Physics-embedded graph network for accelerating phase-field simulation of microstructure evolution in additive manufacturing

Abstract

The phase-field (PF) method is a physics-based computational approach for simulating interfacial morphology. It has been used to model powder melting, rapid solidification, and grain structure evolution in metal additive manufacturing (AM). However, traditional direct numerical simulation (DNS) of the PF method is computationally expensive due to sufficiently small mesh size. Here, a physics-embedded graph network (PEGN) is proposed to leverage an elegant graph representation of the grain structure and embed the classic PF theory into the graph network. By reformulating the classic PF problem as an unsupervised machine learning task on a graph network, PEGN efficiently solves temperature field, liquid/solid phase fraction, and grain orientation variables to minimize a physics-based loss/energy function. The approach is at least 50 times faster than DNS in both CPU and GPU implementation while still capturing key physical features. Hence, PEGN allows to simulate large-scale multi-layer and multi-track AM build effectively.

Ultrafast and accurate prediction of polycrystalline hafnium oxide phase-field ferroelectric hysteresis using graph neural networks

“…With the objective to accelerate phase-eld simulations, machine-learning approaches have been explored to build surrogate models. 13,[26][27][28][29][30][31][32][33][34][35][36][37] Many of these models are based on dense neural networks, 31,33 recurrent neural networks, 29,32 or convolutional neural networks (CNNs). [26][27][28]32 CNNs typically operate on Euclidean data, conventionally utilizing the microstructure as input in the form of a 2D or 3D array.…”

Accelerate microstructure evolution simulation using graph neural networks with adaptive spatiotemporal resolution

“…The graph-based architecture allows for more flexible geometric representations when compared to CNNs [21]. Very recently, Xue et al adopted a graph-based representations to solve multi-physics PFM of microstructure evolution in additive manufacturing with 50x speedup [37]. Qin et al developed a long-short-term-memory based surrogate model of 2D epitaxial grain growth in additive manufacturing conditions [38].…”

Machine Learning Surrogate Model for Acceleration of Ferroelectric Phase-Field Modeling

“…They took advantage of the inductive biases inherent to graph neural network architecture and used a fully unsupervised loss based on the Landau energy. 20 Currently, the use of physical constraints in machine learning is being explored in physics-informed neural networks. Since free energy is at the core of phase-field approaches, energyinformed neural networks have been developed that directly minimizes it as a training target.…”

“…In another relevant work, Xue et al embedded the phase-field model in a graph neural network for metal additive manufacturing. They took advantage of the inductive biases inherent to graph neural network architecture and used a fully unsupervised loss based on the Landau energy . Currently, the use of physical constraints in machine learning is being explored in physics-informed neural networks.…”