Do you have results for test examples?:
Example files: crackIIcum, crackIIint, crackIIprin
I’m asking because I have it for version 2.20 but not for version 2.21 (Windows), the program suddenly ends the calculations
I can’t find this one. But the remaining two work for me in ccx 2.21. Do you have the masterII.frd file in the same directory ? It’s used by *CRACK PROPAGATION,INPUT=masterII.frd
I guess I’m using it correctly because running it in ccx_2.20 works, but the same example running on 2.21 doesn’t.
Are you using Windows or Linux?
I though that maybe you have separate folders for them (and one is missing this file) or something.
I’m using Windows 10.
Don Guido has been working on Cracks very recently. Not sure if something was fixed or just improved.
Yes but it was after the release of 2.21. Apart from that, the most recent reported activity related to cracks is from December 2022. However, 2.20 was released in August 2022 so those changes should be included and might be the cause of potential issues.
It’s very recent.
5 Dec 2023 - check crack length modification.
8 Dec 2023 - finished crack length modification.
Yeah, so it will be in 2.22 since 2.21 was released in August 2023.
Talking about source code updates, I wonder what is being changed in the shape functions of some elements:
You can see the changes here but I’m not familiar with the source code so no idea what it does in practice: "changed shape3tr, shape4q, shape6tri and shape8q.f" · Dhondtguido/CalculiX@37317f7 · GitHub
Yes, actually I tried to understand it but had no luck hahaha
I think Guido is trying to handle very small values in the calculation. Specifically, it appears to be used as a modified value of the Jacobian vector (xsj) when certain conditions are met - like if d1.lt.1.d-30 or something. I assume it has to do with assigning the right directions for the components.
I could be wrong…