The Asparagus Thread

[shynung]

If I build from a design, and then test and add as small an improvement as possible as quickly as possible (Soviet Style), than is that considered typical?

Yes.

In the engineering world, it is very rare to see anything work right in the first try. Even simple parts can receive tens to hundreds of design iterations, depending on how the part is supposed to interact with other parts. Especially in the case of turbomachinery, a micrometer (1/1000^th of a millimeter) could make or break the design.

Wait, aren't turbopumps effectively compressors? Is it that hard to build a compressor? Once you have the molds you can do it easily...........

Molds for turbocharger housing, as far as I know, are made out of sand. The factory first make a clay model of the final product, then encase it in sand. They then heat the sand, melting the clay out, leaving the sand mold. After pouring in liquid aluminium, the mold is broken, freeing the final product. Effectively, the molds are one-time use.

BTW: 40/200 = 1/5. So, that's actually quite a lot. (Sorry, I had to.)

I either use nothing: 200000

Or a comma: 200,000

Pitiful you are for using different writing conventions to defend yourself.

[KASASpace]

Molds for turbocharger housing, as far as I know, are made out of sand. The factory first make a clay model of the final product, then encase it in sand. They then heat the sand, melting the clay out, leaving the sand mold. After pouring in liquid aluminium, the mold is broken, freeing the final product. Effectively, the molds are one-time use.

If I recall correctly they have a base mold, and then make the other molds based off of that one mold. For some things. Maybe not turbo pumps. But so much hasn't yet been applied to rocketry that should have been years ago. Industrialization for example.

Pitiful you are for using different writing conventions to defend yourself.

Actually, that was a joke.

Pitiful you couldn't see that.

[shynung]

If I recall correctly they have a base mold, and then make the other molds based off of that one mold. For some things. Maybe not turbo pumps. But so much hasn't yet been applied to rocketry that should have been years ago. Industrialization for example.

Wasn't the military missile mass production an industrialized example of the rocket industry? Granted, they're smaller, but they are quite similar.

Actually, that was a joke.

Pitiful you couldn't see that.

[sgt_flyer]

The problem is, turbopumps is the single most expensive part of a rocket engine (up to 55%)

Before even thinking about asparagus, we need cheap, lightweight & ultrareliable pumps and the combination of the 3 is extremely difficult to get.

(Heck, the RL-10 engine is the single most expensive part on rockets that use it - excluding the payload)

So the gains of building 'smaller engines' would soon be offset by the additional turbopumps needed to transfer fuel - because you'll need to drive the pump with the rocket engine, you'll also lose efficiency (more kinetic energy lost due to the need to drive the pumps), and thus ISP, so the thrust / fuel savings are even more diminished at greatly increased costs.

[KOCOUR]

Yes.

In the engineering world, it is very rare to see anything work right in the first try.

Dont know, where i work, if yor design doesnt work on first try, management gets all "Are you still fit for this job? Aren´t we paying you too much?" :-)

[TerLoki]

Dont know, where i work, if yor design doesnt work on first try, management gets all "Are you still fit for this job? Aren´t we paying you too much?" :-)

Nobody ever said managers were engineers, or particularly bright.

[shynung]

Dont know, where i work, if yor design doesnt work on first try, management gets all "Are you still fit for this job? Aren´t we paying you too much?" :-)

That would happen if said design completely fails. If it kinda works, just not perfect, it will be revised at least once.

[sgt_flyer]

In the end the problem with asparagus staging in real life can be resumed as such : does the benefits (basically, slightly higher payload fraction, outweights the inconvenients : increased complexity (with all that can bring in lower reliability, and higher part cost & R&D) After all, fuel is a very small fraction of the launch vehicle price. (And you still have to pay for the logistics and manpower for mission control.

In this case, spaceX is trying to devellop this technology - if they are successful, others will try to copy it if they think they can achieve similar benefits to outweight the inconvenients.

[Ralathon]

In the end the problem with asparagus staging in real life can be resumed as such : does the benefits (basically, slightly higher payload fraction, outweights the inconvenients : increased complexity (with all that can bring in lower reliability, and higher part cost & R&D) After all, fuel is a very small fraction of the launch vehicle price. (And you still have to pay for the logistics and manpower for mission control.

In this case, spaceX is trying to devellop this technology - if they are successful, others will try to copy it if they think they can achieve similar benefits to outweight the inconvenients.

