Application of Evolutionary Algorithm Technique to Minimize Torsion for Plan and Vertical Asymmetrical RC Buildings

Abstract

Plan and vertical irregularities in reinforced concrete (RC) medium-rise buildings pose significant challenges in seismic analysis due to their susceptibility to torsional effects. Irregularities such as mass, stiffness, and geometric discontinuities cause eccentricities between the building’s centers (i.e.; mass and rigidity centers), inducing lateral and rotational responses that compromise structural safety. This work presents the use of evolutionary algorithms to minimize the stiffness eccentricity and thereby optimize the seismic torsional behavior for three-dimensional asymmetrical RC buildings. In order to achieve the torsional stiff structures, efforts are being made to attain an ideal preprogrammed design of the rotational deformation of the diaphragms of the proposed RC structures with asymmetries (in-plan and vertical). Both static as well as dynamic forces cause the torsional drift of floor diaphragms and they can be precisely described with regard to the design variables for the orientation of vertical structure elements. To illustrate the effectiveness and viability of the suggested optimization strategy, two examples were attempted successfully. The procedures outlined in the current seismic standards were followed to evaluate these structures' performance. Finite element mathematical models were employed to perform seismic assessments of said asymmetrical buildings. The Genetic Algorithm model was used to address the orientation optimization problem. Outcomes of the proposed study confirmed that the proposed Genetic Algorithm (GA) can solve the orientation optimization problem for three-dimensional reinforced concrete asymmetrical buildings in an efficient and optimal manner. Linear static analysis method, though easy and computationally efficient, fail to capture the torsional responses accurately. Linear dynamic analysis method, incorporating modal contributions, provide better accuracy but are limited in addressing inelastic behavior of the structural elements. Nonlinear dynamic analysis method emerges as the most reliable technique, offering precise modeling of torsional amplification under real earthquake records. The study is found to be very productive and useful as the stiffness eccentricity and torsional irregularity ratio get reduced significantly by adopting this method.