Some further information:
- no errors are detected by PETSc. By using it in place of Trilinos the
code seem to work OK.
- if only the rhs is distributed, no errors again. Errors are due to the
matrix distribution in Trilinos.
Thank you again
Alberto
Il giorno mercoledì 23 dicembre 2020 alle 0
Dear community
After Timo's suggestions I attempted at assembling matrix and rhs for
problems with an interface, using continuum elements (not DG).
I mean at this: take two domains connected by springs and pull them apart,
so to separate the domains and elongate the spring.
Since the forces prov
So it does seem to be the Cmake version. The problem showed up with Cmake
3.19.1 and updating to 3.19.2 fixed it apparently. Wanted to share if
anyone else comes across this.
On Friday, December 18, 2020 at 11:42:41 AM UTC-6 d.arnd...@gmail.com wrote:
> Zachary,
>
> are you using CMake 3.19 b
I would like to bring your attention to the publication
The deal.II finite element library: design, features, and insights
Computers & Mathematics with Applications, vol. 81, pages 407-422, 2021.
https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoi.org%2F10.1016%2Fj.camwa.2020
Timo,
thank you. These two files are extremely helpful and I guess I understood a
lot from them. Extremely well written, by the way.
I really appreciate it.
I would like to associate to each gauss point on the interface, identified
by the command
const auto &q_points = fiv.get_quadrature_po