individual scattering XS arrays

[vedantkm]

Dear Thomas -

Is there a quick way to create scattering XS arrays separately?

E.g.,
mat = NC.load(NCMAT)
bragg_array = mat.scatter.bragg.xsect(E)
incoh_elas_array = mat.scatter.inelas.xsect(E)

Thank you!

[tkittel]

Sorry for taking so long to answer, the question arrived during my vacation and I missed it! Anyway, the answer is to use the comp cfg-string parameter to select out the component in question (read more about it on https://github.com/mctools/ncrystal/wiki/CfgRefDoc). Here is a complete script showing what I mean. It also shows how to use the detect_scattering_components function from the NCrystal.misc module to more generically detect the available components:

#!/usr/bin/env python3

import NCrystal as NC
import numpy as np

#Energy array (default NCrystal unit is eV):
energies = np.geomspace( 1e-7, 10.0, 10000 )

#cfg-string to be investigated:
mycfg = 'Al_sg225.ncmat;temp=200K'

##############################################################################
#Approach #1, if you already know which component you want, you can use the
#comp=xxx cfg-string parameter to directly select it
#(cf. https://github.com/mctools/ncrystal/wiki/CfgRefDoc for more info about the
#"comp" parameter):

mat_bragg = NC.load(f'{mycfg};comp=bragg')

#Getting the cross sections is simple (note: if dealing with an oriented
#single-crystal material you would have to also specify the neutron direction):

xs_bragg = mat_bragg.scatter.xsect( ekin = energies )

#xs_bragg is now a numpy array of cross sections of the corresponding neutron
#energies in the "energies" array.

##############################################################################
#Approach #2, if you do not already know which component you want, since you do
#not know what is available in the material, you can figure this out with the
#NCrystal.misc.detect_scattering_components(..) function. Internally, it simply
#tries all possible components, and then asks the resulting scattering process
#if it is a null (vanishing) process or not:

from NCrystal.misc import detect_scattering_components

comp2xs = {}
for compname in detect_scattering_components( mycfg ):
    mat_comp = NC.load(f'{mycfg};comp={compname}')
    comp2xs[compname] = mat_comp.scatter.xsect( ekin = energies )

#Voila, "comp2xs" is now a dictionary with component names as keys, and the
#corresponding cross section arrays as values.

import pprint
pprint.pprint(comp2xs)

#We can of course try to plot these curves:
import matplotlib.pyplot as plt
for compname, xs in sorted(comp2xs.items()):
    plt.plot( energies, xs, label = compname )

#We also add a "total" curve before showing the plot:
plt.plot( energies,
          NC.load(mycfg).scatter.xsect(ekin = energies),
          label='Total' )
plt.title(mycfg)
plt.loglog()
plt.grid()
plt.legend()
plt.show()

Example output:

{'coh_elas': array([0.        , 0.        , 0.        , ..., 0.01081875, 0.01077895,
       0.01073931]),
 'incoh_elas': array([8.19999441e-03, 8.19999438e-03, 8.19999436e-03, ...,
       6.05355297e-05, 6.03128745e-05, 6.00910381e-05]),
 'inelas': array([9.10845581, 9.09168951, 9.07495407, ..., 1.38554395, 1.38558486,
       1.38562563])}