Jump to content

Ingeniorator

Members
  • Content Count

    25
  • Joined

  • Last visited

About Ingeniorator

  • Rank
    Beginner

Profile Information

  • Are you University user?
    No

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Which tutorial is it? Most likely, the result is due to your constraints. The solver tries to satisfy the constraints at first, and only then optimises towards the objective even if the minimum found is above the initial value. It is simply the optimum feasible with the given constraints.
  2. Hi again, I ran into another issue with the import functionality. When I import an input file like the one attached to this post, HM sometimes throws the following warning: Warning: Dimensions of PBEAML ID 4 are different from Hyperbeam comments, a new Hyperbeam will be created Warning: Dimensions of PBEAML ID 5 are different from Hyperbeam comments, a new Hyperbeam will be created Warning: Dimensions of PBEAML ID 5 are different from Hyperbeam comments, a new Hyperbeam will be created Warning: Dimensions of PBEAML ID 6 are different from Hyperbeam comments, a new Hyperbeam will be created Summary:Feinput finished with 0 errors and 4 warnings. This differs of course per model and does not always show up. The odd thing is that all PBEAML entries are defined the exact same way and I did not find a difference to the way HM does it when comparing it to a corrected and exported input file from HM. It's very odd as well that the PBEAML entry with ID 5 gets mentioned twice. Does anyone have any advice? testlattice_from_matlab.fem testlattice_from_HM.fem
  3. Coming from other FEA/CFD packages and having used Hypermesh previously only as a preprocessor, I taught myself Optistruct on the fly. But I see that I have a lot of catching up to do. Thank you very much for the explanation and the links.
  4. Hello, that's promising news. Can you give a few more details about this? When might it be released and what kind of external formats will be supported? Only Optistruct files for example or others as well?
  5. Thank you very much, it works with RBE2 elements now. I wonder however why RBE3 elements cannot be used for a displacement constraint in this case as it runs without issues as a standard analysis or topology optimization.
  6. Hi all, I tried all day to get a shape optimization running, but I keep getting an error: *** ERROR # 339 *** The dependent d.o.f. is constrained by grid or spc data. RBE3 element id = 25502. grid id = 28613. component = 1. This does not happen when no RBE3 elements are used. I interpret this error message that a constrained node is also supposed to be optimized, but these nodes should be separated in this model. I'd be very grateful if someone could point me to my error. testblock.fem
  7. Hi, I have a similar error in a free shape optimization: *** See next message about line 55 from file: F:/Temp/Optistruct/shapetest/HM19.fem (this is continuation line 2 of DSHAPE,1 bulk data) "+ GRID ID 0" *** ERROR # 1000 *** in the input data: Incorrect data in field # 4. A simple analysis runs perfectly fine and I don't know where this error comes from exactly. When looking at the section in the input file: $HMNAME DESVARS 1designvar1 DSHAPE 1 + PERT BOTH 0.5 10 10.0 10.0 1 1 + GRID ID 0 I defined the nodes to be optimized as a set which has the ID 3, however in the input file this is not represented correctly! When importing the same input file for testing, an error message shows up: Warning: No grids are defined for DSHAPE with id 1 Resolution: This may result in incorrect definition. Summary:Feinput finished with 0 errors and 1 warnings. When I manually change the DSHAPE section to reference the node set 3 and start Optistruct as standalone instance, it asks for a design objective reference to proceed with a minmax problem, which is not what I defined. What's going on here? HM19.fem EDIT: I found out why this was the case. I read somewhere in the HM files that you should use the menu on the lower part of the screen instead of the model tree on the left and it worked now!
  8. Go to the Tools panel and select Scale in the middle column.
  9. Yes of course you could.
  10. Are the nodes at the domain borders connected? Check by running equivalence.
  11. Hi again. For whom it may concern, I managed to get it to work correctly with all the elements moved to the right components and their properties, materials and beam sections assigned as they should be. For this to work, the import option has to be set to "By HM comments" and HMMOVE is necessary to move the elements to the correct component, otherwise manual editing is needed.
  12. You can access the Organize panel either by pressing Shift+F11, or choosing the Tools panel (meaning the bullet points on the right side of the lower screen menu) and picking Organize in the leftmost column. Also, why don't you use general contact in Abaqus instead of multiple contact pairs?
  13. Oh that's it! Great, thank you very very much! That solves one issue at least I may be back in the next days if I can't figure out the rest. Another issue, this time a graphical one. Sometimes, when manipulating the properties and/or beam sections, all elements or just part of them vanish. They are still "displayed" as they are still considered as displayed (for example Choose by displayed gives the correct element number) and I can even select them manually, but they can't easily - or at all - be rendered visible again.
  14. Well, if I don't use $HMMOVE, the elements just get all thrown in the misc1 component which is not what I want. They need to be organised by component, each component needs the right property and each property needs the right beam section assigned to it. But what's the error in my input file since it's not imported this way? Apparently, some elements are moved to the wrong collector and some aren't moved at all. Is there a line limit for $HMMOVE? Also, what does $HMDPRP do exactly? The description is very short and does not give much information and it does not seem to work the way I think it might since it makes no difference in my input file.
  15. Hi again, I've progressed somewhat with the task. However, I still encounter a few issues for which I don't find the solution at the moment. As you can see in the input file created with Matlab, there are 6 components, beam sections and properties as well as one material and one beam section collector. However, the beam sections, properties and material don't get associated correctly and a number of elements get put into a newly created "misc" collector upon import in Hypermesh. I suspect the $HMMOVE, $HMDPRP and beam section definitions to be faulty. However, when compared to an exported file from HM after having organised everything (as much as possible), I don't see a big difference. What's the cause of this? Is it the free format vs. fixed format? thicklattice_fine_from_matlab.fem
×
×
  • Create New...