Hello,

When using the NeutronHP model, the inelastic cross sections (CS) for the
inelastic interaction of  neutron on He is much too high at low energy.
The reason of this bug  is that  no  He4 inelastic data exist in the G4NDL
database. Because of this  only the channels for inelastic scattering of neutron
on He3 are used for this specfic physics.
The reaction n(He3,H)p having an important CS at low energy, this CS
is the main part of the total inelastic CS.
This should not be the case because He3 is only 0.00014 % abundant in the
nature. While He4 has a 99.999 % abundance.

As far as I understand the NeutronHPModel, when computing  the CS for a specific
element, looks at all isotopes of this element for which data exist in the G4NDL
database and sum all  the CS from these data by weighting the contribution of
each isotope according to their natural abundance.  The problem is that if the
inelastic data are missing or no channel exist for the most abundant isotope
then  the CS  of the second most abundant isotope predominates.
Once again He3 is only 0.00014 % abundant in the nature, while He4 has a 99.999
% abundance. But as no He4 data exist He3 becomes artificially very abundant.


To solve this problem I did copy all the He3 inelastic data file in  He4
equivalent inelastic data files, and put in the He4 inelastic CS data the CS
value to nearly 0.
This remove the bug.

This problem occur probably also for Hydrogen, while no inelastic CS data exist
for H1  in the G4NDL database. But in this case as the inelastic scattering CS
on H2 and H3 is low it has not an important impact.

In general the NeutronHP way of computimg the inelastic CS should be modified to
avoid this kind of bug


Best regards


Laurent Desorgher