AI-Enabled Large-Scale Structural Optimization

Xu Guo

Dalian University of Technology

Abstract:   Structural analysis for structures composed of highly heterogeneous materials which often involves the solution of large-scale linear algebraic equations is very time-consuming even within the linear elastic regime. Furthermore, the tremendous computational cost of iterative large-scale finite element analysis also prevents the widespread use of topology optimization as a powerful design tool especially when the desired design resolution is very high. In order to break the bottleneck hindering the efficient solutions of large-scale structural analysis a nd design optimization problems, a general machine learning (ML) enhanced substructure-based framework is proposed. The essential idea is resorting to the classical substructure-based finite element analysis approach and establishing an implicit mapping between the parameters characterizing the material distribution within a substructure and the corresponding condensed stiffness matrix/ numerical shape functions through offline trained deep neural networks. In contrast to most of the existing ML enhanced approaches, the proposed framework is truly problem independent machine learning (PIML), i.e., which is independent of the forms of structural geometry, boundary condition and external load, and can be applied to solve various boundary value problems governing by the same type of partial differential equation once the offline training is completed. Compared with the traditional paradigm, it can achieve 103 -104 times solution efficiency for tested large-scale examples with satisfactory accuracy. The effectiveness of the approach was also validated for the non-adjoint topology optimization problems, i.e., 3D compliant mechanism design. Finally, to demonstrate the proposed approach's capability in dealing with extremely large-scale three-dimensional problems, a full-parallel framework based on PIML approach is developed. A cantilever beam example over 100 billion degrees of freedom is optimized in 200 iterations with an average time of only 42.0 seconds per step.

Biography:   Xu Guo, academician of the Chinese Academy of Science, a recipient of the National Natural Science Fund for Distinguished Young Scholars, and Chang Jiang Distinguished Professorship, is now a professor at Dalian University of Technology and holds the deputy director of the State Key Laboratory of Structural Analysis, Optimization and CAE Software for Industrial Equipment. Prof. Guo has long been devoted to research on computational mechanics, structural optimization, and solid mechanics, as well as the development of autonomous and controllable CAE optimization software, and has made significant contributions in these fields. He won two second-class awards of the National Natural Science Prize, two first-class awards of the Natural Science Prize for the Chinese Ministry of Education. He is also a vice president of the Chinese Society of Mechanics, vice president of the International Society of Structural and Multidisciplinary Optimization (ISSMO), associate editor of the Journal of Mechanical Design-Transactions of the ASME, review editor of Structural and Multidisciplinary Optimization, editorial board member of Computer Methods in Applied Mechanics and Engineering, International Journal for Numerical Methods in Engineering, etc.

Nonlinear Dynamics of Machining Processes

Marian Wiercigroch

University of Aberdeen

Abstract:   Comprehensive understanding of chatter, bifurcations and stability is essential for improving productivity, quality, and efficiency of manufacturing operations. Chatter manifests itself as undesired large amplitude self-excited vibration, whilst bifurcation and stability analysis hold the keys to its effective control and suppression. Bifurcations in manufacturing occur when processing and system parameters cross certain values and(or) they are results of complex interactions resulting predominantly in creating or disappearance of limits cycles. Bifurcations in manufacturing processes belong to the most complex ones, where nonlinearities are very strong and play the dominant role. Process stability is vital to ensure high-quality surface finish, dimensional accuracy, and extended machine and tool life. In this lecture, first I will define nonlinearity, chatter, bifurcations and stability in manufacturing processes with the focus on metal cutting. Then I will discuss on the frictional chatter, which was introduced and investigated with my group and collaborators. Chatter in precision grinding has been a subject of many our investigations, where low dimensional strongly nonlinear models were used to undertaken in depth analytical and numerical studies. Finally, examples of advanced nonlinear dynamics techniques such as path following bifurcation analysis, basins of attraction, Poincare maps and Lyapunov exponents will be discussed, which have been used to examine stability and determine a practically important measure so-called cutting safety.

Biography:   Professor Marian Wiercigroch holds a prestigious Sixth Century Chair in Applied Dynamics and he is a founding director of the Centre for Applied Dynamics Research at the University of Aberdeen. His area of research is theoretical and experimental nonlinear dynamics. Wiercigroch has published extensively (over 500 journal and conference papers) and sits on a dozen editorial boards of peer review journals. He is a frequent keynote and plenary speaker at major international conferences and the Editor-In-Chief of International Journal of Mechanical Sciences. He is the inventor of new patented drilling technology called Resonance Enhanced Drilling and the Founder and Chief Technology Officer of a spinoff company iVDynamics Ltd. He has established in Aberdeen unique experimental laboratories. Marian is a Scottish Champion of Knowledge Exchange (2020) and he served as a panelist in the Research Excellence Framework (2014, 2021). He has received many awards and distinctions including a Senior Fulbright Scholarship (1994), Fellowship of the Royal Society of Edinburgh (2009), DSc honoris causa from the Lodz University of Technology (2013), Distinguished Honorary Professorships from the Perm National Research Polytechnic University (2017), Balseiro Institute (2018), Yanshan University (2021) and University of Nottingham Ningbo (2023).

