sevenperforce

While I doubt this is the case, seeing headlines like "SpaceX lofts satellite; botches landing" make me want to tear my hair out.

Manned mission to Mars on a kickstarter? Nah. Mars flyby on a kickstarter? Sure. Step 1. Identify some yet-unrealized science objective that can be accomplished at Mars via flyby. Maybe it's measuring magnetic field fluctuations, maybe it's sampling upper-atmosphere isotopic ratios. I don't know. Step 2. Partner with a university that has an active planetary research group and get one of the grad student groups to design a lightweight instrument to achieve that science objective. Get their advisor to write a grant proposal for money to build the instrument. Step 3. Design a spacecraft to carry the instrument, make some sexy mockups, and crowdfund the cost of building it, promising that anyone who donates over $100 will have their name engraved on it. Step 4. Respectfully ask Elon and NASA to use a reusable Falcon Heavy in place of a Falcon 9 for one of their ISS resupply missions, so that the upper stage has enough dV remaining to put our intrepid voyager on a Mars Transfer.

I got to watch the live stream in real time with my two kids...they're both toddlers but they loved it. Kept shrieking and pointing and exhulting about the "Fawcon Wocketship" and making big whooshing sound effects. The older one was particularly interested in the SECO and reignition sequence. "Da wocketship is fyying faster dan an airpwane?!"

That's what compression is for.

I definitely like the helium loop being used to run the compressor and so forth. I hadn't been aware of that before. The reason I want to use onboard oxidizer is that it will allow a much higher-volume flow of compressed air. If you're injecting into the combustion chamber, then you can only increase mass flow up to a limit, and then you start to lose efficiency in your burn. Inject downstream, and your air mass flow can be almost arbitrarily high. You can still use a fuel rich combustion if applicable to afterburn.

Indeed. To use air as your working mass, you need to add kinetic energy to it. You can do that at low speeds with a simple fan, but supersonic flow doesn't mix with fans, so if you're trying to book it, you need to use an expansion cycle, converting thermal energy to kinetic energy. But that is wildly inefficient unless you have plenty of expansion, which means you need compression at the start. Compressing with a turbine compressor runs into the same problem as a fan unless you slow down the airstream...but doing so heats it up and will melt your compressor. So you need to precool aggressively, or you're stuck with ram compression, which is limited and inefficient. Ah, the tyrannies of physics. I still say they would do better to skip the air breathing aspect and just use the air as working mass injected downstream of the combustor.

Isn't that because the compressor-to-expander cycle really isn't set up to work with really high inlet speed? I'm convinced (possibly without merit, but who knows) that some kind of high-bypass air turborocket expander is the key to making SSTO instantly realizable.

In the interest of eliminating heavy moving parts... There has to be a way of designing inlet geometry such that increase in speed moves the shock surface out of the inlet precisely in accordance with the optimal efficiency requirements.

What would lead you to the conclusion that coking isn't a consideration?

Well, this would be a high pressure device. Moreover, the compressor serves only to inject the compressed working mass into the exhaust bell; the kinetic energy in the exhaust bell does the work on the mass. It was my understanding that supersonic jets use low-bypass turbofans or turbojets primarily due to heat dissipation problems.

Well, regarding energy, I'm not interested in adding an equivalent amount of energy to the airflow. I'm interested in adding substantially lower energy, actually, but adding it to a much larger mass flow, so that the momentum impulse is higher. Lower fuel consumption but greater thrust. Curious, about the helium loop. How does the engine get started, with no initial airflow?

