Learning Forum

[qian qian]

Dear friends,

 

I am currently attempting to simulate wafer grinding, with the wafer rotating at 250 rpm. However, deformations occur in both the wafer and the support frame during the simulation, causing non-convergence.

 

Two potential solutions come to mind:

 

Setting the initial rotation speed of the wafer to 250 rpm. However, I haven't found the function to apply the initial rotation speed yet.

Increasing the number of simulation substeps. I have already increased it to 400 steps, but deformations still occur. I'm unsure of how many more steps need to be added to resolve this issue.

Could you please advise on how to improve the simulation?

Thanks if anyone can help solve this.

[dlooman]

This isn't an easy analysis, but the ICROTATE command can specify an initial rotation.

[qian qian]

Hello~ Thank you very much for your response.

1. Does setting the initial rotation speed ensure that no deformation will occur?
2. Apart from setting the initial rotation, are there more ideal methods to set high-speed rotation, such as setting substeps?

 

[dlooman]

High speed rotation analysis with friction is borderline impossible for a couple reasons.  One, finite elements can't be rotated more than 2 or 3 degree per step.  Two, the element faces for contact are constantly changing so the rotation should not produce translational motion of more than 1/2 element length per step.  Even with small time steps convergence will constantly be slow making the run time almost unlimited.  Technology Showcase Example 28 in the APDL doc. is a simpler analysis than yours, yet takes 24 hours to run.  What is the object of the analysis?  What do you want to find out?  

[qian qian]

Dear Dave Looman,

I want to add random vibration load during the wafer grinding process and analyze the stress distribution on the wafer, but I encountered problems during the pure grinding process. I have thought about using the ls-dyna module before. Analysis, but there seems to be no way to increase the random vibration load.

Is there any recommended way for me to conduct simulation analysis?

Thank you for your time and consideration.

Sincerely,
Qian

[dlooman]

Random vibration is a linear mode based analysis.  It might be possible to include the grinding analysis as the static portion of a prestressed modal analysis, but as you appear to have found out, it's hard to analyze grinding in Ansys Mechanical.  LS-DYNA would be a better choice for the grinding analysis.  Random Vibration analysis will require physical testing to determine the input PSD.