A recurrent neural network for modeling crack growth of aluminium alloy

A new recurrent neural model for crack growth process of aluminium alloy is developed in this work. It is shown that a recurrent neural network with the feedback loops at the output layer is constructed to model the dynamic relationship between the crack growth and cyclic stress excitations of aluminium alloy. The output feedback loops in the neural model play the role of capturing the fine changes of crack growth dynamics. The Extreme Learning Machine is then used to uniformly randomly assign the input weights in a proper range and globally optimize both the output weights and feedback parameters, to ensure that the dynamics of crack growth under variable-amplitude loading can be accurately modeled. The simulation results with the averaged experimental data of the 2024-T351 aluminium alloy show that the excellent modeling and prediction performance of the recurrent neural model can be achieved for fatigue crack growth of aluminium alloys.