Seminar: Topology Optimization of Multi-physics Problems using Level Sets and the Extended Finite Element Method

Abstract:

Ad-hoc or intuitive approaches for designing engineering systems that are dominated by nonlinear interactions of multiple physical fields often yield suboptimal solutions and cause delays in the design process. Model-based approaches provide promising avenues for systematically designing such engineering systems. In this presentation, we will review and discuss design optimization methods, in particular topology optimization approaches, that explore the unconventional and interesting design spaces of engineering problems dominated by multi-physics phenomena, such as fluid-structure interaction, thermo-mechanical coupling, and natural convection flows. Traditionally topology optimization methods are solved by a density approach, representing the structural geometry via the distribution of a fictitious material within a given design domain. More recently level set methods have enjoyed increasing popularity; these methods describe the geometry via dynamic implicit surfaces. As interface phenomena often play an important role in multi-physics problems, we will show that level set methods provide distinctive advantages for optimizing multi-physics problems. For selected problems, we will compare density and level-set approaches with respect to their accuracy and efficiency. In particular, we will highlight the advantages and disadvantages of combining a level-set description of the structural geometry with an XFEM discretization of the underlying physics model.

About the Speaker:

Dr. Maute is a professor in the Department of Aerospace Engineering Sciences and the Department of Mechanical Engineering, by courtesy. He is currently serving as director of the Center of Aerospace Structures (CAS) at the University of Colorado Boulder. He received his Ph.D. degree in Civil Engineering in 1998 and his Bs/Ms in Aerospace Engineering in 1992, both from the University of Stuttgart, Germany. His research is concerned with computational mechanics and design optimization methods, applied to fundamental problems in solid and fluid mechanics and heat transfer. Dr. Maute has published his work in over 180 journal articles, book chapters, and papers in conference proceedings. He is a fellow of USACM and a senior advisor of Structural and Multidisciplinary Optimization.

Hosted by Professor Soheil Soghrati.