Manufacturability-Driven Multi-Component Topology Optimization Of Thin-Walled Structures Based On A Level Set Method
Ayinde, Samuel
2020
Abstract
Thin-walled structures (TWS) are suitable for lightweight, load-bearing enclosures with various external geometries with internal reinforcements. Thin-walled structures find application in automobiles, aircrafts, ships, and industrial facilities. Past research in the field of structural design optimization have been done to make single-piece thin-walled structures less costly, lighter and of better performance. The primary drawback of these research is that complex structures are scarcely manufactured as a single piece, and this has made the optimization of single-piece structures to be of little industrial relevance. The goal of this dissertation is to develop a computational method for simultaneous design and partitioning of assemblies made of thin-walled components, driven by component manufacturability. First, the conventional level set function for monolithic topology optimization based on a signed distance function is extended to realize a simple representation of monolithic thin-walled structures with uniform thickness, by taking advantage of the signed-distance property. Second, a new multi-domain representation within a level set, inspired by level-set methods for multi-material topology optimization, is introduced to model multiple components, where the additional level sets specify partitioning of the level set for a monolithic thin walled structures. Finally, the geometric constraints imposed by a manufacturing process for thin-walled components, sheet metal stamping as an example, are introduced to formulate the manufacturability-driven, multi-component topology optimization of thin-walled structures. The optimization problem is formulated as continuous optimization with respect of the level set parameters that specify overall structural geometry and its partitioning, which can be solved efficiently by gradient-based optimization algorithms. A few examples inspired by the sheet metal structures for automotive applications demonstrated the effectiveness of the new formulation to automatically design thin-walled structures made of multiple component each of which satisfies process-specific geometric constraint for component manufacturing. The conventional approach for design and partitioning is a two-step process in which the optimization of the single-piece geometry is first carried out, followed by the decomposition of the optimized single-piece geometry to refine part boundaries and joint configurations. Since the outcome of the second step largely depend on the first step, the two-step approach is likely to yield suboptimal solution. Although the improvement resulting from the new formulation of simultaneous design and partitioning cannot be quantify, it is expected to bring about improvement when joint modeling is implemented. This dissertation advances the state of the art of the simultaneous designing and partitioning of thin-walled structures driven by manufacturability. While the dissertation focuses on the auto-body application, it is expected that the methodology will be applicable to other domains of thin-walled structures.Subjects
Thin-walled structures Level set Topology optimization Manufacturability-driven Multi-component Manufacturing
Types
Thesis
Metadata
Show full item recordCollections
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.