16 thoughts on “SN9”

  1. Looks to me like they’re at least largely repeating the flight profile of SN8. They’ve also added helium pressurization to the methane header tank to avoid a repeat of SN8’s issues.

    My guess is they’ll land this one. If so, I expect a more challenging flight profile (going a lot higher) shortly.

    I also think they are a lot closer to having a superheavy booster ready than most people realize; it’s been under construction for a while. We might well see an orbital launch in spring. I think it’s going to be a very interesting year at Boca Chica.

    Meanwhile, SLS is advancing too. At the rate they’re going, they’ll be able to complete a wet dress rehearsal by late summer, and the Green Run by this time next year, maybe. Then, there’s the software issue, as it seems to have a lot of the issues Boeing Starliner had, so that may well add a delay commensurate with Starliner’s (and, um, when, exactly, is the repeat of Starliner’s OFT going to happen?). I’m still thinking that the odds favor Starship seeing orbit well before SLS. I also think SLS will be slowly, quietly, eased out of Artemis, assuming Artemis survives, in the same way SLS is being eased out of the Europa mission.

  2. Elon knows they have to move away from helium, in part because they’re not going to find any on the Moon or Mars. Pressurizing the full CH4 header tank and the downcomer to 70 psi at 90 K, till empty, would take about 205 standard cubic meters of helium (36.7 kg), which would cost about $1,000.

    Switching to compressed hydrogen would make sense for ISRU, since if they can make CH4, they can make a bit of hydrogen by steam reforming.

    Another possibility is to add some large screw impellers to the system to keep the pressure up at the infeed to the pumps, either in the tank or as part of the pump infeed system.

    1. Upon further thought, the CH4 turbine exhaust heading over to the Raptor’s combustion chamber is at 774 Kelvin. 773 Kelvin, or 500 C, is a temperature where some of the new, more efficient LTSR (low temperature steam reforming) catalysts work well for steam reforming methane into hydrogen gas. This is currently a very active research topic for more efficient hydrogen production.

      So the methane is at the right temperature, and at about ten times the required pressure, for steam reforming into hydrogen. All that’s needed is a water feed system so the engine can be used to flash it into steam, running at 40 to 400 psi. The hot steam combines with the methane to form CO and 3H2.

      The mix of CO and H2 could be fed back up to the header tanks. CO freezes at 68 K, so it’s not going to cause icing issues if it ends up mixed with the CH4 in the tank, and it will end up running back through the turbopumps and getting burned in the combustion chamber.

      The tanks would still needs a small COPV tank to initially pressurize them when their nearly full, but once the engines are fired up, there’s a possibility of just diverting a bit of the methane for conversion to hydrogen gas to use as a pressurant for the tank as the fuel volume drops.

      I’m almost tempted to see if I can find some data on reaction rates, catalyst sizes, and other details to see if it has a prayer of working in practice. Of course the LOX tank pressurant is still a problem, but if the hydrogen from the CH4 reforming can be fused into helium, which is inert… ^_^

      1. I Am Not An Engineer but would it be possible to compress gaseous CH4 and O2 from the main tanks, store it in COPVs and return it for the landing phase?

  3. Isn’t there a risk that H2 would dissolve in either O2 or CH4, or that it would form an inflammable gaseous mixture in the O2 tank?

    1. Oh, I wouldn’t let the liberated H2 anywhere near the oxygen tank! Rocketry is exciting enough as it is!

      @FC, I think that’s what they were doing with the CH4 tank on SN8, using autogenous pressurization, similar to an LP gas tank, but with external heating of some of the CH4 so that the gas is warm. But of course it’s not going to stay warm over time, because the tank and the fuel are very cold. SN8 perhaps encountered that problem during the long descent with the engines off, as some of the hot methane from the engine is tapped off for tank pressurization. With the engines off, the gas temperature in the tank would’ve been dropping and dropping.

      My idea is just to turn some of that tapped-off methane into hydrogen, taking advantage of the high temperatures in the CH4 turbine exhaust duct. The hydrogen isn’t going to condense out at liquid methane temperatures. You could use a separate hydrogen tank, but at that point you might as well stick with helium, ignoring the ISRU issues with the lack of concentrated, indigenous helium sources in the inner solar system.

