Mobile menu

                                     

7/10(五)演講公告

會員評等:  / 0

者:Shengyen Li 博士

演講題目:A machine-learning assisted ICME framework for materials design

演講人Shengyen Li 博士

服務單位美國西南研究院

演講題目A machine-learning assisted ICME framework for materials design

演講時間109710(星期五)早上1030

演講地點工綜館228教室

個人經歷

美國西南研究院 / Research Engineer

個人簡歷

Shengyen Li received his Ph.D. from the Department of Mechanical Engineering at Texas A&M University in 2013. From 2008 to 2013, he studied the phase transformation of transformation-induced plasticity (TRIP) steels and optimized the processing conditions to maximize the work-to-necking of these medium carbon steels. He joined the Material Genome Initiative (MGI) program at the National Institute of Standards and Technology (NIST) in 2013. During his five years there, he developed an ICME framework called the Material Design Toolkit to assist the decision-making for material design (https://mgi.nist.gov/materials-design-toolkit). While working at NIST, Dr. Li also participated in many activities in the Center for Hierarchical Materials Design (CHiMAD) at Northwestern University. CHiMAD is the center of excellence sponsored by NIST focused on material design. Dr. Li was the first exchange fellow to develop a long-term relationship between the two institutes and he brought collaborations in developing material design strategies and ICME frameworks. Since 2018, Dr. Li is leading the development of the ICME program for metal additive manufacturing at Southwest Research Institute.

演講摘要

With the announcement of the Materials Genome Initiative by the White House advocating for a 50% reduction in time and cost to develop and deploy new materials, the need to accelerate computational material design approaches has become essential. It has been demonstrated that integrated computational materials engineering (ICME) is capable of bringing deep insights into the process-structure-properties relations for various applications [1], [2]. However, the underdevelopment of the physics-based models is one of the critical gaps to create a modeling hierarchy for an ICME approach [3]. Machine-Learning/Artificial-Intelligence (ML/AI) are the alternative methods to learn the present knowledge and bridge the gaps for model integrations. Because of their efficient numerical processes, the ML tools can also assist to formulate the problems and validate models. In this seminar, I will compare a physics-based ICME workflow to a ML-based approach for the designs of a Ni-based superalloy and a transformation induced plasticity (TRIP) steel. The potential risks of employing ML/AI will be shared as well.

[1] D. G. Backman, D. Y. Wei, D. D. Whitis, M. B. Buczek, P. M. Finnigan, and D. Gao, “ICME at GE: accelerating the insertion of new materials and processes,” JOM, vol. 58, no. 11, pp. 36–41, 2006.

[2] J. Allison, D. Backman, and L. Christodoulou, “Integrated computational materials engineering: a new paradigm for the global materials profession,” Jom, vol. 58, no. 11, pp. 25–27, 2006.

[3] X. Liu, D. Furrer, J. Kosters, and J. Holmes, “Vision 2040: a roadmap for integrated, multiscale modeling and simulation of materials and systems,” 2018.

20200710-2