Assoc. Prof. Zhanli Liu

Tsinghua University, China

Speech Title: Solving direct and inverse computational engineering problems through a deep learning based data-driven method

Abstract

With the steady development of computer science, machine learning and data science have made significant progress in recent decades. These techniques generally rely on a substantial amount of data samples to extract the abstract mapping hidden within the data. Hence, these technologies have gradually attracted the attention of researchers in the field of computational mechanics modeling and material design. This study aims to interpret several forms of applications that integrate machine learning and data science with computational mechanics modeling and material design. In the first application, a method is proposed to establish the implicit mapping between the effective mechanical property and the mesoscale structure of heterogeneous materials. Shale is employed in this study as an example to illustrate the method. A convolutional neural network is trained based on the images of stochastic shale samples and their effective moduli. The trained network is validated to be able to predict the effective moduli of real shale samples accurately and efficiently. Not limited to shale, the proposed method can be further extended to predict the effective mechanical properties of various heterogeneous materials. In the second application, assisted by image-based finite element analysis and deep learning, a data-driven approach is proposed for designing phononic crystals. An auto-encoder is trained to extract the topological features from sample images. Finite element analysis is employed to study the band gaps of samples. A multi-layer perceptron is trained to establish the inherent relation between band gaps and topological features. The trained models are ultimately employed to design phononic crystals with anticipated band gaps. Not limited to this material, the proposed method could be further extended to design various structured mechanical materials with specific functionalities.

Biography

Professor Zhanli Liu is an Associate Professor at the Department of Engineering Mechanics of the School of Aerospace in Tsinghua University. He got his B.SC. in Mechanics by Tsinghua University in 2004 and his Doctoral Degree in 2009 also in Tsinghua. In 2010 he was a visiting scholar in the Department of Mechanical Engineering at Northwestern University. In 2011 he was awarded the National Excellent Doctoral Dissertation of P.R. China. Since 2008 he has published several articles both in national and international journals. His research interests focus on Multi-scale crystal plasticity, Advanced numerical methods in the simulation of fracture and damage in solids, and Micromechanics simulation of advanced composite materials.

清华大学航天航空学院长聘副教授、博导。现任清华大学航天航空学院工程力学系副主任，中国力学学会计算力学专委会副主任，北京力学会秘书长，国际期刊International Journal of Fracture, Regional Editor，应用数学和力学期刊编委，力学学报青年编委等。主要围绕固体强度与失效的力学建模、数值仿真及工程应用开展研究，包括材料动态力学行为表征与建模、轻质综合防护材料及结构设计、人体冲击致伤及防护、基于机器学习的计算力学及反向工程设计等。研究成果应用于冲击波防护装备研制、页岩水力压裂施工设计、飞行器穿盖弹射救生等国家重大工程。所获奖励包括基金委优秀青年基金、中国力学学会青年科技奖、教育部自然科学奖一等奖（排名2）、航空学会科学技术奖一等奖（排名5）、钱令希计算力学青年奖等。