      But the big question is how much catalyst and equipment would it really take to make it work at the required flow rates, and is that going to end up adding significant complexity to the system, along with potential reliability problems? I don’t know, because I only thought of it this morning.

  4. The take on SN8 was that it was a great success — lots of data gathered, and only the landing itself failed. If that’s true, shouldn’t SN9 be going for a more aggressive flight profile? Why go to 12.5 km again instead of pushing further (unless there’s an FAA issue?)?

    1. They could go for a more aggressive flight profile but I think they really want to nail down the landing of Starship. It doesn’t do any good to go faster and farther if you can’t get the ship back in one piece.

      1. Indeed. As much as I support the notion you can learn a lot from failure; you can learn a lot from an intact vehicle. A test regime should flush out the vulnerabilities, and those vulnerabilities should get addressed and retested. I would say preferably prior to moving to the next step, but exceptions can occur if the next step avoids the previous vulnerabilities. In this case, SN9 would still need to land even if it flew to a higher altitude.

          1. You might as well suggest the previous 100m hop was pointless, because if they could do that, why not do 18km. Most of the reasons why those are different apply to 12.5km vs 18km. Energy loads (producing and shedding), external cooling (longer time in near vacuum), fuel margins (she can only carry so much, and she was never intended to be an Earth launcher), wider impact zone if vehicle explodes at altitude, etc.

  5. Okay, instead of getting some sleep, I took at look at a paper on Rhodium catalysts on a variety of substrates, at a range of temperatures, especially 500 C (the Raptor’s CH4 turbine exhaust temp), and pressures ranging from 1 atm to 10 atm. The S/C ratio in the paper is the steam (moles) to carbon (moles) ratio of the input stream, which was generally 1.5 to 1.

    The conversion seems to yield from 13% to 36% H2 in the dry gas stream, so some of the CH4 would circulate back up to the header tank. If the Raptor burned 1 kg of fuel, the header tank liquid volume would decrease by 2.33 liters, and at 112 K and 70 psi, it would take 1.2 moles of H2 to replace that volume, or 2.4 grams out of the original 1 kg of methane sent to the engine. Considering the extra recirculation of CH4 to the tank, and thus missing the engine on the first pass, the pre-burner throttles would need to be 1.85% higher at 13% H2 in the dry gas stream, and 0.67% higher if there was 36% H2 in the dry gas stream.

    But the stream initially is wet, because it’s mostly steam that has to be condensed out and separated from the dry gas. That can be recirculated and reheated, but it is adding complexity and weight, though I don’t know how much. The huge tanks of cryogenic fluid just above the engines are an obvious idea for a way to cause the condensation, though. But the cryogenics are also flowing into the Raptors, right next to the turbine exhaust, so the whole system could probably be a pretty compact unit with one dry gas line running up to the header tank.

  6. I was trying to come up with a simpler solution to the tank pressurant problems, and may have found one. The problem occurs when the autogenous pressurant cools down because the engines haven’t been fired in a long while during re-entry, and the only time the header tanks are required is for that final landing burn.

    So, is there some other heat source available during re-entry that could be used to warm up the gaseous LOX and CH4 in the header tanks?

    Hrm… Perhaps the hot glowing outer skin of the vehicle could be used as a heat source. The main tanks are empty, so just pump gas from the header tanks to almost anywhere else, and back, and the problem should take care of itself. The one time this won’t work is on the little test hops because on those, nothing but the engines get hot.

    1. That’s a very interesting approach. But as far as I know, right now the plan calls for the heat shield panels to be attached to the skin of Starship with bolts not designed to transfer heat from the panels to anywhere. But the idea of using active cooling on those panels has been kicked around by Elon before. So it wouldn’t surprise me one bit that some form of active gas cooling / pressurization would return as a partial heat sink for those hexagonal heat shield tiles. However fine that might work on Earth or possibly Mars it is of no help for the Moon. Might you be better off to just fire off one of the Raptors to maintain tank header pressure as needed and save all the extraneous plumbing? Tight is right.

  7. Beat me to it, George. I figured this was only a problem for these test flights and the heating available from re-entry might solve the problem. The tiles will only be on the windward side. The other side will still get hot. Will there be enough re-entry heating in the engine bay to use the autogenous plumbing on the Raptors? The best part is no part. I have taken Elon’s wise words to heart.

Comments are closed.