Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by Q.Nguyen-Dai

  1. Element Quality checking gives you some ideas about the quality of your mesh. It's not so easy to say "good" or "bad". Depending to your analysis goal, to analysis type, to your FE solver, ... and depending to your experiences too. A "not-so-good" meshing for one solver maybe "good enough" for another solver, another analysis. Some general rules I do with TET10 meshing with SAMCEF solver: Aspect <= 3~4 Jacobian >= 0.6~0.7
  2. Try my code: *createmarkpanel sets 1 "Select sets " set set_ids [hm_getmark sets 1] *clearmark sets 1 foreach id [lsort $set_ids] { set name [hm_entityinfo name sets $id] *createentity comps name=$name *createmark elems 1 "by set id" $id *movemark elements 1 "$name" *clearmark elems 1 } HTH,
  3. Your black solid is not enough fixed yet. So you can't load your structure. Please check again your constraints.
  4. How do you do to get these H3D? Normally, at the end of anlysis, we got only one H3D with several loadcases, right?
  5. IGES is only my latest working option. It's better with STEP, Parasolid.
  6. 1. Extract face from 2 current components 2. Split geometrical surfaces by faces from 1. 3. Do surface mesh on surfaces created on 2. 4. From 1. & 3. you got closed 2D mesh => Fill with TET mesh to get volume. HTH,
  7. Yes, Hyperview Player is your (best) solution. Just export your results into H3D format and use free software "Hyperview Player" to visualize.
  8. -nt nnn Specify number of CPU cores in multiprocessor (SMP) run -nthread nnn -''- I can't found the option "-ncpu" in help page of Optistruct. Which version do you use?
  9. Could you share a sample of Polyflow output? I know nothing about this software. But if you show me the format, maybe I'll try a script.
  10. It's not complicated. Just do in loop: mesh => check => if > 0.35 then fixe that => check => ...
  11. Sometimes I have the same problem. I have to go out completly the current panel to continue without issue. I think v14.0 has some graphic bugs.
  12. Yes, you should have a good and supported graphic card. I use Quadro 2000M with 2048 Mo DDR3 on my working laptop.
  13. Do you know something about TCL programming?
  14. Tell us about: Your hardware (cpu, ram, hdd, graphique card,...) Your HM model: size, complexity,...
  15. Write a TCL script to get out all mid-side nodes from your selected nodes.
  16. Have a look at this thread The same question, the same problem, why changed topic?
  17. If mid-surface extraction tool does not work, you can build yourself mid-surfaces, of course manually.
  18. At your place; I'll try fix that manually. By this way, I'm sure for quality of the mesh.
  19. That depend to your solver, to your objective. For example, with FE solver, 1st order TET elements are not good (too stiff) for Static analysis. 2nd order TET elements are better. However, with Explicite solver (like LS-Dyna, MSC/Dytran,...) you can not use high-order elements.
  20. You can correct this issue manually. It's not so complicated. Have a look at my test (I used HW14. Zip file have FEM format). sent-2.zip sent-2.hm
  21. That's simple. Create 2 shell properties: one for embbed 2D & another for enclosing 2D mesh. Create two different components for each 2D mesh. Apply the property for each component.
  22. Make TRIA mesh for your box firstly and then try TET mesh. Have a look at my test: my TRIA 2D mesh is yellow. My 2D mesh of the box is green (closed volume). The final 3D shares the same nodes as TRIA 2D mesh.
  23. If you work with external/internal faces, you can define "offset" for your shell. Create new component Organize => surf => Select solid surfaces => copy
  24. Before to woth with this kind of meshing, don't forget that mid-surface is NOT always the best choice. Some times it's better to work with external/internal faces.
  • Create New...