Appropriate Use of Jet Lifting Stages

[thiosk]

I've become interested in SSTO payload launches for operations on environments like Duna and Laythe.

I'm curious about jet-turbine powered initial ascents. It seems like for Kerbin and Laythe, jets are quite powerful for working within the atmosphere. Do folks have experience with an initial stage that burns up to 22,000 and then fires a central core to get into orbit? I'm interested in knowing what how the use of non-detach jets affect design and flight profile. I'm using couplers with 2 ramintakes per jet turbine.

If one uses a stage like that, are there any ballpark guesses for how many dV such a launch might shave off my rocket stage? in my last failed run (clustered center stage fried the skiper, I had a nice profile that carried me up to 22000 m and let me speed up into the 800s m/s before going back to 45 deg when the rocket fired. I'm interested in knowing folks' ascent profiles.

My ultimate goal would be to enable a kethane tank return from laythe for net fuel gain, though now that I'm working with the process that seems untenable. I'm not averse to those kethane-powered engines. It seems somewhat unweildy for kerbin, but I'm thinking the lower gravity of laythe would be workable with normal parts.

[Deathsoul097]

Well. Technically you are still using the same amount of DV to get to orbit, just using a much more efficient engine. However, although I see your point here, you wouldn't actually gain much from this. Frankly, if you want to lift kethane, your best bet will be to use a plane, not a rocket, because planes allow you to gain a massive of speed while still in the atmosphere, so a significant amount less fuel is needed than a rocket. Jets are best used to gain a lot of horizontal velocity incredibly efficiently, as opposed to rockets, which need to get up and out of the atmosphere as quickly to be as efficient as possible. I know this kind of thing is possible, as people have done it before, but I am saying that this particular design and plan you have is not the best way (In terms of efficiency) to go about what you want to do.

I'd say you would shave an alright amount of DV off of your rocket stage, but not an awful amount of fuel. Simply: Although you would be using less fuel in the beginning, due to the high efficiency of the jets, you would then have to carry much more fuel to lug those jets to orbit. Net result would probably be about the same, probably leaning to slightly lower consumption from the jet based lifter, but the most efficient thing to do would be to drop the jets once they run out of fuel or are about to flameout.

[tavert]

You can do vertical takeoff with jets as long as you have enough of them. Keep in mind the turbojet and Rapier only provide half their rated thrust at 0 speed. If you have enough jets to do this, then you can skip the wings and wheels and all the complexity of a horizontal takeoff.

Once you're in the air, you first want to get to high altitude to reduce drag losses, then you want to get as much horizontal speed off the jets as you possibly can. If you're running exclusively on jets when you're in the atmosphere then you want to have an effectively horizontal trajectory at the highest altitude your intakes can keep running the jets. Build up as much horizontal speed at that altitude as you can. If you have enough intakes and you're using turbojets, this can get you past orbital speed, with an apoapsis outside of the atmosphere. If you don't have enough intakes to reach orbital speed, just keep accelerating at constant altitude until you reach a steady speed. Then switch from jets to rockets and pitch up a bit.

[Bean]

I typically use 16 jets on my 4 way comms sat launches, gets me up to about 12km straight up then I light the rockets and drop the jets.

[Skorpychan]

I've tried it. It was a hell of a lot of complexity and fiddling with intakes and dumping engines, all for the sake of using jets.

Sure, it's more efficient on fuel, but what's stopping you from just using moar rockets anyway?

Just make sure that if you're using jets attached to the sides of the rocket, you don't let them flame out individually. I ended up having to decouple engines as I went up, leaving me with a central core of jets, which was then dumped when I ignited the rocket.

Make sure any engine assemblies have decouplers on top, for if you have to drop them still fueled and running because they're not giving enough thrust. Stops it pushing up into the rocket as you crank the throttle up.

[ComradeGoat]

Agree with Skorpychan - trying to make jet ascent rage rockets is annoying. Leave 'em for planes and just stick another Mainsail stage on your rocket. You'll pull less hair out that way, IME.

Edited January 26, 2014 by ComradeGoat

[Claw]

I've become interested in SSTO payload launches for operations on environments like Duna and Laythe.

I'm curious about jet-turbine powered initial ascents. It seems like for Kerbin and Laythe, jets are quite powerful for working within the atmosphere. Do folks have experience with an initial stage that burns up to 22,000 and then fires a central core to get into orbit? I'm interested in knowing what how the use of non-detach jets affect design and flight profile. I'm using couplers with 2 ramintakes per jet turbine.

