Research theme
JSIAM research activity group : Mathematical design
JSIAM research activity group : Mathematical design
Background of researches on shape and topology optimization
The drag-minimizing shape of an isolated object in a Stokes flow was analytically obtained by O. Pironneau. It was subsequently shown that a similar solution could be obtained numerically, and related researches has been widely conducted, including applications to problems with changed flow field conditions and real problems. This research has led to the development of shape optimization techniques, which optimize boundary shapes, and topology optimization techniques, which create appropriate holes based on design objectives. Since the 2010s, when 3D printers began to become widespread, topology optimization technology has attracted renewed attention, and in recent years, many sessions on the subject have been scheduled at academic conferences and other events. Against this background, our research group is holding research symposiums focusing on shape optimization, topology optimization, and related inverse analysis techniques.
Fig, Example of 3D topology optimization result
Application of theory of topology optimization for non-destructive testing
The number of structures that are over 50 years old is increasing every year, and quantitative and accurate inspection of defects in structures is an urgent task. Topology optimization theory is a methodology for creating optimal structures to achieve a certain design objective under mass constraints. However, if the objective in this methodology is replaced with the problem of creating an appropriate hole shape so that the inspection index matches the measured value, topology optimization theory can also be applied to non-destructive inspection. Research into such applications of topology optimization has also been progressing in recent years.
Fig. Example of void topology identification (*Internal bodies indicate void region.)
Future view of development of optimization methodologies
In recent years, with the development of AI, there has been an increase in research into the application of machine learning. There is a need to develop new optimization methodologies so that structural shape optimization results can be instantly created based on pre-learning results. We will also have organized sessions at international conferences, so please feel free to join us if you are interested.