Prof. Jacob Fish
Department of Civil Engineering and Engineering Mechanics
Columbia University

Predictive Multiscale Paradigm for Integrated Manufacturing and Component Design

ABSTRACT: Recent advances in composite manufacturing and multiscale modeling have opened the door to a new paradigm in which process design, material behavior, and structural performance are treated as an integrated system, rather than isolated stages of engineering analysis. This lecture introduces a unified computational framework bridging manufacturing processes and component-level performance through both physics-based and data-driven model-reduction strategies. At the manufacturing scale, we present nonlinear homogenization methods for ultra-fast resin transfer molding, enabling accurate simulations of mold filling under challenging conditions, including low resin viscosity and high injection pressure. A complementary multiphysics framework for overmolded thermoplastics rigorously couples heat conduction, crystallization, deformation, and nanoscale polymer diffusion, providing the first predictive link between processing conditions, interfacial microstructure, and ultimate bond strength. At the component scale, an Eigenstate-Based Homogenization (EBH) approach is presented for efficient multiscale modeling of inelastic heterogeneous materials. EBH captures fine-scale plasticity and damage evolution through eigenmode representations and controllable fixed-point solvers, achieving finite-element accuracy at less than five percent of the computational cost. Finally, the Weighted Eigenseparation-based Residual framework extends model reduction to interface failure, enabling accurate and scalable simulation of decohesion in composites and polycrystalline materials. Together, these advances establish a cohesive digital thread that connects manufacturing processes, microstructural evolution, and structural reliability. This paves the way toward predictive, sustainable, and rapid development of composite components.

BIOGRAPHY: Dr. Jacob Fish is the Robert A.W. and Christine S. Carleton Professor and Chair of Civil Engineering and Engineering Mechanics at Columbia University. Over the past 35 years, he has made foundational contributions to multiscale computational science and engineering, including scale-separation–free homogenization, reduced-order, stochastic multiscale methods, temporal and dispersive multiscale approaches, and the coupling of thermo-chemo-electro-mechanical processes across multiple spatial and temporal scales. His pioneering work in data-physics-driven multiscale modeling, integrated manufacturing product design simulation, and hybrid computational frameworks for inelasticity has significantly advanced both academic research and industrial practice. Dr. Fish founded Multiscale Design Systems, whose Multiscale Designer software has been adopted by more than 250 major companies worldwide and is regarded as the gold standard in predictive materials modeling across aerospace, automotive, energy, electronics, manufacturing, and healthcare sectors. He is the recipient of the John von Neumann Medal (United States Association for Computational Mechanics [USACM], 2021), the Grand Prize (Japan Society for Computational Engineering and Science, 2018), the Computational Mechanics Award (International Association of Computational Mechanics [IACM], 2010), and the Ted Belytschko Medal (USACM, 2005). Dr. Fish is a Fellow of the American Academy of Mechanics, USACM, and IACM and has authored more than 280 publications and several textbooks including, “A First Course in Finite Elements and Practical Multiscaling.” He currently serves as Vice President for the Americas of the IACM, Founding Editor-in-Chief of the International Journal of Multiscale Computational Engineering, and Director of Columbia’s Initiative for Computational Science and Engineering.