I'm not sure if you asking about burn profiles on Duna/Laythe, or dV requirements for Kerbin w/ TurboJets. If you're asking about rough rocket dV requirements using TurboJet lifters off of Kerbin, I find that I usually need around 1700-2000 dV with rockets to get into orbit AFTER the TurboJets. Up until then, I find the dV estimates for TurboJets will be overly generous because the fuel available will inflate the estimated performance. This will depend a lot on your ascent profile and rocket build, but I can get about 20t payload into LKO with a 65t rocket (about 30% payload). I only use 1 intake per engine, but get going a lot faster than 800m/s before firing the rocket.

That being said, whether you decouple the TurboJet boosters or keep them for orbit, the dV requirement will be the same. What will change is needing to carry more fuel or use a more efficient rocket engine to get to orbit (giving less payload). You can typically get better payload percentages with space planes, but they're also trickier to design (IMO).

- By the way, I DON'T limit myself to terminal velocity when ascending on TurboJets. Why? Because they are already so efficient that I take advantage of the speed boost. It also gives you a little more ram air and, hence, use the jets a little longer.

- Whether you're ditching or keeping the TurboJets, toggle them using Action Groups. If you're going to ditch them, shut them down before staging to keep them from running you over.

[numerobis]

- By the way, I DON'T limit myself to terminal velocity when ascending on TurboJets. Why? Because they are already so efficient that I take advantage of the speed boost. It also gives you a little more ram air and, hence, use the jets a little longer.

Another reason: below 1km/s, the fuel efficiency of your engine grows with your speed. That wipes out a basic assumption of the proof that terminal velocity optimizes the Goddard problem.

Basically, the Goddard problem says that gravity loss decreases with speed but drag loss increases with speed, and the optimal tradeoff between those losses is at terminal velocity (where gravity loss in m/s == drag loss in m/s, pretty much by definition). For standard rockets, drag loss in m/s is directly related to drag loss in units of fuel. But that's not the case for KSP turbojets because the engine becomes more efficient when you go faster. Example: if you're going 500 m/s rather than 400 m/s, your gravity loss goes down by 20%. Drag loss in m/s goes up 56%, so with a rocket, by speeding up you burn 56% more fuel to fight drag and 20% less fuel to fight gravity. But with a jet, drag loss in fuel terms only goes up 12%. So the optimal tradeoff point will make you want to fly faster than terminal velocity -- at least until 1km/s. By that time, you're no longer flying a vertical ascent, so the Goddard problem no longer applies.

I'm not able to perfectly quantify the tradeoff with jets. For the most part, it's not really relevant: the Goddard problem assumes you have infinite thrust and want to minimize fuel usage, but in fact, extra thrust capacity is very expensive in terms of mass. So your real mass-optimal solution will have you going less than terminal velocity, with engines at max throttle.

[tavert]

below 1km/s, the fuel efficiency of your engine grows with your speed.

Uh, you mean thrust, right? Specific impulse of jets only depends on altitude.

And thanks to KSP's wonky aerodynamics, there is nothing (well, except solving the potential problem of an uneven jet engine flameout) preventing you from flying exactly the same ascent profile with your jet-powered rocket as you would with your space plane...

Agreed. Wings are overrated, and don't do much once you're at the altitudes you want to get to before building up most of your speed anyway.

Edited January 26, 2014 by tavert

[Awaras]

Frankly, if you want to lift kethane, your best bet will be to use a plane, not a rocket, because planes allow you to gain a massive of speed while still in the atmosphere, so a significant amount less fuel is needed than a rocket.

And thanks to KSP's wonky aerodynamics, there is nothing (well, except solving the potential problem of an uneven jet engine flameout) preventing you from flying exactly the same ascent profile with your jet-powered rocket as you would with your space plane and ejecting your jet engines when they are no longer useful.

[Fractal_UK]

Uh, you mean thrust, right? Specific impulse of jets only depends on altitude.

KSP displayed specific impulse doesn't change based on velocity but the technical definition for specific impulse is thrust per unit fuel mass flow. The velocity curve on those engines halves the thrust at sea level compared to high velocity flight without changing the fuel usage. In other words specific impulse has changed, even though the game doesn't tell you.

Specific impulse readings are correct for rockets in KSP but for jets, they might as well be made up they are so inaccurate.

[tavert]

The velocity curve on those engines halves the thrust at sea level compared to high velocity flight without changing the fuel usage.

Going to have to go test this now, I don't think it works that way. If fuel flow for jets were always based on max nominal thrust instead of actual velocity-dependent delivered thrust, then at super high speeds we'd continue to see high fuel usage. As far as I understand it, the only thing that's strange about the way specific impulse works with jets is that 15/16th's of the propellant mass is coming from intake air.

[Fractal_UK]

Going to have to go test this now, I don't think it works that way. If fuel flow for jets were always based on max nominal thrust instead of actual velocity-dependent delivered thrust, then at super high speeds we'd continue to see high fuel usage. As far as I understand it, the only thing that's strange about the way specific impulse works with jets is that 15/16th's of the propellant mass is coming from intake air.

