Learn about differences in the mental approach to two techniques toward refining a design.
- [Instructor] Finite element analysis has been available to engineers for decades. Over time engineers have begin to use the tools to refine their design components. First by trial and error, and then more recently the software has begun to make suggestions on how the model could be optimized. In some circles people have suggested that Generative Design is really the same thing as Shape Optimization, but the two, while they might appear similar in their results, have some significant differences. Both are based on Cloud studies in Fusion 360, but Generative Design will iterate based on changes to the stress loads on portions of the mesh, so the mesh will actually change form as the stresses move around. Shape Optimization is based on load paths, so the geometry will be made to conform to initial load paths. There's also no limit on Generative Design to the initial geometry. It creates the geometry as it sees needed based on the stress. For Shape Optimization it will refine the shape that you began with. There's also no accommodation in Shape Optimization to work with multiple materials. If you want to test aluminum versus steel you need to run two studies. With generative you construct a list of materials you want evaluated and they're evaluated simultaneously. The final major difference is accounting for manufacturing. It's entirely possible through Shape Optimization to get a result that cannot be manufactured at all. With Generative Design you have the option to choose what sort of manufacturing process you want to use. Whether it be two-axis cutting, three-axis cutting, five-axis milling, die casting or additive manufacturing, you have a broad variety of options that will be kept in mind while the shape is developed. As a comparison the initial design that I'd come up with for this part when I ran static stress analysis on it in Fusion 360, the design weighed over 87 ounces, had a minimum safety factor of 13.3 and a peak stress of over 2000 PSI, so there was a lot of extra weight that wasn't needed in the design. Using Shape Optimization I was able to bring the mass down to just over 53 ounces, reduced the minimum safety factor, but still have it well above what I needed and the peak stress increased slightly Using Generative Design I was able to drastically reduce the weight to 1/10 of Shape Optimization with a slight increase in safety factor and peak stresses, but a much lighter part that will clearly get the job done. The key here is that Generative Design breaks down barriers to developing an idea by not only focusing on the engineering criteria, but also keeping an eye on the manufacturing process.
- Generative design vs. shape optimization
- Starting a generative study
- Selecting portions of a model to preserve
- Defining obstacles
- Setting design conditions and criteria
- Reviewing and comparing outcomes
- Exporting results
- Cloning an existing study