Prediction of Remaining Life of Structure Based on Phase-Field Method

In this work, a new approach for forecasting the remaining useful lifespan (RUL) of mechanical structure is described, which combines phase-field method (PFM) technology and neural network (NN) technology. Firstly, brittle fracture damage model established according to PFM. And sample data, are strain value at observation locations model, obtained by numerical simulation. Then, set input into NN as training sample. mathematical can forecast structure’s RUL, training. Finally, we assess rationality proposed standpoint means simulation cracked plate. It demonstrated that presented structure, PFM technology, has high performance.