For material ordering, the code determines how many materials and Legendre tables can fit into the available storage. It then makes one pass through the library file, produces the cross sections for the first set of materials and orders, and writes them out on the output file. It then rewinds the data file and repeats the entire process for the second set of materials and orders. This process continues until all the desired materials have been produced. The minimum set is one Legendre order and one material; if this does not fit, group ordering can be used.
For group ordering, the code determines how many groups
(NGMAX) can fit in memory at one time. It then makes one
pass through the data file for each set of NGMAX groups,
writing out the group cross sections after each pass.
This paging strategy allows TRANSX to be run on small machines if
necessary, but if a large memory is available, the code will run
much more efficiently. Storage is assigned to a single container
array A with NAMAX words
(NAMAX is 300000
in the Los Alamos CRAY version). The allocation can be changed
easily by changing the two statements DIMENSION A(300000)
and NAMAX=300000.
A similar scheme is used during the calculation of self-shielding
$\sigma_0$ values. Cross section data are read in for groups
JGLO to JGHI, the $\sigma_0$ values are computed
and printed, and the code repeats the process for the next
group range, JGLO to JGHI.
Part of the NAMAX words of storage is reserved for
reading the records of the MATXS cross section file by the
statement INMAX=6000. This should be sufficient for MATXS
files prepared with the default value of MAXW (5000).
TRANSX HyperText Manual
TRANSX HyperText Help Package
T-2 Nuclear Information Service