Anyone here know anything about Orion- or Casaba-Howitzer-style shaped charge design and how it scales (or doesn't) to the multi-megaton range?
I'm trying to figure out if it's actually possible to get to Proxima Centauri in a human lifetime using the Orion concept. The yield-to-weight ratio of the pulse units needs to be very high - at or above the Taylor limit - to have a reasonable shot at it, and that's charged particles only; neutron and photon emissions don't count because I'm trying to use a Mag Medusa and it can't deflect those. (I don't expect an ablative pusher to work well at the exhaust velocities necessary for this application, but maybe I'm wrong... Medusa is lighter anyway, which is important for mass ratio.)
I was thinking about an advanced Ripple-type device using a solid-core lithium-6 deuteride secondary with a small D-T spark plug. I don't actually know if the Ripple concept can be made to work with a mostly solid-fuel secondary (I suspect there might be neutron budget issues due to a mismatch between the thermonuclear detonation wave width and the D-T neutron mean free path), but for the sake of argument I've assumed it can. The advantages of solid fuel are lower potential neutron losses (the deuterium-lithium-6 cycle doesn't produce net neutrons), higher effective propellant density, and amelioration of the problem of cryogenic storage of deuterium and especially tritium over several decades of flight time (with a magsail for braking you don't need a lot of delta-V at the destination, but you do need some). This type of physics package should result in a very high yield-to-weight ratio - in 1960, Edward Teller and Harold Brown predicted 50 Mt from a 6000 lb device by 1965 (assuming adequate testing, which didn't happen), and deuterium-lithium-6 is nearly as energetic per unit mass as D-T.
The problem is that the bomb gets very hot when it goes off, and apparently most of the energy (80-90%?) leaves as X-rays before it can expand much. The idea of a nuclear shaped charge (a specialized channel filler like beryllium oxide behind a propellant plate in a rocket nozzle-shaped extension of the radiation case) seems to work fine for a small bomb, but can this type of approach possibly mitigate thermal radiation losses efficiently enough to stay in the Isp regime needed for starflight? Is there anything that can? Or is a large bomb with an inert radiation case already good enough? I'd try to simulate it but I need that part of my brain for my job...
I'm beginning to suspect that basically nothing can materially reduce the X-ray losses. Thermal radiation is so much faster than expansion that it basically has to shed that much energy before the timescales get similar.