Hi

Hi,

I would like to simulate compton current. This means I would like to score things like how many electrons that were born in region 4 died in region 11.

1) Is it possible to do this with egs_track_scoring? I read the code and it seems to me, that it may splits tracks, that in fact belong together. For visualization, this makes of cource no difference. My understanding is limited however, can someone comment on this?

2) Is there another tool/code that provides a good starting point?

Hello EGSnrc community

Hello EGSnrc community,

I recently tried to use dosxyznrc in conjuction with a beam simulation on a parellel run. However the bandwidth demand of this setup skyrocketed so high that effectively one node was grabbing all the bandwidth alone (making the idea to run it on a cluster pretty useless). I did not experience this behaviour when using a beam simulation coupled with egs_chamber. My questions to you as a community now are: Is this behaviour normal, and are there any good strategies you found to deal with it?

Thank you in advance and Best Regards,

Markus

[CBCT] Few days ago I asked about the bow tie filters for Varian Trubeam, but I got no response (just few likes :D).

[CBCT] Few days ago I asked about the bow tie filters for Varian Trubeam, but I got no response (just few likes :D).

So I simulated the filter with XYZGeometry (see the comment below) and I used gunion to put them together. But I have few problems with this arrangement:

1. When I put only celyfantom (phantom) and bowtie to gunion, the scoring seems to happen only in the bowtie. Is there some condition that the objects in gunion must overlap? Or what could be the problem? My solution for now – I put there box filled with air which goes through both objects.

2. Is there some better way how to connect the source and bowtie filter, so I could use the CBCT setup for rotation of the source/detector?

For egs++ applications, there is a ‘simple run’ option for the run control object which, as far as I understand,…

For egs++ applications, there is a ‘simple run’ option for the run control object which, as far as I understand, does not rely on a lock file for parallel runs. Every job just blindly runs the number of histories in the input file.

1) Is there a similar option for Mortran applications?

2) Wouldn’t it make more sense if each job ran ncase / nparallel histories? It’s the expected behaviour from the typical ‘non-simple’ parallel runs.