Abstract
In this study, we join the thermoplastic composite using thermoplastic composite fastener (TPCF), and a physics-guided neural network is designed to predict the fatigue life of the all-composite joint. The experimental results show that TPCF is only 19.0% of the weight of the titanium rivet. Under a proper joining pressure of 1.50 MPa, the all-composite joint is well connected by an ‘S-shape’ interlayer, the fatigue resistance is improved. After hot-press joining, the lower cooling rate of 1 °C/min prolongs the crystallization process of TPCF and composite hole wall, thus enhancing the fatigue life to 105.8 cycles. The experimental results are summarized as physics constraints guiding the neural network. Similar fatigue data are migrated by producing a significant constraint. The loss function is the negative logarithm of the likelihood, which can predict both mean and deviation of fatigue life. The predicted results agree well with the experiments.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
