Development of a Process for Determining Minimum Feature Size in Additive Manufacturing with Applications to Topology Optimization
Weiss, Benjamin M
MetadataShow full item record
Additive manufacturing (AM) is opening a new creative landscape for designers. Freedom of form and material creates opportunities for customization, increased functionality, and higher complexity in manufactured objects. Characterizing and understanding the unique manufacturing constraints of AM is a key to enabling the designer to leverage the full potential of AM to produce highly complex geometries. In this work, “minimum feature size”, defined as the smallest scale at which a feature of a particular shape in a particular orientation can be manufactured, is examined in detail. A parameterization for shape and orientation of small features is developed, and an adaptive, iterative experimental procedure is created to accurately estimate the minimum feature size for a particular feature type in the parameter space. A design of experiments process is used to systematically explore the minimum feature size over the space of parameterized shapes and orientations. This process is applied to three different AM platforms, and in each case minimum feature size spans an order of magnitude over the set of considered features. The data collected are used to create a parametric design rule for each process, a function which provides a detailed map of the minimum feature size achievable over the parameter space. The parametric design rule is applied to typical design problems, used to assess existing 3D models for manufacture, and applied to design optimization frameworks. The parametric design rules produced are found to more tightly follow the actual capabilities of the AM processes, increasing the envelope for complexity compared with existing design guidelines consisting of one or two constant minimum feature sizes and improving accuracy in predicting the minimum feature sizes for new geometries by as much as 50%. Application is made to three design problems, realizing significant improvements in each case. The experimental study of minimum feature size is complemented by an adaptation of topology optimization to incorporate the parametric design rules into an automated design process. A new variant of the Moving Morphable Components (MMC) approach is created, improving convergence through a “bootstrapping” approach and integrating the feature-dependent design rules to ensure manufacturability of the result. Experimentally-verified manufacturable outputs are achieved while sacrificing less than 5% of the objective function value for several test problems. By bringing together experimental assessment of minimum feature size and design optimization through the creation of parametric design rules, a novel, reproducible, and practical process for ensuring manufacturability of optimized designs is created.
- Mechanical engineering