sevenperforce

Centaur SEC uses an RL-10 as well, the RL-10A-4-2. At 451 s it's actually got a poorer specific impulse than the RS-25. Its advantage is its extremely low mass -- just 168 kg -- which is just barely over half the specific impulse mass of the RL10C-2-1 on the Delta IV's upper stage. That, combined with Centaur's extraordinarily lightweight tank design, makes Centaur's mass fraction excellent and makes it extremely good for high-energy orbits. But like all other RL-10s it has very low thrust for its nozzle area. Basically, it takes up too much space. TLI is not terribly high energy; you need good thrust for your final push to orbit or you burn too much of your propellant in a lofted trajectory. That's why the J-2X would have made such a good Earth Departure Stage engine despite its lower specific impulse (448 seconds).

Our ability to kinetic-kill missiles during boost is between 40% and 70%, I believe. The chances that we could actually target and impact the engine cluster directly, particularly in a tumbling stage, is virtually nil.

By my math, these engines are going to hit the ground at just over Mach 1 with the equivalent kinetic energy of a warhead with about 40 kg of TNT. A little bit less than the yield of the AGM-88 HARM tactical air-to-surface missile. This is the size of the crater left in loose soil by the Qassam 3 rocket, which has about half that much yield:

Most of the rocket body will be shredded by the heat of re-entry and what reaches the ground will be small pieces of debris falling at fairly low speeds. But the helium COPV(s) for the LOX tank will reach the ground. Not sure how big the Long March 5 COPVs are. And the two engines will reach the ground. Each one is 2,700 kg or just under 6,000 pounds.

It really was. You could see the inside of each Draco thruster firing and see the cool-down after each pulse. Each of those little thrusters produces the equivalent force of a 90 lb weight.

I started watching the stream at EXACTLY the right time to see the undocking happen live.

Stay indoors. Preferably as low as possible. Park cars in garages.

Here's a thought (hear me out)....... Space Shuttle external tank with three SSMEs mounted underneath it and a Centaur V on top. Two Delta CBC strapped onto the sides cross-feeding to two of the three SSMEs. Alternately, move the core tank bulkhead and use two Pyrios boosters crossfeeding LOX only to two of the three SSMEs.

I'm not so worried about cost -- if we're paying for something pricey but the performance is good enough then I'm fine with it -- I'm worried about thrust per area and cooling scheme. How is the EUS even going to handle mounting four RL-10C-3 engines? It has a regeneratively-cooled nozzle AND a fixed radiatively-cooled nozzle extension AND an extensible radiatively-cooled nozzle extension. Like, that's a LOT. If Ares V ran into problems with mounting non-regeneratively-cooled RS-68s too close to each other, how is the EUS going to manage? The highest-performing RL-10 that could be clustered ad nauseum, with regenerative cooling handling everything, would have been the RL-10A-4-2 without the nozzle extension that flew on Centaur D-5 for the Atlas V, at 446.4 seconds and 97.9 kN in a 1-meter bell. But if that's the best you can do, why not just bite the bullet and figure out how to air-start an RS-25? At 2.4 meters, an air-started RS-25 takes up the same area as four clustered RL-10A-4-2 engines but produces 475% as much thrust. It's T/W ratio is almost exactly the same (0.42% higher, actually!!) and it has a 1.32% advantage in specific impulse.

The RL-10 has awe-inspiring specific impulse (up to 470 seconds in some designs) and its T/W ratio is not terrible but its thrust is SO low for its nozzle area that it's pretty useless for a second stage. The Japanese LE-5A and LE-5B are open expander cycle hydrolox and sport 447-452 seconds so I'm not sure why the BE-3U would be as low as 430 seconds. Expander cycle engines generally are not very powerful for their size. The RD-146D (upper-stage engine for Angara) comes in at 470 seconds as well but it only delivers 67 kN with a 1.9-meter bell diameter. I assume it would need a complete turbopump redesign.

The good news is that based on launch inclination we know it will land somewhere above New Zealand but below New York. The bad news is that's most of the people.

Answering my own question...

Could prove useful.

There are also more exotic solutions, like jettisoning SSMEs mid-flight. But seriously that core is.....just so heavy.

If you reuse the Superheavy booster but expend the Starship, its TLI performance comes very close to SLS Block 1. It depends on how stripped-down you can make the Starship. At 60-65 tonnes dry it matches Block 1 almost perfectly. Block 1B of course has significantly higher throw. Not sure whether expending the Superheavy booster would raise dV enough to match. The tanks are the same diameter, but otherwise unrelated, and at vastly higher cost. There is no reason for the core stage (tanks) to cost more than maybe $200M, and the cost is unknown except 3 tanks for $6.7B including dev. Shuttle derived would have been sticking engines on a Shuttle external tank with an interstage added, and some upper stage. Yeah, if SLS was still "Shuttle-derived" after stretching the main tank, un-stretching it wouldn't make it any less Shuttle-derived. The most Shuttle-derived part of SLS is the concept of taking a massive first-stage tank with heavy first-stage engines just to the edge of orbit and then dropping it. Which was also the most inefficient thing about the Shuttle.

But for all its splendidness Starship+Superheavy can't actually match the TLI throw of SLS, not without refueling. Two stages with methalox can't beat 2.5 stages of hydrolox boosted by solids. Obviously refueling is Starship's whole shtick, but still. I'm all about encouraging criticism of Shuttle-derived designs but all I'm saying to keep is the tank diameter and total vehicle height, which is hardly Shuttle-derived at all. I suppose you could shrink the core vertically and stretch the upper stage and see if that helps. That 452 seconds of vacuum isp on the RS-25 is pretty hard to beat. Six or seven BE-4s will definitely beat the hell out of the SRBs. So would a cluster of Raptors in the Big Falcon configuration.

But if you were going to go 2.5 stages...... Assume you could use any engines.

Hating on SLS, as always, but it occurred to me that coming up with something better isn't particularly straightforward. I mused that it would be nice to see those SRBs swapped out for paired F-1Bs (e.g. Pyrios concept), the RS-25s swapped out for a cluster of Raptors, and extra RL-10s added to the EUS. But after running the numbers, I actually ended up with a lower throw weight to TLI. So that was disappointing. If the objective of SLS was to throw something actually useful to TLI and you could swap out any of the engines (keeping in mind that you will need to adjust tank strength to account for increased propellant load), what would you choose? You can use any engines that have actually been built and that you have specs for. You just have to keep the overall stack dimensions the same within a 2.5-stage architecture (otherwise it is no longer an apples-to-apples comparison).

Then it demonstrates that there is something very wrong with their design. Flight safety is predicated on control. If SpaceX can't control the vehicle well enough to keep the flight within the designated envelope then we have a big problem and we will not be going to space the next day.

NGL I always get chills when the signal drops right before the touchdown

OOOOOMF that is an amazing landing

Well that's not sexist at all Yeah that was me.

He is such an a$$

Static fire. no hop yet