Neural network prediction of the ultimate capacity of shear stud connectors on composite beams with proled steel sheeting

In this paper, the eciency of dierent articial neural networks (anns) inpredicting the ultimate shear capacity of shear stud connectors is explored. experimentaldata involving push-out test specimens of 118 composite beams from an existing databasein the literature were used to develop the ann model. the input parameters aecting theshear capacity were selected as sheeting, stud dimensions, slab dimensions, reinforcementin the slab and concrete compression strength. each parameter was arranged in an inputvector and a corresponding output vector, which includes the ultimate shear capacity ofcomposite beams. for the experimental test results, the ann models were trained andtested using three layered back-propagation methods. the prediction performance of theann was obtained. in addition to these, the paper presents a short review of the codesin relation to the design of composite beams. the accuracy of the codes in predicting theultimate shear capacity of composite beams was also examined in a comparable way usingthe same test data. at the end of the study, the eect of all parameters is also discussed.the study concludes that all ann models predict the ultimate shear capacity of beamsbetter than codes.