search

skaersoe / Reactive

Reactive Reactor R&D
http://reactive.dk
1 stars 0 forks source link

pstudy #7

Open trsch opened 10 years ago

trsch commented 10 years ago

We need to implement some pstudy script into the deck we develop. This should map keff as a function of hexagonial height and cylinder radius. We must be able to run such a parameter mapping on any of the isotopes of interest, moderators of interest, and i believe we should run it twice for each of these, one with fuel running through pipes in the moderator, and one where moderator pipes are pushed into the fluide (i.e. swap the materials in cell 100 and 200)

EsbenKlinkby commented 10 years ago

The first version of the pstudy script is completed: deck_pstudy_v1 In the present version, the hexagonial height and cylinder radius are varied in broad ranges. -sub2.sh runs pstudy and submits the jobs to the ESS cluster -plots.C_root contains keff() which is meant as the macro to retrieve keff, radius and height from output files using KcodeReader from TSCKcodeTools.h. Present version only reads keff, but has the functionality to access files from different directories (in the style that pstudy outputs)

trsch commented 10 years ago

@Esben is it possible to get a five line guide in how to run the pstudy?

EsbenKlinkby commented 10 years ago

To run pstudy, the command you need is in the first line of subs2.sh Of course you need the pstudy script, which I'll commit as soon as I get a chance (at present I am in a car in austria, hopefully i get internet i 2-3 hours)

EsbenKlinkby commented 10 years ago

Sorry for the delay . I just pushed pstudy.pl to the repository

trsch commented 10 years ago

Esben: "uniform" means uniform random distribution form a to b - not uniformly increasing between a and b. This needs to be fixed - i love the "nbins low high" notation, but "uniform" but uniform does not do the job...

EsbenKlinkby commented 10 years ago

Fixed in _v3 In fact, after working on this with Troels today, I think all the main features of the pstudy script/deck are ready for a test run. So I submitted 3040 jobs varying: size of fuel element, size of moderator, fissile and furtile isotopes. I've also run prepared input files that redo the above for different salts (12160 jobs :-) ). When/if Troels confirms that the 3040 finished jobs contains all what is needed, I'll submit (better do this at night) Next step is to repeat for various moderators. The input file is prepared for this as well, so it is merely a matter of pushing the button (and hide so that the sysadmin cannot find you)