You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Based on Sean’s Charge Trapping App with Am-241, the Charge Trapping correction can be linear based on the boundary conditions found in the app:
Slack message from Sean September 2025: “My intention was to just do a linear correction so we don't have to do any simulations or anything. Since we're anticipating a small effect for the undamaged detectors, I figured the linear approximation is probably sufficient.”
With this, we need to also determine the mechanics of running the data through the first steps of Nuclearizer, implementing the charge trapping correction, refitting in Melinator, and then running the data through Nuclearizer again.
Activity
fhagemann commentedon Aug 18, 2026
After all, it looks like we still ended up going for the simulation-based approach in #184 ^^