Modifications for Density Based Topology Optimisation #80
Replies: 3 comments 16 replies
-
We recently added the capability to compute dF/dBeta, where beta is a scalar field in the turbulence equation. We are about to release a new version 2.2.1 that will allow computing dF/dAlpha, where alpha is a source term (alpha*U) added to the U equation (This happens to be exactly what you need). So stay tuned. |
Beta Was this translation helpful? Give feedback.
-
Dear friedenhe, first of all thank you very much for providing this great framework! I am also very interested in testing the possibilities of DAFoam regarding density based topology optimization. If it is not possible yet, do you plan to provide such functionality in the future? Thank you very much in advance! |
Beta Was this translation helpful? Give feedback.
-
@jjahammond @schewef Check this new tutorial for density-based topology optimization. |
Beta Was this translation helpful? Give feedback.
-
Hi Ping,
I'm a PhD student and so far my research has focused on turbulence modelling in density based topology optimisation using continuous adjoints in OpenFoam. I am interested in trying a discrete adjoint approach as our current continuous adjoint implementations still rely on a constant eddy viscosity assumption.
That said I'm interested in trying to modify DAFoam to run density based topology optimisations (e.g. by including a resistance term
fvm::Sp(alpha,U)
into the UEqnSimple.H of DASimpleFoam and iteratively updatingalpha
based upon the sensitivitydC/dalpha
). I wonder how tractable you think this would be given the current focus of the DAFoam library is for shape optimisation?Thanks and I look forward to hearing from you,
James
Beta Was this translation helpful? Give feedback.
All reactions