Tags: nuc-astro/WinNet
Tags
Movie script (#15) * (-) typos in docu Removed typos in the documentation that Jan made me aware of. * (*) Default number of threads=1 Set the default number of OMP threads to 1 in case OMP_NUM_THREADS was not set. Before it was taking the maximum cores. Since PARDISO does not parallize well, 1 is the better choice. * (*) Ascii flow not outputted in zeroth iteration The flow was producing an empty file in the zeroth iteration as there was no flow yet. Removed this output. * (+) movie script Added a script to create a movie of a WinNet run. Therefore also changed the Dockerfile that all dependencies are installed in docker as well.
(*) Rehaul of fission treatment into different channels (#11) * (*) Rewriting of top_engen_every parameter The top_engen_every parameter was only working when nuclear heating was turned on (and even then it was called several times instead of one time). Moved everything to analysis.f90 and also implemented that not only the changing isotopes but also the top contributing reactions are outputted. All these changes also have been described in the documentation of the parameters. * (*) Small format change in nuclear energy output Appearantly it could happen that a number like 1....-101 (without E) for the format were printed when number smaller than 1e-99 occured. This makes it difficult to read the number later. Introduced a third digit for the exponent. * (-) bug in tabulated rates The amount of counted tabulated rates and actual read tabulated rates could mismatch resulting in an later not understandable error. Fixed it to count tabulated rates correctly.... * (+) Major improvement in fission Fission rates got split into different files for different types of fission. This removes the necessity of having hardcoded flags in the code. Furthermore, beta-delayed fission can now be given in terms of probability from the total beta-decay rate (as given mostly in literature). Additionally, added the possibility to have beta-delayed neutron emission fission. The different fission types can be given in different formats that are specified with new parameters. Unfortunately, this commit breaks compatibility to the older versions as the paramter fission_rates got removed. The example cases still have to be adapted accordingly... * (*,+) Fission module, new fissflag Restructered fission more, kicked out rrate_fiss as it was not used. Made fission fragment subroutines easier to read. Added more comments. Added a new fissflag (4) to completely customize your fragment distribution, this comes with a bunch of new parameters (sfission_file, fission_frag_spontaneous,...). * (*) Parameter files for example cases and documentation Changed the example cases to work with the new parameters for splitted fission treatment. Added beta-delayed fission rate file of Mumpower et al. 2022. Changed the documentation and description accordingly. * (+) variable tabulated temperature grid * (*) made tabulated_rate non-private for compatability with rate_variation * (+) documentation, monotonicity check, test case Added a check for monotonicity in the tabulated temperature grid. Made the grid a bit more flexible in the sense that it does not have to be separated by exactly one space. Added test case. * (*) Used wrong error code * (-) bugs from merging The fission needed access to the tabulated rates and their treatment had changed. Made changes to make everything work again. * (+) error message for fission probabilities larger than 1. * (*) default value of beta-delayed fission format Now the default value is 2: probability format * (*) Fixed rosswog fission example The fissflag was accidentally set wrong when implementing the fission treatment. * (-) bug when compiling with gfortran One statement in the legacy abla subroutines had an "inout" flag. Changed it to intent(in) since it is not changed. This has no impact in any case since the routine is not used.... --------- Co-authored-by: Max Jacobi <[email protected]>
Automatic package creation on release. (#10) * (+) automatic docker file via gh packages Added workflow to automatically create and deploy a docker image with a working WinNet environment. This environment contains the intel fortran compiler and the hdf5 library compiled with ifx. The workflow is triggered whenever a new release is published and the package will be visible under "packages" on the github page. An instruction on how to use this is added to the documentation. * (-) bug in makerun.py. The folders "snaps/" and "flow/" were not created by the makerun when only the custom_snapshot parameter was present and not the "snaps_every" or "flow_every". This resulted in an error when trying to run winnet as winnet was unable to save the snapshots.