: Soil is of medium stiff and the entire building is supported on isolated rigid column footings. Frames are infilled with lightweight sheets. Lumped weight due to dead loads is 12 kN/m2 on floors and 10 kN/m2 on the roof. Floors are to cater for a live load of 4 kN/m2 on floors and 1.5 kN/m2 on the roof[1]. Start from here 3. ELF Procedure

The ELF method is the most commonly used analysis procedure, and has been codified in various forms since perhaps the 1960s. In this method, design seismic forces are determined by linear static analysis of the structure. In effect, time-varying inertial forces are replaced by equivalent static forces that are applied to each floor level. The forces applied to each floor equilibrate the base shear and are distributed over the height of the structure in proportion to the weight of the floor and its height. The effects of seismic actions are combined with the effects of gravity loads according to the load combinations of the governing code to determine the member design forces. Story drifts due to the lateral forces computed must not exceed limits that are specified. The concept employed in equivalent static lateral force procedures is to place concentrated static loads on a structure with magnitudes and directions that are closely approximate the effects of dynamic loading caused by earthquakes. These lateral forces tend to occur at floor and roof levels in building frames, where concentration of mass is assumed. Furthermore, lateral forces tend to be larger at higher elevations in a structure. Thus, the greatest lateral displacements and the largest lateral forces often occur at the top level of a structure (particularly for tall buildings). These effects are modeled in

equivalent static lateral force procedures by placing a force at each story level in a structure, as shown in Figure 2. In general, the distribution of lateral

storey forces is associated with the first (fundamental) mode of vibration. The summation of the lateral storey forces must be equivalent to the base shear force (Vb) applied to the structure due to seismic ground motion.

IS 1893(Part-1)-2002 addresses the probability of significant seismic activity in various locations by categorizing geographic regions of India as seismic zones II through V. seismic zone II indicates a geographic location where very small seismic activity is expected to occur. Seismic zone V indicates a geographic location with a high probability of significant seismic activity. The equivalent static force procedure specifies the following formula for calculating base shear Vb=Ah Ws where Ah (2) (1)

Ah= design horizontal seismic coefficient associated with structural sensitivity to the velocity and acceleration. Ws = Seismic weight of the structure. It is the sum of total dead load and appropriate amounts of specified imposed load. Z = Zone factor for the maximum considered earthquake and service life of structure in a zone. The factor 2 in the denominator of Z is used so as to reduce the maximum considered earthquake to design basis earthquake. I= Importance factor; used to obtain the design seismic force depending on the functional use of the structure.

R= Response reduction factor; by which the actual base shears force that would be generated if the structure were to remain elastic during its response to the design basis earthquake.

(Sa/g) =Structural response factor.

The R factor is intended to account for inelastic structural behavior and the ability of a structure to deform and dissipate energy without failing. Since all R factors specified are greater than unity (R > 1.0), the R factor effectively reduces the calculated base shear (Vb) depending on the ductility of a structure. The code provides the following simplified equation for estimating fundamental time period (Ta) of the steel structures;

Ta = 0.085 (h)0.75

(3)

Figure 3 shows the proposed 5 percent damping spectra for rocky,medium and soft soils sites to obtaining spectral values. Lateral forces that counteract the base shear are assumed to act at each storey level of the structure. The magnitude of each storey force (Fi) is determined from the following formula:

Fi=Vb[(wihi2)/(wihi2)] where

(4)

wi= the portion of the building weight assumed to be lumped at level i. wi typically includes the total weight of the floor or ceiling/roof system at level i, plus half the weight of the vertical elements (walls; columns) located immediately below level i and half the weight of the vertical elements located immediately above level i hi = height from the base of the structure to floor level i Lateral load distributions of equivalent static forces with height for 3,6,9,12 and 15 storey frames are presented in Table 1.

Mathematical Problem Formulation

This section deals with the Interior Penalty Function (IPF) method of optimization technique, which is applicable to the solution of the constrained optimization problem Find X which minimizes f (X) Where, subject to gj (X) 0, j = 1, 2, . . . ,m hk(X) = 0, k = 1, 2, . . . , p where
s

(5)

is the unit weight of the member material, Li is length of i th

member(column or beam) and Ai is the cross sectional area of ith member. 5 Structural Optimization

The constrained problem is solved as a sequence of unconstrained minimization problems. In the IPF method, a new function ( function) is constructed by augmenting a penalty term to the objective function. The penalty term is chosen such that its value will be small at points away from the constraint boundaries and will tend to infinity as the constraint boundaries are approached. The function defined originally by Carroll [7.14] is given in equation (6).

(x,rk)=f(x)- rk (1/gj(x)) X =set of design variable

(6)

f(X)= objective function expressed as a function of weight to be minimized, gj(X)= set of inequality constraints. hk(X)= set of equality constraints. where (x,rk) is known as penalty function. The second term on the right side of the equation rk1/gj(X) is called penalty parameter. To solve numerically and for

ease of discussion, the constrained function gj(x) is normalized to the standard form. If the unconstrained minimization of function is repeated for a sequence of penalty parameter rk (rk=1, 2, 3, . . . ..) the solution may be brought to convergence to that of original problem stated.The optimization problem is an iterative procedure ,whos flow chart is presented in Figure 4. The design variables, objective function and the constraints are briefly explained as follows; 5.1 Design variables: x={x1, x2, x3, x4} T (7)

x1=flange thickness ,x2= width of the flange,x3=web thickness,x4=depth of web for the selected welded steel section as shown in Figure 5.

The present work considers the design of members subjected to combined forces, such as shear force and bending for beam elements and axial compressive force and bending moments for column elements of MRFs. The constraints function includes both side and behavior constraints described as follows; 5.2 Side Constraints: g1=-x10; g2=-x20; g3=-x30; g4=x2-8.4; g5=x4-2.4 x20; where =(250/fy); fy= yield stress of the member material (8.6) (8.1) (8.2) (8.3) (8.4) (8.5)

5.3 Behavior Constraints Members subjected to combined axial compression and biaxial bending, the section strength behavior constraints for plastic sections are;

(8.7)

Members subjected to combined axial compression and biaxial bending, the overall member strength behavior constraints are; (8.8)

(8.9) where Cmy,Cmz, = equivalent uniform moment factor as per Table 18 of IS 800-2007; P = factored axial compressive load; My,Mz= maximum factored applied bending moments about y-y and z-z axis of the member, respectively; Py.Pz=design strength under axial compression as governed by buckling about minor (y-y) and major (z-z) axis respectively; Mdy,Mdz, =design bending strength about y-y (minor) or z-z (major) axis considering laterally unsupported length of the cross-section. Cmy,Cmz=Moment amplification factor about respective axes Ky,Kz and KLT= bending stress reduction factors to account for lateral torisonal buckling, 6 Results and Discussions The analysis method and optimization algorithm, explained in the previous sections, were applied successfully to 3-, 6-, 9-, 12- and 15-storey steel MRF, thus obtained results are presented in the form of tables and graphs. The analysis results (Design Forces and Moments for 3,6,9,12,and 15 storey steel MRFs) computed are presented in Table 2 to Table 6. Also, the optimum design variables and the values of objective function are presented in Table 7 to Table 11.

Figure 6 illustrates the variation of achieved minimum weight using IPF method for different values of penalty parameters. All penalties are able to reach the same minimum weight.

Conclusions

You might also like