Improve algorithm for criticality search

[paulromano]

Our search_for_keff Python function currently uses optimization functions from SciPy that do not take into account the uncertainty in keff estimates. A few papers in the literature have proposed methods for doing a criticality search while accounting for an uncertain estimate in keff. A few references:

• A keff search capability in MC21
• Movable geometry and eigenvalue search capability in the MC21 Monte Carlo code

There are also a few more sophisticated methods, but these may not be as generalizable:

• Perturbation based Monte Carlo criticality search in density, enrichment and concentration
• Development of the universe based geometry criticality search capability in RMC and analysis of NHR200-II reactor

There are likely other methods from the larger field of stochastic approximation that may be relevant.

[paulromano]

To make sure it doesn't get lost in the noise, @gridley made a good suggestion here #2693 (comment) based on a more recent paper:

• Method for control drum position critical search with Monte Carlo codes