I keep on seeing accelerator-driven solutions to astronuclear n flux deficiencies suggested... While it's not a bad idea in some cases, there's a few considerations that I'm curious about: A thread.
1: How sensitive is the reactor to n spectrum? Many NTRs run in the epithermal region, which is NOTORIOUSLY complex when dealing with n capture resulting in fission in most fuels. Adding ADR n spectrum complexities seems to be (AFG) just adding to the headaches.
2/
2/
2: accelerator and associated equipment mass in dry mass considerations. While LINACs are getting TINY, getting a high enough energy beam with enough throughput to make a difference in multi-hundred MWt reactors won't happen out of a shoe box - and that's JUST the LINAC!
3/
3/
There's also the support structure, shielding for the accelerator, fueling and powering it. BNTRs are an exception for a lot of reasons, mostly due to complexity, so at some point it may become more effective to just add either an RTG or Kilopower class reactor... Ooff.
4/
4/
3: spallation target. Most accelerator driven n sources use a Be/Hg/etc target to produce the neutrons, with a clear line of sight btwn the beam and the target. While possible, or would require a very different architecture from typical NTR design (of any sort)
5/
5/
This also brings up thermal limits: Be is the major thermal limiter in LEU NTP. All spallation targets I've seen are either solid or liquid. What about gaseous/vapor? Dusty plasma? I have no clue. If anyone has papers about this subject, PLEASE link them.
6/
6/
4: alternatives. As @NukeSpaceCadet has shown, for non-solid fueled reactors using a % of solid fuel can increase n flux and achieve criticality in lower temperature regions. This concept is both more complex to model and simpler to actually run during a mission.
7/
7/
In summary: I'm not against being convinced that ADR astronuclear drive systems are unworkable, but "just use an accelerator" has no excuse for the use of the word "just." It's a complex design choice with lots of complications and basically requires a rewrite of the arch.
8/8
8/8