Dear Victor,
I am not sure to understand the goal of your study.
I just have a couple of comments.
The time for initialisation is not related to the number of primaries.
The average cpu time per primary is instead dependent on the number of
primaries. When you increase the number of primaries, this value should
converge, because you become less and less sensitive to
statistical fluctuations.
In order to have a fair comparison of cpu time you should make sure to
use the same machine in the same conditions (if other processes are
running and using your CPU at the same time as your FLUKA calculations,
then your estimate would be biased).
Best regards
Ilaria
On 18 Sep 2019, at 01:50, VICTOR HUGO ALTAMIRANO CARPIO <victor.altamirano_at_epn.edu.ec<mailto:victor.altamirano_at_epn.edu.ec>> wrote:
Dear Fluka experts
Q1. I measured the time used for initialization and I plotted these data vs. the number of primaries particles. I read the paper "FLUKA Realistic Modeling of Radiation Induced Damage", which stipulate that the time for initialization depend of the time for calculate through a multiple integral the number of defect.
Please, help me to understand. Why does the less time correspond to 2x10^6 primaries particles?
Q2. When I tried to plot the average time used to follow a primary particle vs. the number of primaries particles. I did not have the same (similar) values I found strange values, though. Please, which is the reason?
PD. I defined the total number of particles as the number of primaries particles * number of cycles.
Best regards;
Victor Altamirano C.
victor.altamirano_at_epn.edu.ec<mailto:victor.altamirano_at_epn.edu.ec>
EPN Physics Department
593-982436747
Our life dreams the Utopia. Our death achieves the Ideal. V.H.
<Análsis del tiempo de inicialización de cada ciclo.png><Análsis del tiempo de cada partícula.png>
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Wed Sep 18 2019 - 23:59:14 CEST