Abstarct: The delamination phenomenon of multi-laxyer composites with curved fibers and the prediction of the necessary force for the initiation of subsequent crack growth have been analyzed and modeled in the presents study. In order to predict the force required to initiate crack growth in multi-laxyered composite double-sided clamping beams with curved fibers, an inverse method is presented baxsed on neural network and genetic optimization algorithm. For validation purposes, the results of the proposed inverse method were compared with the results of the finite element and sticky region models. According to the results of this research, it can be stated that the proposed inverse method has good accuracy and is able to predict the force required to initiate crack growth in multi-laxyered composite double-sided clamping beams with curved fibers. The desired neural network consists of six inputs, which include pre-crack length, number of laxyers, curved fiber rule, fracture toughness, and bending stiffness of the desired multilxayer composite beam, and the output of the neural network for these six inputs includes the force required to initiate crack growth in is a multi-laxyer composite beam with curved fibers. It can also be stated that the force required for crack growth in the desired multilxayer composite beam increases with the increase in the number of laxyers, and with the increase in pre-crack length of the desired multilxayer composite beam, the force required for crack growth in the desired multilxayer composite beam is reduced. It should be noted that the angle of the fibers in the multi-laxyer composite beam affects the force required to initiate crack growth in the multi-laxyer composite beam. The results of the present study revealed that the input and output angles of the fibers compared to the angle of the fibers in the center of each laxyer has a greater effect on the delamination phenomenon and the force required for crack growth in the multi-laxyer composite beam.