Application of Gaussian process autoregressive models for capturing the time evolution of microstructure statistics from phase-field simulations for sintering of polycrystalline ceramics

摘要: While phase-field models have been demonstrated to be highly versatile in performing physics-based simulations of a large variety materials phenomena involving microstructure evolution (e.g., phase transformation, recrystallization, sintering), they are not practical for rapid exploration the process design space due their high demand computational resources. The extraction reliable and robust reduced-order from datasets produced by such sophisticated continues an unsolved problem. Recent advances fast computation comprehensive set statistics data-driven low-dimensional representations using principal component analyses (PCA) resulted successful practically useful connecting effective properties exhibited material. In this paper, we explore first time, viability these establishing capable learning important details predicted computationally expensive models. More specifically, will applying Gaussian autoregressive (GPAR) used fields signal processing problems evolution. This accomplished specific case study dealing with time porous microstructures sintering polycrystalline ceramics.