Re: Problems of Y-90 beta source with FLUKA

Hai Guoqing,

I tested with:

- a 1 MeV electron source;

- a 2 keV electron source.
In both cases FLUKA ran 2 to 5 times faster.

For the 1 MeV run, both MCNPX and FLUKA simulated EM showers consisting of
Moller, bremsstrahlung, Compton, photo-electric etc. Note, however, that
for the 2 keV run, MCNPX simulated hardly any interactions (we can see
this in the particle creation tables). Under such circumstances the
simulation would surely finish in a flash.

I see that this post is similar to your earlier one,'Why is FLUKA slower
than MCNPX?' which Alberto had responded to. I therefore take the liberty
to add that Class I versus class II (a term used by the EGS community) or
random-hinge (the term used by the PENELOPE community) is not only about
the implicit/explicit simulation of Mollers, but also bremsstrahlungs. In
MCNPX one can turn on/off Moller and/or bremsstrahlung too; the switches
are in the PHYS:E card.

:) mary

On Sat, 30 Jul 2011, zhangguoqing01_at_tsinghua.org.cn wrote:

>
> Dear FLUKA Users,
>
> I am using FLUKA to make simulations with a Y-90 beta source. The geometry
> is rather simple: one point Y-90 source inside a two-layer shielding,
> which include 1 mm plexiglass and 1 mm lead.
>
> The simulation runs very slow. To speed up the simulation, someone told me
> that I can turn the delta-production off by setting a production cut
> higher than the maximum source energy, but this seems not possible. I
> found in the following link that FLUKA will automatically set the
> transport threshold equal to the production one.
> http://www.fluka.org/web_archive/earchive/new-fluka-discuss/3678.html
>
> It's written in the FLUKA manual (pp. 8) that the lowest energy of primary
> electrons is about several tens of keV. When I use the multiple scattering
> transport, the electron current on the outer surface of the shielding
> looks strange. In the first energy bin, the current is abnormally high
> (see the attachment). If single scattering transport is used, the peak at
> the first bin disappears. The problem is the single scattering transport
> is extremely slow. Can somebody tell me how to get rid of the strange peak
> and make the simulation faster?
>
>
>
>
> Thanks a lot!
>
> Vielen Dank!
>
>
Received on Thu Aug 18 2011 - 18:55:05 CEST