Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
How to perform the topology optimisation of a structure by considering the stress constraints?
Posted 14.01.2022, 00:22 GMT-5 Optimization Version 5.6 2 Replies
Please login with a confirmed email address before reporting spam
Hello everyone, can anyone suggest me how to perform the topology optimisation of a structure by considering stress as a constraint and minimise mass as an objective function in the Comsol Multiphysics software? Generally, in stress constraints, we either use global or local stress constraints. There is a inbuilt module present in comsol where it is explained about the traditional topology optimisation formulation, i.e, minimise compliance subjected to volume constraint. But i am not aware how to incorporate stress constraints in the topology optimisation problem formulation. Can anyone help me in this mattter? Thanks in advance.
Please login with a confirmed email address before reporting spam
There is a Max operator that you can apply to a domain, which keeps track of the maximum value of a variable. Or you can apply an average operator or integral operator to look more globally at the stress.
Please login with a confirmed email address before reporting spam
There are many academic articles dedicated to this topic. COMSOL supports the density method via our Density Model feature. One can use a stress aggregation strategy similar to what is done for the flywheel_profile model (in the application library), but oftentimes you need to use a smaller exponent for topology optimization (~10) than for shape optimization.
Keep in mind that for topology optimization one has to interpolate the stiffness used for evaluation of the stress constraint different from the stiffness used in the actual forward evaluation.
Finally, one often has to augment the stress constraint with a compliance constraint. Otherwise, you risk getting a very soft design.