Expertise: Scientific computing using the Finite Element Method, software programming (C++, Python), machine learning with Deep Neural Networks, high performance computing, applications to mechanical engineering.
Degree: MS (Applied) Mathematics, PhD Mechanical Engineering, University of Michigan
Greg is a computational science applications specialist with a background in computational materials science. He has developed scientific software in C++ for modeling solid mechanics and phase field dynamics in materials, using the finite element method. He has experience with the computational libraries PETSc, Trilinos, deal.II, and PetIGA. His research has involved using machine learning techniques, including deep neural networks and the TensorFlow library, to learn free energy surfaces to predict material microstructure.