You're forgetting all the bits and pieces needed to attach the actual LM's descent tanks to the descent engine. Here, I'll actually run the numbers so you can see what I'm talking about. My hypothesis is that with a throttleable engine and break-away drop tanks (neither of which were possible from an engineering perspective during Apollo), a single engine would result in a lower initial LM mass. The initial LM massed 15.2 tonnes on separation from the CM, for an effective dry mass of 7 tonnes. Since we know the specific impulse of the descent engine was 311 seconds, this allows us to estimate the onboard dV: 2,363 m/s. However, this included two minutes of hover time, or 194.6 seconds, reducing the actual descent dV to 2,169 m/s. The 7 tonnes of dry mass comprised the descent engine (180 kg), the descent tanks (mTD), the landing legs (mLL), and the ascent launch mass (mAL). We don't know those three values; that's why I used symbols. However, if the ascent dV was comparable to the descent dV, then based on the ascent engine's specific impulse we can estimate the mass fraction at 51%, with mAL = 4.517 tonnes and an ascent dry mass of 2.217 tonnes. T/W ratio at launch was 2.18. The descent tanks and landing legs together (mTD + mLL) mass 2,663 kg. Let's suppose I use the stock descent engine for both burns (even though a smaller engine would be able to do the trick, if my hypothesis is correct). Ascent launch dry mass goes from 2,317 kg to 2,412 kg...bad, I know. But I also know that the ascent engine was fixed-thrust; it couldn't be throttled. Using its thrust, exhaust velocity, and mass flow, I can determine that its burn period was 438.6 seconds, representing 711.4 m/s of gravity drag losses. Since the thrust of the stock descent engine is 2.8 times greater, that shaves off 458.46 m/s of gravity drag, reducing our required ascent dV to 1,711 m/s. Our fuel requirement? 1,817 kg, for an ascent launch mass of 4,229 kg. So...what does our dry mass look like on the descent stage? The descent stage dry mass is now the ascent launch mass (4,229 kg) and the descent tanks and landing legs (2,663 kg) for a total dry mass of 6,892 kg. This means we only need 8,072 kg of fuel to match the descent dV of 2,363 m/s. A net total LM mass savings of 237 kg. Of course, a smaller engine could therefore be used, further reducing fuel and weight. Right, it was the "or" that I was talking about. Thrust won't be a problem; the Dragon V2 has plenty of that for a moon landing. I was talking about a crasher stage configuration, where a separate stage "drops" the Dragon at a negligible velocity. It would then return to orbit on its own for LOR.

The goal would be to have a lower fuel+oxidizer flow, actually, but with a mass flow several times greater. You already have to have a preburner or gas generator to run your turbopump, so beef it up slightly and run it to your compressor. Cool the incoming airstream only as much as is needed to keep your compressor from melting, and inject the pressurized air into the exhaust stream. So, basically a precooled air turborocket. Come to think of it, you don't actually have to have a preburner or gas generator. You can bleed high-pressure gas straight from the combustion chamber and use it to run your compressor and turbopump, then exhaust it into the airflow that is compressed by the compressor. Up to around 3 km/s, the exhaust flow would be more air than propellant. The Russians used oxy-rich preburning for the turbopumps because preburning fuel-rich RP-1 will coke and choke. Don't think it's an efficiency thing. Oxy-rich preburning is horribly corrosive and generally a nasty nasty affair but it's better than filling your engine with coal tar. I like this idea. However, it's not too far from what SpaceX is already doing. Staging speed for the reusable first stage of the Falcon 9 is around 2 km/s and it takes place out of atmo. You'd have to show how doing this with a partial airbreather would be better than doing the same identical thing with the Falcon 9 first stage. Also, what would the second stage engine need to look like?

Trolling apparently. My first guess was that paint would have helped, but as it turns out, it didn't -- the aerodynamic forces on the insulation foam were just too great, and there was always a little shedding even with paint. Using a mesh or other layer to try and contain the foam would have likely caused larger pieces to break off en masse and result in more damage. After Columbia, they mostly just tried to reduce the amount of foam and keep it as smooth as possible. Totally. Just grab one of those massive, fully-filled fuel tanks that happens to be floating by in orbit and...oh, wait. Seriously, the heat shields weren't the problem. They never were. Having a really, really ridiculously large spaceplane with a lot of exposed surface area, particularly exposed to debris from a giant tank suspended above it covered in fragile insulation, was the problem. Heat shields are not a problem. We've been using them for decades and they work really, really well. Sure, it seems scary to ride from space to Earth on a column of flame...but then again, that's what a rocket launch is in reverse.