Space X is working on this because they can retrieve the boosters with their soft landing schtick. If you can retrieve your hardware intact asperagus can be worth it. It just doesn't make sense as disposable launch system.

[R0cketC0der]

The reason SpaceX is using fuel crossfeed in the FH is to make the payload capacity higher, not to reduce cost. Fuel crossfeed will only be used if the payload mass is over 45 tons, for any payloads below that crossfeed won't be used and the center core will be throttled down at launch and throttled back up after the booster cores separate.

When they reuse it, fuel cost will still only be a tiny fraction of the costs.

[Streetwind]

There's one large problem with KSP style asparagus staging that nobody mentioned yet, and that is the fact that KSP conveniently handwaves a law of physics that real life unfortunately insists on maintaining: the conservation of momentum.

When you pump fuel from one side booster to the next, you create torque on the rocket. And since fuel is the single largest mass component in a booster, you're pumping about nearly the entire booster's mass, and thus generate that much torque. Your rocket will spin like a top as it ascends, and if there's one thing that kills rockets, it's spinning. Even if the structural integrity holds up to the forces, fuel will be centrifuged to the walls of the tanks and the engines will flameout.

SpaceX can do fuel crossfeed on their Falcon Heavy because they have only two outer boosters, both of which pump into the center booster. This is not asparagus staging, it's onion staging. It creates no torque. If you have more than two outside boosters, you can still keep having all of them pump into the center; but that's still onion staging. Asparagus specifically implies that you pump from one outer booster to the next, and that creates massive torque.

There is a small number of configurations that can produce something roughly like a two-stage asparagus, for example a setup with eight boosters set up so that two of them pump left and two of them pump right in a completely symmetric fashion, then drop 4 boosters and continue with 4, then drop the second 4 and continue with the center. Or one with six boosters where four boosters pump into two in a balanced symmetric fashion, then you drop first 4 then 2. Or there's even the possibility of having 4 cores, where two of them pump both left and right at the same time, so you can actually drop pairs. But this last approach means you need not only the pumping hardware for regular fuel crossfeed - no, you need a double set of that hardware in each core! And if there's even a slight imbalance between pumping speeds, the rocket still starts spinning. And going up to a three-stage asparagus gets even more complicated (a six-booster could work, again with doubled pumping hardware on every booster).

Overall, it's an engineering nightmare where so much can go wrong. It's not worth it.

When they reuse it, fuel cost will still only be a tiny fraction of the costs.

Case in point: Elon Musk quoted the fuel cost of a Falcon 9 v1.1 rocket at 0.3% of the total launch price during a recent press conference.

[Red Iron Crown]

When you pump fuel from one side booster to the next, you create torque on the rocket. And since fuel is the single largest mass component in a booster, you're pumping about nearly the entire booster's mass, and thus generate that much torque. Your rocket will spin like a top as it ascends, and if there's one thing that kills rockets, it's spinning. Even if the structural integrity holds up to the forces, fuel will be centrifuged to the walls of the tanks and the engines will flameout.

SpaceX can do fuel crossfeed on their Falcon Heavy because they have only two outer boosters, both of which pump into the center booster. This is not asparagus staging, it's onion staging. It creates no torque. If you have more than two outside boosters, you can still keep having all of them pump into the center; but that's still onion staging. Asparagus specifically implies that you pump from one outer booster to the next, and that creates massive torque.

You're ignoring the torque applied by the fuel stopping its lateral motion when it reaches the destination tank. There is a small net torque from the center of mass of each half of the booster array moving around the rocket, but at worst that's 180 degrees of rotation over the burn of the entire asparagus stack and stops of its own accord when the booster ring burns out. If it's important to keep the rocket oriented a particular way, those samller torques are easily counteracted by thrust vectoring or aerodynamic controls.

That doesn't solve any of the other problems of asparagus staging, but the torque from fuel pumping is not a significant problem.

[Dkmdlb]

There's one large problem with KSP style asparagus staging that nobody mentioned yet, and that is the fact that KSP conveniently handwaves a law of physics that real life unfortunately insists on maintaining: the conservation of momentum.

