Exercise 1:
Linearization and Unionization




This exercise will make use of the RECONR module of NJOY to linearize the hydrogen cross sections to be accurate within a specified tolerance.

NJOY runs data through its various modules by using "tapes" to communicate between the modules (well, they are really files, except to the nostalgic). Therefore, NJOY input files give the module name to use, then the input and output units for that module, and then the characteristic input for that module. This is repeated for each module to be used, and then terminated with the module name "stop". Change to your NJOY working directory, and copy t511 to tape20. Type in the following input file (but leave off the comments to the right of the slash symbol): 

           reconr
           20 21
           'exercise 1'/    new tape ID title
           1301 1/          MAT
           .001/            fractional tolerance
           '1-H-1'/         descriptive card for new tape
           0/
           stop

Save the file as in1.

This file says to run RECONR on MAT1301 with a reconstruction tolerance of .001 (.1%). The output will be on tape21. The title on the Tape ID card will be "exercise 1," and the single information line in the output file will say "1-H-1."

Run the NJOY job with this input file. Under unix or DOS, this can be done as follows: 

           xnjoy < in1

Examine the file called "output" to see the messages from RECONR. Examine the file called "tape21" to see what the new PENDF (pointwise ENDF) tape looks like.
  • Check MF=3, MT=102 (radiative capture). Find the energies corresponding to energies on the original ENDF tape. Are the cross sections the same as the original ones?

  • Look at the energies added by RECONR. Note the standard even values added, and also note where the intervals were divided in half for increased accuracy.

  • What is the analytic shape of this radiative capture cross section at low energies? Where does it begin to deviate from this shape?

  • What would the error have been at 1.5e-4 eV if linear interpolation had been used with the grid of the original ENDF tape?

  • [Optional] Derive a formula for the energy grid spacing that would be required to achieve the fractional accuracy x for a 1/v cross section.

  • Check MF=3, MT=2 (elastic scattering). Does the grid match the one found for MF=3, MT=102? What is the analytic shape of this cross section at low energies?

  • Check MF=3, MT=1 (total cross section). Does the energy grid match those from the other sections? We call this the union grid. Are the cross section values really equal to the sum of the partial cross sections? This is one of the things that unionization makes possible.

NEXT
INDEX

23 January 1998 T-2 Nuclear Information Service ryxm@lanl.gov