Benchmark C5G7

User Support
1

Hi all,

I just worked on the Benchmark C5G7 for a multigroup, for calculating the flux and the reaction rates in each energy group, but I don’t have any information concerning the energy bin.

How can have the flux and the reaction rate in each energy group without filter of the energy groups?
Here are the tallies

<?xml version='1.0' encoding='utf-8'?> 85 1 1 -53.55 -0.63 -1e+50 53.55 0.63 1e+50 ????? 1 1 2 flux fission nu-fission kappa-fission flux fission nu-fission

Thanks in advance!

2

Hi Mohamed,

You will eed the energy group filter to get the groupwise results, just as your xml shows. The values to use for these energy bins are those that match the 7-group structure. Since that isn’t specified by the benchmark (as far as I know), this should simply match the group structure used in the model’s multigroup cross section library (the .h5 file). If you have obtained the C5G7 library from here (https://github.com/mit-crpg/benchmarks/tree/master/c5g7/openmc) then the group structure is:

`
1.0E-5 5.179474679231213E-4 0.026826957952797246 1.389495494373136 71.96856730011514 3727.593720314938 193069.77288832457 1.0E7

`

These are 7 equal-width, in log-space, bins from 1E-5 to 1E7 eV created with the following python numpy command:
np.logspace(-5, 7, 8)

If you are not using the library from there, the best way is to view the"group structure" attribute of the multigroup xs library you are using. This can be done on a computer with hdf5 and the hdf5-tools package (named that wa on ubuntu at least) installed (which should be the case for the one running openmc) with the following command where you should replace mgxs.h5 with the particular filename of the library you will be using:

h5dump -N "group structure" mgxs.h5

The output of this will be, for example:

HDF5 "mgxs.h5" { ATTRIBUTE "group structure" { DATATYPE H5T_IEEE_F64LE DATASPACE SIMPLE { ( 8 ) / ( 8 ) } DATA { (0): 1e-05, 0.000517947, 0.026827, 1.3895, 71.9686, 3727.59, 193070, 1e+07 } } }

The information you want is in the DATA { block.

Alternatively, you can use a graphical viewer, such as HDFView (in fact you can get more bits of precision using the graphical viewer).

Hope this helps!
Adam

3

Thanks Adam!

4

Hi all,
In the same C5G7 benchmark, I try to calculate the power in each pin cell from the kappa-fission, but the results are zero in each pin cell.

<?xml version='1.0' encoding='utf-8'?> 51 1 1 -32.13 -0.63 -1e+50 32.13 0.63 1e+50 1e-05 0.000517947 0.026827 1.3895 71.9686 3727.59 193070 1e+07 1 1 2 flux fission nu-fission 2 flux fission nu-fission kappa-fission

Thanks in advance!

5

I would double check your mesh. Do you really want 51 bins in the x-direction and only 1 in the y-direction?

6

Yes, I have defined a one dimensional benchmark problem from 2D C5G7 as shown in the figure below.
And I want to compute the power distribution in 51 bins in the x-direction only.

Geometry.png