This presentation was made at CAASE18, The Conference on Advancing Analysis & Simulation in Engineering. CAASE18 brought together the leading visionaries, developers, and practitioners of CAE-related technologies in an open forum, to share experiences, discuss relevant trends, discover common themes, and explore future issues.
Consider the traditional design cycle; an engineering team takes a set of requirements like loads, constraints and form. It then designs a nominal structure that may or may not satisfy the requirements. This nominal structure is analyzed and then redesigned over several iterations until it meets the original requirements. This process takes a long time and is particularly laborious when requirements change mid-process.
In the generative design workflow, the user inputs all requirements into the software. They also input design options like different materials, build orientations and safety factors. This allows the software to generate many designs that are fully analyzed. Instead of having to iterate and redesign, the engineer can navigate a range of designs for which all requirements have been met. Should requirements or specifications change, they can simply choose another design from the solution space.
This paper considers the well known Alcoa-GrabCAD bracket challenge and applies the generative design workflow. One of the many potential options is taken from design, to build, to test, and fully verified with detailed non-linear finite element models. It is shown that the engineer can very accurately simulate the real mechanical behavior of these designs. The test method and test apparatus is described in detail with jig design etc. The method of correlation is fully described and the relevant challenges highlighted. Correlation is based on non-linear static models (geometric and material non-linearity).
Comparisons are drawn between the mechanical performance of the original bracket and the generative design bracket. The efficiency of shape is shown to not only improve mass, but also ductility and strain to failure. Simulations are used to show redistribution of stresses in the plastic range. Nuances about additive versus subtractive materials are also explored.
Finally in-service design, validation and verification recommendations are given. This relates to type of material and desired additive material properties.
|Date||6th June 2018|