ESTIMATING SHEAR STRENGTH OF RECTANGULAR SHORT REINFORCED CONCRETE COLUMNS USING GENETIC PROGRAMMING

The numerous collapse of reinforced concrete (RC) build-ings reported during destructive earthquakes, happeningaround the world, are pertaining to short column fail-ure. Shear collapse mode is the most devastating causesof failure in short RC columns due to intensive seismiccyclic loads. Since the code-based design equation forestimating the strength of these structural elements isbasically based on pragmatic models, they cannot ap-propriately estimate the resistance over a wide rangeof involved variables. Evaluating collapse capacity ofthese structural elements has been an attractive topicof research for the last few decades. The major goal ofthis paper is to propose a model for shear strength ofbeam-column with shear span-to- depth ratio less than2.5. To complete few available experimental samples,the purposed ABAQUS general nite-element methodsoftware is used to model 52 numerical samples to covera wide range of various parameters involved in shearresistance. According to the performed experimentalworks, the e ective shear strength parameters are axialforce, transverse and longitudinal reinforcement, columnaspect ratio, concrete compressive strength, and ductil-ity. The nite-element models are validated against ex-perimental data. To develop a closed form solution forevaluating shear strength, Genetic Programing (GP) isused. GP is an automated method for creating a workingcomputer program from a high-level problem statementof a problem. GP starts from a high-level statementand automatically creates a computer program to solvethe problem. This method is successfully used in vari-ous elds, including symbolic regression, image process-ing, network protocol design, patching buggy codes androbot control.The proposed GP model is compared against shearstrength provided by the existing relations of two majorcodes, e.g., ACI and EC2. Compared with experimentalresults, the average errors from the application of GP,ACI and EC2 are %15, %45, and %43, respectively. Thisshows that the proposed equation using GP is a preciserelationship for the prediction of the shear capacity ofrectangular short RC columns.