A thermodynamically consistent physics-informed deep learning material model for short fiber/polymer nanocomposites

arXiv:2403.18310v1 Announce Type: new
Abstract: This work proposes a physics-informed deep learning (PIDL)-based constitutive model for investigating the viscoelastic-viscoplastic behavior of short fiber-reinforced nanoparticle-filled epoxies under various ambient conditions. The deep-learning model is trained to enforce thermodynamic principles, leading to a thermodynamically consistent constitutive model. To accomplish this, a long short-term memory network is combined with a feed-forward neural network to predict internal variables required for characterizing the internal dissipation of the nanocomposite materials. In addition, another feed-forward neural network …

abstract ambient arxiv behavior consistent cs.ai cs.ce cs.lg cs.na deep learning material math.na physics physics-informed type work