Rombur
commented
2 years ago We may want to consider changing the temperature of deposited material to being its own input parameter since strictly speaking it doesn't have to be the same as the initial substrate temperature.
I actually have a branch that does exactly that. I was planning to create the PR later today :smile:
This PR adds the ability to set the standard deviation of the initial temperature and beam 0 max power from the input file for ensemble simulations. Eventually we can add more parameters that can vary between ensemble members, but these two parameters reflect two main types of parameters -- the initial temperature only changes the initial conditions, it doesn't change the evolution equations at all (and therefore wouldn't be in a parameter-enriched ensemble) meanwhile the max power does change the evolution equations (and therefore could be in a parameter-enriched ensemble).
The current approach stores vectors of the ensemble parameters. Strictly speaking this isn't necessary and we could do random draws for each ensemble member separately, but I think having them all in a vector might make the creation of parameter-enriched ensembles easier.
In this PR the temperature of the material that is added during the build is at the nominal (mean) initial temperature rather than the initial temperature for that ensemble member. We may want to consider changing the temperature of deposited material to being its own input parameter since strictly speaking it doesn't have to be the same as the initial substrate temperature.
This PR sets the stage for a future PR that does DA for ensemble members with different initial temperatures. That can be followed by another PR with DA for ensemble members with a parameter-enriched ensemble that includes the max power.