Hi! Collaboration on some Geant4 to spice interaction ?

[leochand101]

Hey Steven!
Is Geant something you are using actively ? I would like to use if for radiation hardened by design modelling. I have some work on bias-dependent modelling of single-events (radiation hits on integrated circuits).
I am not a C/Cpp expert but would like to collaborate perhaps at a python wrapper level (with as much Cpp as needed)
I would like to model IC stacks and then get potential charge generation as a result of radiation interaction on Silicon substrates.
This will produce some scholarship and papers. We would explore the physics of interactions and then modeling for circuit design.
If this is something interesting let me know.

Cheers and best of luck with your research , we also make camera ROICs for radiation and non-rad environments, not sure if your computer vision research involves custom hardware or not.

[stevenwalton]

Hey @leochand101, I don't work in this space anymore so unfortunately I can't give you too much advice. These tutorials were specifically made as I was learning Geant and trying to break down the included tutorials, which are nowhere near sufficient. I just wanted to help others not go through the pain that I did. Spending so many hours reading the docs and source code just to create basic simulations. Even with good C++ chops Geant is an unwieldy beast. I'm glad to see that others have benefited from this and hope that at least in some part I've helped this community that I've sense left. (Same goes for my H5Easy project that is similarly out of date)

I would like to model IC stacks and then get potential charge generation as a result of radiation interaction on Silicon substrates.

Interestingly enough this is not too different from the project I was working on at the time. Though I never published anything and I'm not sure I can offer too much advice, especially given ITAR and NDA restrictions.

At this time I'm not really interested in this type of work. I'm focused on ML research and trying to finish my PhD.

For the python wrapper, I seem to recall finding a project that was doing this (similarly a CUDA implementation, which would now be useful but wasn't back then). But my memory is failing me (this was like 7 years ago...). I also expect that this project is much harder than it appears at face value. So go easy on whoever you end up hiring to do this. They're going to have to do a really deep dive in the Geant code and that's never fun.