Random Ray Void Accuracy Fix

[jtramm]

Description

This PR fixes an accuracy issue in the random ray solver stemming from PR #3299 affecting flux tallies in void cells with external source terms. More info is available in the writeup for issue #3315. This issue was originally missed in my testing as this error decayed away as the mesh resolution was refined, so seemed like a legitimate method bias for this use case rather than a logical error. With the new fix, a much coarser mesh is now sufficient to resolve flux tallies in void regions with external source terms.

Fixes #3315

Validation

To test this, I redid the analysis in #3299. This model is a modified 3D Kobayashi dog leg problem, where the fixed source area and duct work have been converted to pure void materials. Below is the flux tally in the void area that contains the external source term, using a 5x5x5 cm^3 mesh:

	Flux	Error vs. MGMC
MGMC	9.92E+00	0.00%
Flat Source (Old)	7.57E+00	-23.73%
Linear Source (Old)	7.68E+00	-22.57%
Flat Source (This PR)	9.80E+00	-1.22%
Linear Source (This PR)	9.91E+00	-0.06%

Formatting change

I also noticed there was an annoying loop index variable name that was causing a discrepancy in formatting between LLVM 15 vs. 19, so changed that variable name to a shorter one to stop triggering the condition that caused the formatting discrepancy.

Checklist

• I have performed a self-review of my own code
• I have run clang-format (version 15) on any C++ source files (if applicable)
• I have followed the style guidelines for Python source files (if applicable)
• I have made corresponding changes to the documentation (if applicable)
• I have added tests that prove my fix is effective or that my feature works (if applicable)

[paulromano]

Thanks @jtramm!