Heavy. Really, really heavy. They stopped painting the external tank because the weight of the white paint was too great. (IIRC, paint probably would have kept the foam from breaking off and striking Columbia in the first place.) The fact is, we never should have put our crew vehicle anywhere other than on the very top of our rockets. The side-slung configuration was chosen so that the Shuttle's engines could thrust from launch to orbit.

Better than using liquid hydrogen.

...what extra dry mass? The ascent stage has a SLIGHTLY heavier engine, sure, but that's more than compensated for by the fact that you didn't have to carry a whole extra stinking engine down with you. Step 1: You launch an engine, a lunar module, a re-entry capsule stocked with consumables, landing legs, and four sets of fuel tanks to lunar transfer orbit. Step 2: The engine uses 1 set of tanks to enter LLO, then drops them. Step 3: The engine, 2 sets of tanks, the lunar module, and the landing legs break off from the capsule and last fuel tank set. Step 4: Engine uses one set of tanks to land propulsively. Conduct lunar excursion. Step 5: Engine, lunar module, and full tank break away from empty tank, landing legs; ascend to LLO and dock with waiting capsule/tankage. Step 6: Drop empty tank; head home on remaining tank.

And I'd wager that their primary boost to effective ISP comes from passing a large bulk of nitrogen through their engine. In fact, I bet you could get significantly better effective ISP and a lower dry mass by relaxing the combustion requirements and opting for a less demanding, denser cryogen like slush-liquid methane. Plan on carrying all your oxidizer, use fuel-rich staged combustion for your turbine, and set up the turbine to also run an air compressor. The compressed air gets pumped straight into the neck of your exhaust bell, with an intended air mass flow greatly exceeding your fuel mass flow. As you speed up, use the slush methane to partially precool the incoming air...but not as much as SABRE, since you aren't going to need to combust it. Should be able to operate further than SABRE can and with a far greater effective ISP.

No experience getting that kind of mass fraction to orbit.

Yeah, turnaround on the first few relaunches for each Falcon 9 will be fairly short, with little refurbishment. Probably more work required after the 4th or 5th relaunch, then chopped for parts after a dozen launches total. Though chopping will be more along the lines of rebuilding each component to reuse abt 80% in a "new" rocket.

Getting that kind of mass fraction to orbit is definitely unproven.

That was my point; it's not a Direct Ascent. Still a LOR. But why would the LM be heavier? It should be substantially lighter. It doesn't have to carry a second engine. In fact, it would be virtually identical to the LM of Apollo, minus the weight of the ascent engine. Meanwhile, the orbiting command module is much smaller, containing only return-trip fuel, return-trip consumables, and the re-entry capsule. So the overall launch mass is considerably lower. Compared to Apollo, you save on the weight of an entire pressurized command module, an entire engine, and part of the descent fuel. Now, if that crasher stage could carry enough fuel to boost itself back into orbit.... Actually, if you gave a Dragon V2 external tanks, I'm fairly sure that such a configuration could manage a manned moon landing and return with a Falcon Heavy and a Dragon V2 alone.

Not quite. The LOR lander carried two engines down, leaving a third engine in orbit with its own fuel tanks. I'm saying to keep the LOR approach, but use the same engine for all three burns. Separate fuel tanks for the trip to Earth stay with the command module in orbit.

SpaceX is betting on humanity. Big Dumb Boosters are betting on...something else?

Yeah, did a little digging and it doesn't look like methane will coke even when run very fuel-rich. If that's the case I'm not sure why they wouldn't ditch the oxy preburner and just run everything off the fuel-rich preburner. Hot oxy is bad stuff.