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.

Minimum element quality statistic documentation

Bjarne Christian Hagen

Please login with a confirmed email address before reporting spam

Hi you all

I'm looking for documentation of Minimum element quality. I found this quote:

"Minimum element quality statistic is a telling one. Usually, a value greater than 0.1 is all we need."
in www.comsol.com/blogs/size-parameters-free-tetrahedral-meshing-comsol-multiphysics/

Why is 0.1 OK? There must be a reference somewhere.

Can somebody help me?

Best regards
Bjarne

3 Replies Last Post 1 déc. 2015, 04:04 UTC−5
Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 30 nov. 2015, 13:38 UTC−5
Hello Bjarne,
There is nothing critical about the specific value of 0.1. This number is provided as a rule of thumb.
Best regards,
Jeff
Hello Bjarne, There is nothing critical about the specific value of 0.1. This number is provided as a rule of thumb. Best regards, Jeff

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 30 nov. 2015, 14:20 UTC−5
Hi

the "mesh quality" gives an indication of the length to width ratio of your elements, for a homogeneous material and shape a regular mesh of quality close to 1 is mostly ideal, but on thin elements, that have anisotropic dimensions, one easily get mesh qualities well below 0.1.

In such cases it's often interesting to carefully check the results and remesh, either with brick elements or some swept mesh. A thin membrane, meshed in 3D might only require a few (>=5) elements along the thickness, but could fully solve correctly with a large element size in the tangential directions, even above a ratio of 1:100, it all depends on your model. Which means one need to solve it first, analyze the results and the mesh (i.e. plot the mesh with a filter to look only at the "poor quality meshed") and remesh to solve again, and compare, until you believe its OK, and that the time to solve is acceptable.

You need to reconsider your mesh each time you add a new physics.

--
Good luck
Ivar
Hi the "mesh quality" gives an indication of the length to width ratio of your elements, for a homogeneous material and shape a regular mesh of quality close to 1 is mostly ideal, but on thin elements, that have anisotropic dimensions, one easily get mesh qualities well below 0.1. In such cases it's often interesting to carefully check the results and remesh, either with brick elements or some swept mesh. A thin membrane, meshed in 3D might only require a few (>=5) elements along the thickness, but could fully solve correctly with a large element size in the tangential directions, even above a ratio of 1:100, it all depends on your model. Which means one need to solve it first, analyze the results and the mesh (i.e. plot the mesh with a filter to look only at the "poor quality meshed") and remesh to solve again, and compare, until you believe its OK, and that the time to solve is acceptable. You need to reconsider your mesh each time you add a new physics. -- Good luck Ivar

Bjarne Christian Hagen

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 1 déc. 2015, 04:04 UTC−5
Thanks Ivar and Jeff!!
Thanks Ivar and Jeff!!

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.