Effects of PEATHREShold

I am studying particle production for 30-GeV protons on a nickel target
stack. I have usually prefered to use the PEATHREShold option of the
PHYSICS card. I am now looking at things a bit more carefully.

The attached plot shows some comparisons for the neutron energy
distribution with the default 5-GeV limit and with a 50-GeV limit. In
the upper part, there is a change in slope at about 5 GeV. Is this
realistic? Or is the transition from the default PEANUT model to an
extended version not properly smoothed out?

The default version has a jump near 500 MeV, while the extended version
is smoother (and is one of the reasons I prefer the extended version).
The extended version also seems smoother overall in the lower figure,
which goes up to 100 MeV.

Another issue is a dependence on the upper limit. 50 GeV is well above
any energies in the problem, but I get different results if I change it to
1000 GeV. It appears the random number sequence gets changed very early
on because the numbers of produced particles are very similar. Why should
that be? Is it affected somehow by the scaling of the model parameters?

Finally, I am concerned with changes in the number of particles. With the
extended limit, the number of protons and neutrons increase by about 10%
(mosty at low energies). That is not unreasonable. However, the number
of photons decreases by 1-2% and, more disturbingly, the number of Klongs
decrease by about 20%! Where did they go?

Thank you,
Joe