Re: [fluka-discuss]: Collision tape in two-step simulation

From: Schuetz, Anne <anne.schuetz_at_desy.de>
Date: Thu, 12 Apr 2018 19:07:34 +0200 (CEST)

Dear Luigi,

Thanks so much for your reply.

If I understood correctly, then I can get the total REAL number of neutrons leaving the extraction line of the second step simulation by
scaling the total number of neutrons in the collision tape of the second step with:

REAL # neutrons = # neutrons * (Sum of weights in the collision tape of the first step) / (Sum of weights of the source particles of the first step)

Is that correct?
I am confused by this because it would be totally independent of how many cycles and with how many primaries I was running the second step.


Also, would W3 in your email be
the sum of weights of recorded particles in the collision tape of the second step?

Scaling the number of particles in the Geant4 simulation with W3/W2, where W2 is the sum of weights of recorded particles from the first step (i.e. in the collision tape),
what would this correspond to?
I can kind of understand scaling it to W3/W1 to get the number per original primary particle.
But in both of these scalings, it is ignored how many cycles and primary particles I was running the second step with, isn't it?

Sorry, but I am very confused by this.
Thanks for your help.
Best,
Anne



From: "Luigi Salvatore Esposito" <luigi.salvatore.esposito_at_cern.ch>
To: "Anne Schuetz" <anne.schuetz_at_desy.de>
Cc: "fluka-discuss" <fluka-discuss_at_fluka.org>
Sent: Thursday, 12 April, 2018 15:19:35
Subject: Re: [fluka-discuss]: Collision tape in two-step simulation

Dear Anne,
for the second step simulation, you have to scale the result with the following factor W2/W1, where
W1 = Sum of WEIGHTS of source particle and
W2 = Sum of WEIGHTS of recorder particles from the first step.

For the third step — the Geant simulation— an analogue scaling factor can be computed, W3/W2.

Therefore, for the Geant simulation, the final scaling factor would be W3/W1 to have quantities expressed per primary interaction.

Please bear in mind that you have to consider the WEIGHT of neutrons and not their number.
They would coincide only in a fully not biased simulation, which I guess it is not you case.

An example of the method is given in the advanced course in the lecture of sources
[ https://indico.cern.ch/event/489973/contributions/2148893/attachments/1271819/1884943/11_AdvancedSources2016.pdf | https://indico.cern.ch/event/489973/contributions/2148893/attachments/1271819/1884943/11_AdvancedSources2016.pdf ]
from slide 43.

Hope this could help.
Best, luigi





On 11 Apr 2018, at 17:30, Schuetz, Anne < [ mailto:anne.schuetz_at_desy.de | anne.schuetz_at_desy.de ] > wrote:

Dear experts,

I am simulating the backscattering neutrons from the ILC beam dump in a two-step FLUKA run.
The final goal is to simulate the neutrons that travel back through the extraction line towards the interaction region.

So far, I have successfully scored the neutrons, which are escaping the beam dump in backward direction and crossing a scoring plane, in a collision tape using the mgdraw routine.
With the source routine, these neutrons then served as the primary particles of a second FLUKA simulation of the extraction line.
In the end of this extraction line, I have then another scoring plane to record the neutrons leaving the extraction line in another collision tape.
This final collision tape shall serve as input to a Geant4 simulation of the neutrons in the detector.

My confusion is now about the weights.
How can I find the absolute number of final neutrons and their distributions, when scaling it up to a full ILC beam?

I have read in an old fluka-discuss thread that the results of the second step shall be treated like this:
1. multiply the results by the total weight of the second step
2. divide the results by the total weight of the first step
And then I guess I could normalize this to the ILC bunch population.

Is this correct?

First, I need to find out how many neutrons will find their way through the extraction line from one primary ILC electron bunch being dumped.
I have been running all steps in multi-thread runs, so I get separate collision tapes as final results in the second step.
Will I get the total REAL number of neutrons leaving the extraction line by adding up all neutrons from all collision tapes, and then by applying the steps above on the total number of particles?

Thank you very much for your help.
Best regards,
Anne

-- 
Anne Schütz 
Deutsches Elektronen Synchrotron (DESY) 
FLA group, 1c/01.348, 
Notkestr. 85 
22607 Hamburg 
Germany 
[ mailto:anne.schuetz_at_desy.de | email: anne.schuetz_at_desy.de ] 
phone: +49 40-8998-1465 
__________________________________________________________________________ 
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id�c_info 
__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Thu Apr 12 2018 - 20:36:58 CEST

This archive was generated by hypermail 2.3.0 : Thu Apr 12 2018 - 20:37:10 CEST