Re: piont-wise cross section

From: Mary Chin <>
Date: Tue, 21 Feb 2012 12:49:17 +0100

Hai Weihua,

Thanks for getting back to fluka-discuss about the unresolved issue.
Indeed, the developers hope to implement pointwise cross sections on
a wider scale soonB.

I confirm

- there are monoenergetic peaks, one each for C, O, N, respectively. The
peak for N is relatively low.

- they do not correspond to evaluated cross sections. In
fact the energy of the peaks changes if the source energy is changed.

- they are not due to events upstream (eg interactions in other regions
before entering the 'cavity' region).

- increasing the radius or the density would populate higher energy bins.

In fact the peaks are due to the kerma released from the recoil following
one single collision of a low-energy neutron in the 'cavity' region; the
collision probability is very low due to the low density and there is no
chance for a second collision. This is akin to single scattering
experiments. Experimentally the peaks would have been spread out by the
angular distributions. This, however, does not happen in FLUKA because
there is only one kerma value per group in the multigroup approach.

If we replace all other materials with vacuum we get clean lone peaks on
the pulse height spectrum. If we increase the radius or density we give
second interactions the chance to occur.

Don't worry about the junk lines -- the priority is to resolve the issue,
then add that by the way such and such lines should not be there.
Moreover, a FLUKA expert would pick out the problem no matter how messy an
input file is, as much as a nix/nux user would remove those lines by a
single command line.

The following are non-issues and beside the point but I take the
opportunity to clarify anyway the earlier posts to the same question:

> ENDF/B-VI is just a generic name for several versions

*No. It refers specifically to the LIBRARY released in 1990. Subsequent
INTERMEDIATE RELEASES containing minor updates and corrections to the
general release are thereafter appeneded with a release number.

> Versions older than 8 are non available anymore on NNDC and the IAEA
> websites, so I cannot check if there are differences

*The people who release evaluated data do not play hide-and-seek,
expecting us to plough through hundreds of materials to see which changed.
We do not beed to plot or see the ENDF 'to check if there are
differences'. Data for many materials are inherited from previous
evaluations or cross-inherited from sister libraries, those that have been
changed are fully tabulated and announced at each release.

> The total cross section shows large peaks due to inelastic reactions
> above 0.5 MeV, for both carbon and oxygen.

*There is no such peak for carbon. There is one for oxygen at (0.432610
MeV. 0.83524 mb). It is capture / absorption, not inelastic scatter.

:) mary

On Mon, 20 Feb 2012, Xu Lijun wrote:

> Dear Alberto,
> Thanks for your last return. I have tried the large radius cavity according to your advice. And there is not
> unreasonable peak in the pulse height spectrum.But when the cavity radius shrink to several cm, the unreasonable
> peaks will appear,even there is no wall. why this happen?
> I hope the point wise cross section used for the low energy neutron transportation. As you said, the point-wise
> cross section are not available for any of the materials in my input file. I didnot know how to set the card.Could
> you help me? I attached my input file in which I deleted the unnecessary cards.
> Best regards,
> Zhang Weihua
> 2012-02-20
> ___________________________________________________________________________________________________________________
> Xu Lijun
Received on Tue Feb 21 2012 - 14:11:53 CET

This archive was generated by hypermail 2.2.0 : Tue Feb 21 2012 - 14:11:53 CET