Learning Forum

[thach-hoang.nguyen]

Hello everyone,

I have a question regarding the method ANSYS uses to solve the equation K⋅U= F. Currently, ANSYS employs the Newton-Raphson method for this purpose. However, I am interested in implementing an entirely new mathematical formulation to solve this equation. Is it possible to integrate a new solution method into ANSYS?

Is there someone has done this before? Thank you and have a great day.

[David Weed]

Hello,

This sounds like it might fall under User Programmable Features (UPF)/customization. It doesn't appear that we currently have a way to modify the solving scheme but, can you let me know what alternative solving scheme you would like to employ?

[thach-hoang.nguyen]

Hello David,

I'm interested to employ Newton semi-smooth method. If its possible, i would love to discuss with you further via email?

Thank you and i will wait for your reply

[David Weed]

Hello Hoang,

I would very much like to help but, per ANSYS policy, we are not allowed correspond over email with forum guests. If you have any documentation or reference for the semismooth NR method, I can initiate an internal discussion about it and what alternatives we might have.