Generative design creates shapes that are difficult or impossible to imitate with normal construction methods. In this case, the computer or the cloud are merely given boundary conditions to allow the computer to create the model. Yes, you read that right, the computer does the modeling. With this post I would like to take a look at generative design with Autodesk Fusion 360 and use a small example to show the differences to topology optimization.
Especially in mass-relevant applications (aerospace, automotive, robotics), the weight reduction of components is a decisive factor for increasing efficiency in several areas, which creates wonderful potential. In addition to traditional material savings, the mass and consequently the downstream consumption of energy carriers are reduced or the cycle time in production is shortened through lower moving mass. In both cases, costs are reduced.
Let's look at an example from robotics. In the following picture you can see a holder for holding an electric screwdriver. Such a robot could automatically tighten the screws with the required torque when assembling electric motors and thereby relieve the fitter. In the current case, the holder is conventionally manufactured, and not necessarily relevant for optimization, since it is already lightweight. However, this is of secondary importance for our example, since the principle is also clear.
Without generative design, a topology optimization would take place in order to reduce the mass of the component. We set the boundary conditions of the safety factor, the fixed restraint, the forces and the desired mass reduction. In the present case, a straight line result is calculated. In principle, this is a classic task for calculating the optimum of the current component. As a result, we will find that we “only” receive a modified shape of the original component as a surface model. A feasibility, a production price or other variants are not shown.
The generative design seems identical at first glance, but when defining the boundary conditions it becomes clear that in addition to the above Criteria obviously also the possibilities for manufacturing the elements is considered. In addition to the classic subtractive processes, additive manufacturing processes are also taken into account. This is unique! The sheer number of possibilities that now arise is no longer tangible for the human mind - at least for mine. The large number of boundary conditions is also reflected in the results. So not only one result is shown, but several at the same time. The decision as to which result is the most suitable is still left to the designer. Aesthetic aspects are probably also relevant. Of course, it is not a requirement to only use components that look bionic. This can also be accepted as a creative extension of one's own thoughts in order to develop completely new designs.
Topology optimization vs. generative design
• Start condition: "only" one single part / body
• Result of the study: surface model
• Consideration of two criteria (mass and rigidity)
• "Transfer" in the volume model is done manually
• pure "optimization" and only one "solution"
• Start condition: single part or assembly
• Result of the study: volume model
• Observance of more than two criteria
• No "transfer" necessary in the volume model
• "Optimization" and multiple solutions