The best way to test is to observe fuel usage at sea level and fuel usage at, say, 1000m/s. Testing at high speed is not helpful because you can be affected by intake air availability at high altitude, which reduces actual thrust in a way that does alter fuel consumption.

[Claw]

I agree with numerobis, specific impulse can get pretty high on a TurboJet during flight. You can watch it climb by right clicking on it during flight. I can't remember exactly, but I know I've seen much higher than the 1200 it says in the VAB. I vaguely remember in the 2000+ range. And thats on the way to reaching max thrust.

Also, I wouldn't say wings are useless as you can make space planes with TWR at or less than 1. That's pretty hard to do with vertical launch. In general, I find I need higher TWR with jet boosted craft than space planes.

Edited January 26, 2014 by Claw

[tavert]

The best way to test is to observe fuel usage at sea level and fuel usage at, say, 1000m/s. Testing at high speed is not helpful because you can be affected by intake air availability at high altitude, which reduces actual thrust in a way that does alter fuel consumption.

Well you'd really want to test while stationary vs while moving at 1000 m/s at the same altitude to compare at the same reported specific impulse. Anyway, some testing confirms that for jets, propellant mass flow (fuel + intake air) isn't given by Thrust / (9.82 * Isp), it's given by Thrust / (9.82 * Isp * velocityCurve). Strange. I can imagine where the bug came from though, fuel flow must've been calculated after scaling the rated thrust by the throttle but before scaling it by velocityCurve. The displayed fuel flow in the right-click menu is in tonnes per second including both fuel and intake air, to get to liquid fuel units in the resource bar divide by (16 * 0.005).

Javascript is disabled. View full album

I agree with numerobis, specific impulse can get pretty high on a TurboJet during flight. You can watch it climb by right clicking on it during flight. I can't remember exactly, but I know I've seen much higher than the 1200 it says in the VAB. I vaguely remember in the 2000+ range. And thats on the way to reaching max thrust.

Also, I wouldn't say wings are useless as you can make space planes with TWR at or less than 1. That's pretty hard to do with vertical launch. In general, I find I need higher TWR with jet boosted craft than space planes.

Yes but the reported specific impulse only changes with altitude. If you were to climb extremely slowly you'd still see the same reported specific impulse in the right-click menu, though apparently the fuel flow would be higher per unit thrust because of the VelocityCurve factor described above.

Wings are fairly light, but they aren't completely free. Horizontal launch imposes quite a few more structural constraints on your construction methods and the shapes of payloads you can carry, I'd usually rather just go with slightly higher TWR, skip the wings and call it done.

Edited January 26, 2014 by tavert

[Claw]

Yes but the reported specific impulse only changes with altitude. If you were to climb extremely slowly you'd still see the same reported specific impulse in the right-click menu, though apparently the fuel flow would be higher per unit thrust because of the VelocityCurve factor described above.

Fair enough. Now that I've gone back and looked again it is tied to altitude. Although what you said matches my thoughts on air intake advantage going faster.

Wings are fairly light, but they aren't completely free. Horizontal launch imposes quite a few more structural constraints on your construction methods and the shapes of payloads you can carry, I'd usually rather just go with slightly higher TWR, skip the wings and call it done.

And that's the beauty of the game. You can build/design/fly how you like.

Edited January 26, 2014 by Claw

[CaptainArchmage]

Jet engines will work on Laythe. From what I have read on Kerbin, you may only need 200-400m/s Delta-V to complete an orbit with the rocket engines with an SSTO.

I haven't used Kethane and from what I read about the Kethane jet engines, they will not operate above a certain speed. Duna's atmosphere does not have any oxygen, so you would need to use those.

[Captain Sierra]

Jet engines will work on Laythe. From what I have read on Kerbin, you may only need 200-400m/s Delta-V to complete an orbit with the rocket engines with an SSTO.

I haven't used Kethane and from what I read about the Kethane jet engines, they will not operate above a certain speed. Duna's atmosphere does not have any oxygen, so you would need to use those.

Not entirely true. The mythical 200-400m/s of rocket dV comes from basically being able to get your plane to 35000 meters and establishing an apoapsis outside the atmosphere on jets alone. The ascent profile necessary to do that is pretty demanding. Realistically, your looking at ~1500 unless you air hog like crazy (i.e. anything kisshot builds), and are a master SSTO pilot. Chances are, a person's first SSTO will be a brute-force-to-orbit with rockets rather than the high speed, high altitude approach you run with 10 intakes to an engine.

[numerobis]

Uh, you mean thrust, right? Specific impulse of jets only depends on altitude.

