Hi Edscott,
give AddressSanitizer a try, it is a great tool to find such thing you
describe. Further MemorySanitizer and Undefined Behavior Sanitizer might
turn out to be helpful, too. My best experience with these tools were
when I used the latest Clang compiler. Some of them work with recent
Dear Edscott,
what was your actual problem you are trying to solve? Is this a problem
with Dumux? The code you are showing looks perfectly fine to me.
Or was this just occurring during debugging? Your comments sound like you are
debugging optimized code. If yes, is there any problem if the
OK.
I'll dig into the matter a bit further to see if I can solve where the
problem arises. It might be an incorrect cast somewhere that screws up
memory locations.
Best regards,
Edscott
2018-08-17 15:31 GMT-05:00 Flemisch, Bernd <
bernd.flemi...@iws.uni-stuttgart.de>:
> Hi Edscott,
>
> can
Hi Nikolai,
you have to make sure that you only compute this on Dirichlet faces. On
Neumann faces the fluxes are the ones you set in your neumann() or
neumannAtPos() function and cannot be computed by the flux assembly. So
if you have only no-flow Neumann boundaries just make sure to only
Hi Dennis,
Thanks a lot, I got through elemFluxVarsCache.bind runtime error!
However, a new problem has popped up
When I get to computing (or recovering) the advective flux with
fluxVars.advectiveFlux(nPhaseIdx, upwindTermN), the execution brings me to
caluclating Dracy's flux in
Hi Nikolai,
I think your implementation is almost correct. If you want to balance
masses, you have to add the density to your upwind terms though, i.e.:
auto upwindTermN = [](const auto& volVars) { return
volVars.mobility(nPhaseIdx)*volVars.density(nPhaseIdx); };
...
Also, you have to
Dear DuMuX experts,
Please advise how can I implement a global conservation error check, that is,
evaluate the residual ( mass(t^{n+1}) - mass(t^n) ) / dt + sum of fluxes over
the domain boundaries.
So far I can evaluate the mass in the domain by looping through the elements as
in void
