Recently we have come across some, rather mind-boggling, difficulties concerning the noted Monte-Carlo code. First and foremost, it is of significant importance to us to know what changes are needed to be implemented in the installation, or execution procedures of the code to enable multi-threaded use of all the CPU processing cores. The defined commands are rather complex and demand optimum usage of all available hardware. Secondly, It is extremely vital for us to make sure that the GEANT4 code understands the cross-sections of heavy charged particle interactions with the nanoscale materials apparent in our geometry. Thus, we wanted to know if such directory exists, by default, and if not, wanted guidance on the data, and/or changes which are needed to input the required information. This is an absolute necessity for the accuracy of our results. please inform me of your thoughts and ideas.
Many thanks in advance for all your efforts.
Concerning the use of multi processors, what we normally do is to divide the task into independent runs. For instance, you are simulating an irradiation of a sample by 4000 particles and you have 4 processors, so you run the main program 4x with only 1000 particles.
This has the added benefit that you can extend the method to separated machines, like in the case of cloud computing or even in a dedicated farm.
Regarding the interaction of heavy charged particles with nanoscale materials, Geant4 does not consider the molecular structure of the medium. It only considers the relative proportion of atoms and isotopes, and properly accounts for the interaction cross-sections with individual atoms and their electrons and nucleons. It is up to the user to properly define the materials and elements present in your geometry description.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Particle Physics