Add support for non-fixed neutron source strength

[helen-brooks]

• Currently the neutron source strength is assumed to be a fixed value, which is reasonable for fusion but not fission problems.
• In a k-eigenvalue problem this is not the case and must be deduced at runtime, see more here: https://docs.openmc.org/en/stable/usersguide/tallies.html#normalization-of-tally-results
• We need to implement support for this.