Inspire Topologic Optimization not working

GiovanniBarreca
GiovanniBarreca New Altair Community Member

Hello, I am working with a rather complex geometry that I aim to optimize in Inspire. While the analysis proceeds without any issues, the optimization fails completely. In previous attempts, I received error messages, but now the only feedback is "optimization failed," accompanied by this log:


-------------------- Progress Tracker --------------------

hwfdmfeprocessor process id = 35668
Client = Inspire 24.1 14745
ModelName = [ELC]Design_space_SRAD v1 v24

Opening Model
Preprocessing

Name = [ELC]Design_space_SRAD v1 v24 Rigidità max Massa 25% (8)
stmodOffsetId = 52

RunType: OPTIRUN_ANALYSIS = 1, OPTIRUN_TOPOLOGY = 2, OPTIRUN_TOPOGRAPHY = 4, OPTIRUN_SHAPE = 8, OPTIRUN_GAUGE = 16, OSSMOOTH_REANALYSIS = 128

RunType = 2

Using CAE_HOME='C:/Program Files/Altair/2024.1/Inspire2024.1.1/hwx/bin/win64/'
Time 20:29:5

RUN_STATE = 1 ## Preprocessing Done
RUN_STATE = 3 ## Meshing
#--------Generating a fresh Mesh using Simlab
#--------Simlab version = SimLab Version 2024.1 Beta Built on Jul 18 2024 at 18:16:23
RUN_STATE = 7 ## Meshing completed successfully
RUN_STATE = 15 ## Realizing model
#--------Warning while contact (DESIGN SPACE) realize as 1 features seems got merged during meshing.
#--------Unable to realize contact (Contatto 164) as UdbSet for Primary features can not be created
#--------Unable to realize
RUN_STATE = 527 ## Synthesis run completed

Error - Unable to realize synthesis data

RUN_STATE = 1551 ## hwfdmfeprocessor process completed

I have attached my file, and I would be grateful if you could assist me.

Answers

  • TonyG
    TonyG
    Altair Employee

    Hi Giovanni

    Firstly I tried to run with the OptiStruct solver it appears as if the contacts are not being realised correctly and there is rigid body motion. This will stop the optimisation running correctly.

    So I went looking deeper. The first thing I found was a degenerate piece of geometry. Pieces like this can cause all kinds of issues. So I removed this and tried to run but it failed again…

    Next I tried re-defining all the contacts as part-to-part as I find it can be more robust and on models like yours easier to be confident everything is defined. That worked and the model ran fine.

    So now for the optimisation… with the loads applied the starting SF is ~50 you will need to remove lots of material! It ran OK with a SF of 12 but didn't leave much....

    In cases like this it can be better to run a maximise stiffness at a variety of different percentages. You can multi select and launch all at once by holding the control key.

    You can very quickly generate a number of concept designs, perform re-analysis to evaluate etc.

    One final note… some parts in your model are detached after optimisation. If they are in your design they probably have a purpose and will be loaded - if not delete them :-) Example below…

    Hope this helps! Attached is a running model with baseline analysis.

    Tony