Learning Forum

[William Allouche]

Hello,

I am currently studying the topology optimisation potential on a 3-point bending beam case and I wanted to compare the results between a density-based and a level set based topology optimisation in Structural Optimisation. Density-based optimisation works just fine but when I try to run the level set based optimisation, the solver does not seem to start as I keep getting the following error (taken from the topo.err file):

IDS_LVLS_TOPO_OPTIM_UNKNOWN_ERROR:3: message=DPF error - runtime error: self_call; generalized_inner_product:73<-VolumeForce:72<-Current element is not mapped properly: id=364534 apdl_type=174 n_dofs=3 data_size=192 

Could someone help me understand what this error means? 

I can provide more details on my model if needed.

 

Thank you and best regards,

William

 

[John Doyle]

As you know, the Level Set method deals with the boundary of the shapes during the optimization process.  Your mesh might be too coarse to satisfy all the compatibility conditions.  Have you tried a finer mesh?

[William Allouche]

Hello,

Thank you for your reply and sorry to give my own reply only now.

Changing the mesh for a finer one did not solve this particular issue. However, when I modelled my beam, I also modelled the 2 supports and the indenter:

I noticed I was able to run a Level-set based topology optimisation when 1) I created a new file on the Workbench and 2) when there was only body involved in Mechanical (there has to be only one body in Mechanical, no matter the role of the other bodies I want to put in there). The 2nd point seems a bit strange, since I read online that I should have the possibility to run a topology optimisation with multiple optimisation regions and Level-set based topology optimisation.

But when I try to run a simulation with more than one body, let alone one with multiple optimisation regions, I get exactly the same error as the one I copy-pasted in my first message: do you know if that is normal?