Jump to content
Sign in to follow this  
rorythornton

Meshing Strategy

Recommended Posts

Hello

 

I have been having a little dig through the mesh and aided mesh options in Flux to understand what the best meshing strategy is for simulating the interactions between static magnets (mainly disc and spherical permanent magnets).

 

I have done some research into these settings already (e.g. understanding what bubble packing is for face meshing), but I was wondering if anyone here has good knowledge on what settings get the highest level of simulation accuracy (while trying to minimise the number of required nodes and computation time)? Also, if anyone is able to point me in the direction of good literature about optimising meshes for magnetic interactions, that would be appreciated.

 

I do have some specific questions on the options available for mesh and aided mesh options:

  • Relative precision for distance between 2 points is 1e-5 for both geometry and mesh (despite flux recommending that the geometry is superior to the mesh) 
    • Is this a good starting value to do comparisons with (and is there any research/literature to support this)?
  • Is there any existing literature or experiments that compare the flux mesher to the MeshGems mesher to compare the differences in simulation accuracy?
  • For uniform vs Dynamic mesh points
    • Is the dynamic mesh point generator iterative? It doesn't seem like it based on it's description, but when I research "Dynamic mesh points", all I find is information on types of "Adaptive mesh refinement"
    • If dynamic mesh points are not iterative, is it only useful for reducing computation time?
  • Can anyone recommend good literature that compares relative node deviation (equal number of nodes on a given line) to absolute (equal spacing between all nodes)?
  • I am assuming the only benefit of mesh relaxation is to save computation time?
  • Can anyone recommend literature on the advantages/disadvantages of using, what flux refers to as, the aided shadowface (which is disabled by default)?

 

Another question I have is:

  • how do meshing requirements for electromagnetic simulations vary from structural ones? Can I use literature for structural FEA meshing and apply that to what I am doing here?
  • If anyone is able to recommend literature that compares the two, that would be appreciated

 

If anyone is able to comment on even one of the many points listed above, to help me understand how to optimise my meshing, I would greatly appreciate it!

 

Many Thanks


Rory

 

Share this post


Link to post
Share on other sites

Hi,

 

In the finite element method, the accuracy of the results generally depends on the quality of the mesh used. there are places where we refine the mesh and other, we relaxation of the mesh. You will find attached a PDF file (page 213 to 261) explains the principle and different methods used in flux to mesh the devices. Also, a PDF file about ‘meshgems’, the principal and advantage.

Hope this will help.

Best regards.

01_New_Features_2019.pdf 01_Vol1_Environment_Geometry_Mesh_Import.pdf

Share this post


Link to post
Share on other sites

Hi Abdessamed,

 

Thank you for this, very helpful.

 

One particular section of the "2019 Altair Flux - User Guide" that I found useful was the one on "Examples of criteria to validate a mesh" (Page 219). This lists out checks that can be done on the simulation results to validate the accuracy of the mesh. Examples are:

  • Checking that the reaction forces balance the input loads
  • Looking for cracks in the field lines generated within a single region

These are very useful as they dont require a comparison of the simulation results to an experiment or test. Are there any other ones that you (or anyone else reading this) use to help sanity check your meshes? Or alternatively, is there any good literature on this subject?

 

Moreover, other FEM softwares have in-built mesh checking tools. For example:

  • Element error/quality checking (Distorsion checks for example)
  • A computed model confidence % calculated while solving the solution

Does Altair Flux have any kind of checker or convergence tool?

 

Many Thanks

 

Rory

Share this post


Link to post
Share on other sites

Hi,

In Flux there is an option to check the quality of the mesh. Form the mesh in the top:

  • Select mesh

  • Select check mesh

Once you have clicked on check mesh, you will get:

  •  Number of elements not evaluated: %

  •  Number of excellent quality elements:  %

  • Number of good quality elements:  %

  • Number of average quality elements:  %

  • Number of poor-quality elements: 0 %

  • Number of nodes: 7575

  • Number of line elements: 1035

  • Number of surface elements: 3752

  • Mesh order: 2nd order

You have the possibility to add the mesh defects (poor quality elements) to mesh visualization. To do it: from mesh >> select mesh defects >< right click: add to visualization. This give you and idea about the localization of poor elements.

Best regards.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
You are posting as a guest. If you have an account, please sign in.
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...