Problem with /process/had commands in MT mode

Physics Processes, Models and Cross Sections
1

Hello Geant4 experts,
I am a Geant4 beginner, and I run simulations with some _HP or _AllHP models of Reference Physics Lists now.

I have a question about the UI command: /process/had in the multi-threading (MT) case.

I found that the commands, such as process/had/rdm/thresholdForVeryLongDecayTime 1.0e+60 year, /process/had/particle_hp/skip_missing_isotopes true, /process/had/particle_hp/do_not_adjust_final_state true
are conducted in the serial mode, but NOT applied in the MT mode.

In the MT case, these are always set to the default ones, i.e. threshold time=1 yr, skip missing isotope false, don’t adjust final state false, respectively.

How can I set these parameters as I want in MT case? The following information is about the version I am using now.

If you need additional information to answer my question, please tell me. Thank you.

_Geant4 Version:_11.2.0
_Operating System:_MacOS Ventura 13.6.4
_Compiler/Version:_Apple clang version 14.0.3
_CMake Version:_3.28.1

2

Hi,

I compiled the example rdecay01, using the RunManagerType MT and Serial, and the application of the thresholdForVeryLongDecayTime UI command seems to work (I can see the same gamma lines for both).

What modifications should I apply to rdecay01 example to reproduce the error you mention?

I have to recognize that I see a 0.1% difference in the mean energy of gamma rays of U238 decay when MT built is used, independently of the RunManager type (Serial vs MT). To be investigated with other radioactive isotopes and software stacks (I was using the Geant4 stack provided via CVMFS).

Best,
Alvaro.

3

Hello,

process/had UI commands are working only in PreInit state. After run is initialized these commands are not effective. These commands change nothing in a working thread but only in the master. Worker classes read this information.

VI