Hello,
For benchmarking purpose I would like to use KROME with DVODE instead of DLSODES.
However, I had some issues when I used the -useDvodeF90 option. Is DVODE still supported by KROME?
One issue was that the arguments of fex were requiring proper intent(in)/intent(out) statements. However, this might be compiler dependent.
Another issue was that krome_init seems to be incomplete e.g. the cooling is not initialised at all.
At the end I took krome_all.f90 from the DLSODES version and just replaced the subroutine krome with its DVODE version.
This code could be successfully compiled (with the intent(in)/intent(out) statements in fex).
If I run the code I get a lot of these messages:
At T(=R1) and step size H(=R2), the
corrector convergence failed repeatedly
or with ABS(H) = HMIN.
However, KROME continues and I also get a result at the end but it is very different from the one obtained with DLSODES.
Due to the issues mentioned above I don't know if there is something wrong with the
implementation or if this 'error' message means that DVODE cannot solve the ODE system. But if it indicates a serious error why does KROME continue with the integration?
And if the solution obtained with DVODE is fine why is it so different from the DLSODES one?
Thanks a lot for your help!
Kind regards,
Jan Bolte