The Mechanics of Product Disassembly for Remanufacturing

Duc Truong Pham

University of Birmingham

Abstract:  Remanufacturing is the process of returning a used product to at least the same condition as the original product. Remanufacturing saves resources (raw materials, energy, water, etc.), reduces greenhouse gas emissions and avoids the need for landfill space. Remanufacturing is part of a circular economy. Disassembly is the first operation in remanufacturing. It is labour-intensive and very difficult to automate. The aim of our research is to develop mathematical and computer models of ‘atomic’ disassembly processes such as removing a shaft from a clearance hole, separating press-fitted components and undoing a nut or a bolt. The work forms part of a research programme to investigate disassembly science and derive a fundamental understanding of the mechanics and informatics involved in disassembly. Our hypothesis is that such an understanding would enable the engineering of reliable and efficient disassembly automation equipment. This presentation will start with a discussion of remanufacturing to set the scene for our disassembly research.It will subsequently focus on the results of our investigations into the mechanics of fundamental disassembly tasks to inform how they can best be performed by machines and what tools should be designed to facilitate the automatic performance of those tasks.

Biography:  Prof. Pham, OBE, FREng, FLSW, BE, PhD, DEng, CEng, FIET, FIMechE, SFHEA, holds the Chance Chair of Engineering at the University of Birmingham. His research covers intelligent systems, robotics and autonomous systems and advanced manufacturing and remanufacturing technology. He has published over 600 technical papers and books and has graduated more than 100 PhD students. He has received several awards including five prizes from the Institution of Mechanical Engineers, a Lifetime Achievement Award from the World Automation Congress and a Distinguished International Academic Contribution Award from the IEEE. He is a Fellow of the Royal Academy of Engineering, Learned Society of Wales, Society for Manufacturing Engineers, Institution of Engineering and Technology and Institution of Mechanical Engineers. He is the founding editor of the Springer Series in Advanced Manufacturing and editor-in-chief of Cogent Engineering and the International Journal on Interactive Design and Manufacturing.

A Physics-data Hybrid Framework to Develop Bridge Digital Twin Model in Structural Health Monitoring

Weixin Ren

Shenzhen University

Abstract:   Digital twin in structural health monitoring aims to create a virtual model for a physical structure by combining measurement data. The most important feature is to achieve the physical structure-monitoring data synchronization. For this purpose, a physics-data hybrid framework to develop the bridge digital twin model in structural health monitoring is presented. The physical base is firstly formed by the finite element model of the digital representation for the physical bridge that can fully incorporate with both structural geometry and structural state. The data base is then built by all measurement data of the monitored bridge. By defining the context that is common to both physical base and data base, the mirror relationship between physical base and data base for the specified context is formulated. To achieve the best matching of the mirror relationship by minimizing process, the digital twin model in term of the specified context can be developed. In such away, the proposed framework integrates physical knowledge and data intelligence into one model. A demonstration of a simulated simply supported beam is provided to show how the digital twin model is developed by using proposed physics-data hybrid framework. The presented physics-data hybrid framework is help of clearer understanding of the realization of digital twin model in structural health monitoring, providing a new perspective for smart bridge solutions.

Biography:   Dr. Wei-Xin Ren, foreign member of the Russian Academy of Engineering, is currently chair professor in the college of civil and transportation engineering at Shenzhen University, China. He received his PhD in bridge engineering from Central South University in 1993 and has been a full professor since 1995. His research interests include nonlinear system identification, structural health monitoring, damage detection and model updating. Prof. Ren has authored over 400 academic papers including more than 230 referred international journal papers. Prof. Ren currently serves as Associate Editor of Mechanical System and Signal Processing, Editorial Board Member for Engineering Structures and Deputy Director of National Key Laboratory of Green and Long-Life Road Engineering in Extreme Environment.

Dynamics and Bifurcations of Product-Quadratic Nonlinear Systems 

Albert C. J. Luo

Southern Illinois University Edwardsville

Abstract:   This talk is about nonlinear dynamics and bifurcations of the quadratic dynamical systems. Such a two-dimensional dynamical system is one of simplest dynamical systems in nonlinear dynamics, but the local and global structures of equilibriums and flows in such two-dimensional quadratic systems help us understand other nonlinear dynamical systems, which is also a crucial step toward solving the Hilbert’s sixteenth problem. Possible singular dynamics of the two-dimensional quadratic systems are discussed in detail. The dynamics of equilibriums and one-dimensional flows in two-dimensional systems are presented. Saddle-sink and saddle-source bifurcations are discussed, and saddle-center bifurcations are presented. The infinite-equilibrium states are switching bifurcations for nonlinear systems. From the first integral manifolds, the saddle-center networks are developed, and the networks of saddles, source, and sink are also presented. 

Biography:   Albert C.J. Luo a distinguished research professor at Southern Illinois University Edwardsville. His research interests include: Hilbert 16th problem, discontinuous nonlinear dynamical systems, periodic flows to chaos in time-delayed nonlinear systems, bifurcation trees of periodic flows to chaos in nonlinear systems, synchronization of dynamical systems, discontinuous dynamical system theory, stochastic and resonant layer theory in nonlinear dynamic systems, nonlinear theories for deformable-body dynamics etc. He proposed various new theories and methods, and achieved many new breakthroughs. Professor Luo has published 50 monographs and over 470 journal and conference papers. He organized over 30 international symposiums and conferences on Dynamics and Control. He is also a fellow of ASME and member of editorial boards for several international academic journals.

