Write out Newton Raphson residual maximums

Hey togehter,

I have a simulation which takes about 500 iterations. It has about 500 contacts. I want to figure out the location the simulation has difficulties of convergence. Therefore, I would like to write out the coordinates of the max. of the newton Raphson residuum during the whole simulation into an csv. file.

Maybe also with MAPDL.

Find more posts tagged with

Accepted answers

All comments

James Derrick

Hello @cobonz do you have any code that you've attempted to write already? How far did you get?

cobonz

Hi James,

I am not very educated with APDL code.
I tried to use
"
NLDIAG, EFLG, 1, 999
"

I could also use
"
NLDIAG, NRRE, 1, 999
"
but there I do not really get any coordinates. I do not want to produce a lot of data.

Matthias

@cobonz
If you are sure that the contact convergence is causing most of the issues you can look into the cnd-file. The default name is file.cnd.

If you are using Ansys 2025R1 you can benfit from the cleaned up format of the file. The type of information you can find is (example from 2024R2 file):

Contact Pair ID
Number of Contact Elements in Contact
Number of Contact Elements in Contact (Sticking)
Max. Chattering Level
Max. Penetration/Min. Gap
Max. Geometric Gap
Max. Normal Stiffness
Min. Normal Stiffness
Max. Resulting Pinball
Max. Elastic Slip Distance
Max. Tangential Stiffness
Min. Tangential Stiffness
Max. Sliding Distance of Entire Solution
Max. Contact Pressure
Max. Friction Stress
Average Contact Depth
Max. Geometric Penetration
Number of Contact Points Have Too Much Penetration
Contacting Area
Max. Contact Damping Pressure
Max. Contact Damping Tangential stress
Max. Sliding Distance including near field
Min. Sliding Distance including near field
Max. normal fluid penetration pressure on contact surface
Max. normal fluid penetration pressure on target surface
Total volume loss due to wear on contact surface
Total strain energy due to contact
Frictional dissipation energy
Contact damping dissipation energy
WB contact pair ID
Total contact force due to pressure -x component
Total contact force due to pressure -y component
Total contact force due to pressure -z component
Total contact force due to tangential stress -x component
Total contact force due to tangential stress -y component
Total contact force due to tangential stress -z component
Number of Contact Points Have Too Much sliding
Contact pair force convergence norm
Contact pair force criterion
Max. tangential fluid penetration pressure on contact surface
Max. tangential fluid penetration pressure on target surface
Max. sliding distance of current substep for closed contact

Amongst others there are
Contact pair force convergence norm
Contact pair force criterion

cobonz

Hi Matthias,

thanks that definitely helps. Though usually i have a lot of contacts (~400-800). Therefore, I would like to reduce the number I have to look at by putting out the location of trouble through Newton-Raphson maximums.

Matthias

@cobonz
Or you can look in the gst-file. GST = Graphical Solution Tracking. Unfortunately, the file format is not that user friendly for automated reading. Only one value per column is stored in each iteration.

Time
Load Step
Sub-step
<?WORD FLAGGED?> Iter
Time Incr
Max DOF Incr
Line Search Parameter
Bisection
F CRIT
F L2
U CRIT
U INF
Remesh
Max Resi F
Node with Max DOF Incr
DOF with Max DOF Incr
Node with Max Resi F
DOF with Max Resi F
Max Resi M
Node with Max Resi M
DOF with Max Resi M

Of special interest for you would then be:

F CRIT
F L2
Max Resi F
Node with Max Resi F
DOF with Max Resi F

You could for example then also calculate the ratio of "Max Resi F" to "F L2"