mesh refinment study

nylanyla Member


I would like to perform a sensitivity analysis on the mesh in order to find the best mesh element size to get accurate results.

I know I have to start from a certain element size and reduce it by a factor of 1.3 or 1.5 and check for the output. But I have some questions and I would appreciate if you could clear my doubts.

1. Is it correct if I keep the same general element size and refine the size of elements in a region of interest (body sizing)?

2. Can I look at the von Mises stress output?

3. The plot to study the mesh refinement should include all bodies or only the output related to the region of interest bodies? 

4. How can I understand when to stop to refine the element size?

thank you!


  • peteroznewmanpeteroznewman Member
    edited June 17

    1. Yes, you can refine locally around the maximum stress.

    2. Yes, that is best for ductile materials.

    3. Region of interest

    4. The simple way is when the plot looks like it is converging on a finite value at a zero element size. A stress singularity will not do that.

  • nylanyla Member
    edited June 17

    Hi Peter,

    I have got this plot 

    mesh refinement

    If I try to reduce one more time the element size then simulation interrupts due to hardware performance problems, thus could I decide to use a size 0.2 since it is close enough to the stress value related to the smallest element size? In that way I could spare some computation time and effort.

  • peteroznewmanpeteroznewman Member
    edited June 18

    That plot shows 0.2 mm is an acceptable element size.

  • nylanyla Member
    edited June 18

    ok, many thanks!

Sign In or Register to comment.