sevenperforce

How many rocket scientists does it take to change a nuclear lightbulb?

Easy way to contain a nuclear reaction? Put it under water. https://what-if.xkcd.com/15/

The point of the Orion pusher plate design is to find a way to unlock the extremely power density and energy density of nuclear chain reactions for propulsion. Because the energies are too high to be nominally contained, a "pulsed" approach is the only way to do it. The smallest unboosted nuclear fission warheads have a minimum yield of around 10-20 tonne TNT equivalent and the largest have a maximum yield of about 500 kiloton TNT equivalent. The smallest thermonuclear warheads have a minimum yield of around 5-10 kiloton TNT equivalent and the largest have no theoretical maximum. The smallest possible antimatter bomb is...just one positron. Antimatter will blow up at ANY concentration or quantity if exposed to mass. For that reason, you don't need a pusher-plate approach at all. You can just build a big engine and pump exactly as little antimatter into it as you need in order to run it. Throttle it, vector it, whatever you want to do.

Obligatory:

Liftoff. Sporty!

And it's live!

It would still work.

Successful solar array deploy...and that's a wrap, folks!

You can see the airlock inside Dragon's trunk as it separates.

Successful SECO.

Welcome back booster!!

Interesting that S1 didn't have enough dV for an RTLS but it still did a partial boostback.

More thrust, lower isp.

The atmosphere has to go somewhere. As long as the superheated atmosphere is expanding faster than the airspeed of the spacecraft, a very significant amount of the superheated expanding atmosphere is going to impact the pusher plate. That will provide impulse.

Atmosphere which flows between the pusher plate and the tungsten jet is superheated in the explosion and increases the reaction mass.

Orion is more efficient in atmo.

Yes, that was quite amazing. They were able to use differential thrust to produce some degree of control. Fortunately the control surfaces were trimmed for roughly level flight. That's kind of like if all four of Starship's control surfaces became locked in a stable config and they had to use hot-gas thrusters for roll and pitch. Of course, it may have propellant reserve problems. It can land with extra reserves when carrying humans. Starship will have completely, truly independent drive systems for each of the four control surfaces, so there is no such single failure mode. Helicopters are much closer to Starship for this particular analogy. For a jetliner, the engines are independent of the glide mode; for a helicopter, the engine is the glide mode. If a helicopter lost hydraulic control of its main rotor it would be unable to autorotate.

Yeah, the Convert-O-Tron element is pure hokey. That pretty nearly violates conservation of energy. Antimatter is an energy storage mechanism, not an energy generation mechanism.

High 4-digit at worst. You can use regenerative cooling to keep the throat and nozzle solid. Hell, with antimatter-level energy density, you can afford to use film cooling and still hit high four-digit Isp. That's where radiators come into play.

Specific impulse is the amount of thrust you get out of each kilogram of propellant. Propellant mass being equal, a rocket with 100kN of thrust and a specific impulse of 450 seconds will burn for longer (and therefore deliver greater total impulse) than a rocket with 100kN of thrust and a specific impulse of 320 seconds. With the five-digit specific impulse of an antimatter rocket, you can pretty much burn continuously for as long as you want because you are using so little antimatter each second.

Then inject cool liquid hydrogen. You have a five-digit specific impulse in vacuum; you can carry anything you want. In the atmosphere you probably have a seven-digit specific impulse. This is a torchship.

ACES is the five-meter-wide Centaur, stretched, four RL-10s, IVF. IVF is more than just the ICE; it also includes autogenous pressurization (which we take for granted because Starship has it, but has not yet been done with Centaur or the DCSS) and autogenous ullage thrusters (Centaur currently uses continuously-firing hydrazine).

Incorrect. Antimatter is not thermally limited. If you have trouble with heat, you can use as much atmospheric cooling as you want. By the time you're out of the atmosphere, thrust is not as important, and so you again have no problems. If you have good antimatter containment then just build SABRE but with antimatter.

Unpowered landing is not a huge problem for airplanes. It's tough for helicopters, but doable. With airplanes it is fairly straightforward. Comparing Starship's engines to airplane engines, then, is a bit of a tricky proposition because their role is different. Airplanes can land unpowered but they cannot land without control surfaces. Multiple engine failure on Starship would be less like dual-engine-out on a jetliner and more like locked elevators and ailerons on a jetliner. If a jumbo jet loses its ailerons and elevators, it is dead. In contrast, as long as 1 of Starship's 3 SL engines is working and at least 2 of its 4 control surfaces are working, it can land. I remember watching this landing live and thinking about fault tolerance on landing gear.... If a systematic problem prevents all control surfaces on a jetliner from working, it would crash. This is why control surfaces have triple+ redundant subsystems. The engines on Starship all have independent turbopumps, plumbing, ignition systems, and so forth for the same reason. If the tanks are breached and there is no propellant, then yes, Starship will crash. But that seems obvious. If an airliner's wings break in half, it will crash too.

I believe CRS-19 will comanifest the privately-built Bishop Airlock Module in Dragon's trunk for permanent installation on the ISS. Gunther puts it at 325 kg which doesn't seem like much, but maybe it's enough that they need extra performance.