Multi-Objective Reliability-Based Design Optimization of Structural Dynamical Systems under Stochastic Excitation

Hector Jensen

Santa Maria Technical University

Abstract:  Reliability-based multi-objective optimization formulations give valuable insight for the design of structural dynamical systems under stochastic excitation. Nonetheless, since reliability assessment is often conducted using simulation techniques, the generation of the entire Pareto front or compromise solutions may lead to excessive or even prohibitive computational costs. In this lecture, an approach for exploring specific tradeoff solutions is discussed. To this end, a compromise programming problem is formulated using a weighted Chebyshev metric based on aspiration levels. The formulation leads to a min-max optimization problem which is solved by means of a two-phase stochastic search technique that sequentially explores the corresponding feasible and optimum solution sets. A number of nearly equivalent tradeoff designs are obtained at the end of the solution process, which provide insight and flexibility for decision-making purposes. For improved numerical efficiency, an adaptive surrogate model for reliability measures is implemented. A number of examples are presented to assess the capabilities of the proposed method and to illustrate the effectiveness and usefulness of the optimization approach. The proposed framework encompasses several advantageous features for practical implementation. For example, it allows exploring specific tradeoff solutions even in non-convex Pareto fronts, it gives information about the interaction between the objective functions within the feasible set which can be employed to assist decision-making processes, and it is well suited to handle problems involving multiple nearly optimal and disjoint regions. Overall, the proposed framework constitutes a potentially useful tool to support decision-making processes in a practical class of structural design problems, including general structural dynamical systems under stochastic excitation.

Biography:  Hector Jensen is a Professor Civil Engineering at the Santa Maria Technical University, Valparaiso, Chile, and Professor of Mechanical Engineering at the Catholic University of Chile, Santiago. Chile. He is a Mathematician Civil Engineer from the University of Chile, and he received his PhD in Applied Mechanics from the California Institute of Technology, Pasadena, USA. His research interests include Computational Stochastic Mechanics, Advanced Simulation Methods, Robust and Reliability-Based Optimization, Risk and Sensitivity Analysis. He has been visiting professor in several American and European universities, including University of California at Los Angeles, California Institute of Technology, University of Michigan, University of Innsbruck, etc. He has published numerous journal and conference papers, as well as two books. He has been invited to give lectures, keynotes, semi-plenary and plenary lectures in a number of universities and international conferences. He is member of the editorial board of several journals, and he has been guest editor of different Journals, including Computers and Structures and Mechanical Systems and Signal Processing. In 2018 he was selected by the Recruitment Program of High-end Foreign Experts of the State Administration of Foreign Affairs of the People's Republic of China.

Advancing Vehicle Comfort and Safety with Electrically Interconnected Suspension Systems

Haiping Du

University of Wollongong

Abstract:   Vehicle suspension systems play a crucial role in mitigating vibrations and enhancing ride comfort for passengers and drivers. However, traditional independent suspension designs often struggle to optimise stiffness and damping characteristics independently across various vibration modes, leading to compromises in ride comfort and stability. In response, electrically interconnected suspension systems have emerged as a promising technology to address these challenges. In this keynote, we provide an overview of our research on electrically interconnected suspension systems and associated technologies. We discuss the limitations of conventional suspension systems and explore the principles and technical features of electrically interconnected systems, focusing on the electrical network and decoupling control characteristics. We highlight the concept of synchronous decoupling control for multiple vibration modes, discuss control algorithms, and present performance verification results. Our research findings, derived from recent publications, demonstrate the potential of electrically interconnected suspension systems for real-world implementation. With their rapid responsiveness, affordability, and effectiveness, these systems offer promising advancements in vehicle comfort and safety.

Biography:   Prof. Du received the PhD degree in mechanical design and theory from Shanghai Jiao Tong University, Shanghai, China, in 2002. Since 2016, he has served as a Senior Professor at the School of Electrical, Computer and Telecommunications Engineering at the University of Wollongong, Australia. Prior to his current position, Prof. Du held various research roles, including Research Fellow at the University of Technology, Sydney, and Postdoctoral Research Associate positions at Imperial College London and the University of Hong Kong.Prof. Du is a Subject Editor of the Journal of Franklin Institute, an Associate Editor of IEEE Transactions on Industrial Electronics, IEEE Transactions on Intelligent Vehicles, and IEEE Control Systems Society Conference. He also serves on the editorial boards of several international journals, including the Journal of Sound and Vibration, IMechE Journal of Systems and Control Engineering, and Journal of Low Frequency Noise, Vibration and Active Control. With research interests spanning vibration control, vehicle dynamics and control systems, robust control theory, electric vehicles, robotics and automation, and smart materials and structures, Prof. Du is at the forefront of interdisciplinary research aimed at advancing engineering applications and addressing real-world challenges.