Message: High computational cost with an Electric Field Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Question High computational cost with an Electric Field 

Forum: Fields: Magnetic and Otherwise
Date: 09 Mar, 2018
From: Matteo Bernasconi <Matteo Bernasconi>

Hello,

I need to simulate the propagation of particles inside an uniform electric field, which is present only in a specific volume. I have simply a 10 x 10 x 6 cm box with a local electric field and I want to know the position, momentum and energy of primary and secondary particles when they exit from the volume.

Now the problem: if I run the simulation without electric field, even with millions of primary particles, the maximum time is in the order of minutes. If the electric field is present, it requires days with thousand of particles. It seems that no error is present, it is simply very slow. I thought my simulation was too costly because I dump some variables acting on UserSteppingAction (see attached file), but if I add a magnetic field instead of an electric one, the simulation finishes after half an hour with hundreds of thousands of particles (therefore, much faster). I tried also to increase the minimum step for the integration up to mm (so an order of magnitude lower than the geometry), but nothing changed.

My question, therefore, is: Geant4 is not suitable for dealing with electric field and large amount of particles or am I doing something wrong?

I use geant4.10.01 version.

Best regards, Matteo

 Add Message Add Message
to: "High computational cost with an Electric Field"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews


[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]