I mean efficiency. The Isp that's printed when you right-click only depends on altitude, but it's wrong. Level out at, say, 10km, right-click on the jet, and light up a rocket. As you speed up, you'll notice that the Isp doesn't change, fuel flow doesn't change, but thrust does change -- this is a contradiction! The physics says that F = Ve mdot but you're seeing the numbers on the right side not changing while the number on the left side does. You can independently verify that F is correct, and that mdot is correct; that means that Ve is the one that's displayed wrong.

The way in which it's wrong makes sense from a programming perspective: the KSP physics engine first computes Isp from the pressure and prints that, then computes mass flow from the Isp and the max thrust that's listed in the part and prints that (and draws fuel according to it); and then computes thrust from the velocity and prints that (and applies a force according to it).

Edit: oh, I see you all talked about this on the following page of replies. tavert: "at super high speeds we'd continue to see high fuel usage" -- we do.

Edited January 26, 2014 by numerobis

[tavert]

Yeah, did not catch that quirk about jets until 6 posts later.

The way in which it's wrong makes sense from a programming perspective: the KSP physics engine first computes Isp from the pressure and prints that, then computes mass flow from the Isp and the max thrust that's listed in the part and prints that (and draws fuel according to it); and then computes thrust from the velocity and prints that (and applies a force according to it).

It "makes sense" in that we can see how the bug happened, but it's still undoubtedly a bug. There's also throttle coming into the picture somewhere in that sequence, both user-set and (new in 0.23) from running low on intake air. Those effects should also be tested.

[numerobis]

As for wings: they achieve two things. First, you can reduce the engine mass if you accept more deltaV expenditure. That's a win for jets, not so much with rockets. Second, if the wings are control surfaces, you can get thrust off them via infiniglide.

[Claw]

From what I have read on Kerbin, you may only need 200-400m/s Delta-V to complete an orbit with the rocket engines with an SSTO.

Not entirely true. The mythical 200-400m/s of rocket dV comes from basically being able to get your plane to 35000 meters and establishing an apoapsis outside the atmosphere on jets alone. The ascent profile necessary to do that is pretty demanding. Realistically, your looking at ~1500 unless you air hog like crazy (i.e. anything kisshot builds), and are a master SSTO pilot. Chances are, a person's first SSTO will be a brute-force-to-orbit with rockets rather than the high speed, high altitude approach you run with 10 intakes to an engine.

200-400m/s of rocket dV isn't mythical at all but you're right, it does take a bit of finesse. I have several space planes that get into 100km+ orbit with 85m/s of rocket dV and they aren't particularly tricked out. Just stock with three ram intakes and one turbo jet, no infiniglide.

I feel the original post was in reference to using jets as boosters on a rocket. Although there's no reason you can't fly the same profile, it's a lot harder unless you want to plaster on the jet engines to your vertical launcher and fly it like a space plane. It certainly depends a lot on what kind of payloads you're asking for too.

That's why I said earlier that using jets as boosters on a rocket type profile takes around 1700-2000 m/s.

[Vanamonde]

I think most replies are missing OP's intention. He seems to be talking about using jets at multiple points on an interplanetary flight, which does not promise to be very advantageous. Laythe is the only other world which has an atmosphere that allows jets to generate thrust at all, but even there, they will be less powerful than on Kerbin. In other words, after you leave Kerbin's atmosphere, any jet engines you're carrying will be more hindrance than help.

[thiosk]

I think most replies are missing OP's intention. He seems to be talking about using jets at multiple points on an interplanetary flight, which does not promise to be very advantageous. Laythe is the only other world which has an atmosphere that allows jets to generate thrust at all, but even there, they will be less powerful than on Kerbin. In other words, after you leave Kerbin's atmosphere, any jet engines you're carrying will be more hindrance than help.

This is getting to the issue, though the discussion here is critically important to me as well. I've so far failed anything other than the most basic of planes, and am having a lot of difficulty with the transition from jet flight to rocket.

The true purpose of my mission is to carry a full long kethane tank from laythe surface to the orbital space station. Thats a big, heavy payload, and so far, theres no way I can envision accomplishing that without a net loss in fuel- I tried to put together an SSTO that could land and move a full orange tank payload, and it just wasn't tenable. It may not be tenable even with modded kethane jet engines that can drink off their payload, necessitating less rocket.

Captain Sierra's comment about 10 rams/jet is jaw dropping, I didn't realize it was possible. More jets apparently enable higher altitude flight before flameout, which is really what I expect I'll be needing.

This evening i've got some time to spend working on my designs, so I'll post some pictures of my development process here.

Please feel free to continue debating ascent profile, with the caveat that we are talking about laythe-centric SSTOs for space station resupply so dropping the engines is strictly a no-no.