Menu

Discussion Forum

New Discussion
Filtrer

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Meshing issue - 3D porous model

Posted 20 févr. 2022, 06:50 UTC−5 Geometry, Mesh, Modeling Tools & Definitions Version 5.5 1 Reply

Ibrahim
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi,

I have a porous structure obtained with microCT. I am having issues regarding building a geometry in COMSOL. Mainly, I am getting out of memory error when trying to import the file. The unprocessed .stl file contains 5121132 vertices and 10283881 triangles. Using meshmixer I have tried to fix all the issues and reduce the amount of vertices. However, nothing I have tried to so far has worked.

I can see the mesh in the mesh parts after importing however, due to memory issues I cannot build the geometry. Ideally, I would like to do a fluid flow through the material but I am not really sure how to fix this issue. I am attaching a view of the mesh part imported to the COMSOL.

Any help would be highly appreciated. Thank you.

Win-10 x64 Intel(R) Core(TM) i7-10610U CPU 32 GB RAM

1 Reply Last Post 20 févr. 2022, 19:52 UTC−5
Robert Koslover Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 2 years ago 20 févr. 2022, 19:52 UTC−5

I suggest you consider the following: 1. Model something less complicated, but which still advances your understanding of the problem of interest to you. 2. Use the Comsol application mode with the simplest formulation of the physics that you are trying to study. 3. Customize the mesh in detail (don't simply use Comsol's default/automatic settings) so that you can get the most bang for the buck. Use coarser meshes in regions of slow variations, finer meshes in regions of strong gradients. 4. If there is any possible symmetry you can apply in your model? Take advantage of it. 5. 32 GB is probably not enough. Obtain a far more powerful computer with much, much more RAM (e.g., 100s of GB of RAM would be a good start).
6. Use a more memory efficient (but yes, possibly painfully slow) iterative solver, if you aren't already doing so. Experiment with solver settings. 7. Set discretization to lowest-order (e.g., linear) elements and the coarsest meshes that you can get away with, if you aren't already doing so, at least to start.

Once you have a model that actually executes without running out of memory, you can start improving its validity by (for example) using a finer mesh, modeling a larger volume, etc. Do simpler smaller models of the physics of interest first. Only after you have those working, should you attempt large, complicated, computationally-demanding models. See the Knowledge Base for more. For example, see https://www.comsol.com/support/knowledgebase/830 https://www.comsol.com/support/knowledgebase/866 https://www.comsol.com/support/knowledgebase/1267

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
I suggest you consider the following: 1. Model something less complicated, but which still advances your understanding of the problem of interest to you. 2. Use the Comsol application mode with the simplest formulation of the physics that you are trying to study. 3. Customize the mesh in detail (don't simply use Comsol's default/automatic settings) so that you can get the most bang for the buck. Use coarser meshes in regions of slow variations, finer meshes in regions of strong gradients. 4. If there is any possible symmetry you can apply in your model? Take advantage of it. 5. 32 GB is probably not enough. Obtain a far more powerful computer with much, much more RAM (e.g., 100s of GB of RAM would be a good start). 6. Use a more memory efficient (but yes, possibly painfully slow) iterative solver, if you aren't already doing so. Experiment with solver settings. 7. Set discretization to lowest-order (e.g., linear) elements and the coarsest meshes that you can get away with, if you aren't already doing so, at least to start. Once you have a model that actually *executes* without running out of memory, you can start improving its validity by (for example) using a finer mesh, modeling a larger volume, etc. Do simpler smaller models of the physics of interest first. Only after you have those working, should you attempt large, complicated, computationally-demanding models. See the Knowledge Base for more. For example, see https://www.comsol.com/support/knowledgebase/830 https://www.comsol.com/support/knowledgebase/866 https://www.comsol.com/support/knowledgebase/1267

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content