You can do a modified asparagus staging to solve this problem. Six boosters at the 2, 4, 6, 8, 10, and 12 o'clock positions. The 10 and 2 position boosters pump into the 12 booster, the 4 and 8 boosters pump into the 6 booster. Then you drop the 10, 2, 4, and 8 boosters, leaving the noon and 6 boosters to pump into the core and then get dropped.

I'm making it a little more complicated than it is. You lose a little efficiency by dropping 4 boosters at the first stage, but you gain the benefits of a rocket that doesn't produce any torque via fuel flow.

Or you could just not worry about it and deal with the slight rotation.

[Jouni]

Net torque might not be too relevant here. The lateral forces could be strong enough, that the rocket just rips itself apart from inside.

[Red Iron Crown]

Net torque might not be too relevant here. The lateral forces could be strong enough, that the rocket just rips itself apart from inside.

That's a good point. But the idea that pumping fuel around the ring is going to make the rocket spin like a top is ignoring half the forces involved.

[Streetwind]

You're ignoring the torque applied by the fuel stopping its lateral motion when it reaches the destination tank. There is a small net torque from the center of mass of each half of the booster array moving around the rocket, but at worst that's 180 degrees of rotation over the burn of the entire asparagus stack and stops of its own accord when the booster ring burns out. If it's important to keep the rocket oriented a particular way, those samller torques are easily counteracted by thrust vectoring or aerodynamic controls.

Hmmm, maybe you're right. I'm not that good with physics. But it just seems to me that there's a whole lot of forces involved that can easily destabilize the rocket due to the sheer amount of fuel changing position. I remain convinced that that is a bigger issue than building the hardware... a rocket is at its most vulnerable during atmospheric flight, after all. Desasters happen almost always within the first 2-3 minutes of the 10-minute ascent.

You can do a modified asparagus staging to solve this problem. Six boosters at the 2, 4, 6, 8, 10, and 12 o'clock positions. The 10 and 2 position boosters pump into the 12 booster, the 4 and 8 boosters pump into the 6 booster. Then you drop the 10, 2, 4, and 8 boosters, leaving the noon and 6 boosters to pump into the core and then get dropped.

I'm making it a little more complicated than it is. You lose a little efficiency by dropping 4 boosters at the first stage, but you gain the benefits of a rocket that doesn't produce any torque via fuel flow.

Hate to be snarky, dude, but really... you could at least read my post, before you go trying to explain to me the exact same thing I wrote about

[rkman]

If two launchers have the same dV, the fuel requirements are irrelevant. Only the costs and the reliability are relevant. If your launcher only takes half the fuel but has more seperation events, it is an inferior launcher.

If it is assumed that the seperation events significantly increase cost and reduce reliability.

The practical tricky part that i mentioned would be to develop asparagus separation technology that is not to costly and not to unreliable. You seem to think that is principally impossible, i'm not so sure. Otoh i'm not looking at my crystal ball and say "it will happen".

[sgt_flyer]

Hehe like in programmation which can work in rocketry : it's easy to make something complicated, it's very hard to make something simple.

(And R&D costs grows accordingly too )

(The original say : there are two ways to make a program : the first is to make the program so simple that it's obvious there are no flaws, the second is to make it complicated so that there is no obvious flaws. The first one is by far the most difficult)

[KASASpace]

Let me ask you this N_las:

How many separation events occur on the Soyuz Launch Vehicle?

First it launches

Then after it burns out the fuel in the boosters, it jettisons them

Then the core goes on alone

Then the upper stage is activated after core separation.

At least 2 for the first/second stage alone. Yet it's one of the most reliable LVs ever created. How much of an impact would adding some limited crossfeed capability do? You can use all of the present pumps, you just need extra piping. That adds at most a few hundred kilograms. But the payload increase would be very worthwhile.

[KASASpace]

The problem is, turbopumps is the single most expensive part of a rocket engine (up to 55%)

Before even thinking about asparagus, we need cheap, lightweight & ultrareliable pumps and the combination of the 3 is extremely difficult to get.

(Heck, the RL-10 engine is the single most expensive part on rockets that use it - excluding the payload)

So the gains of building 'smaller engines' would soon be offset by the additional turbopumps needed to transfer fuel - because you'll need to drive the pump with the rocket engine, you'll also lose efficiency (more kinetic energy lost due to the need to drive the pumps), and thus ISP, so the thrust / fuel savings are even more diminished at greatly increased costs.

I know it is the most expensive part. The Russians were actually quite ingenious in their engine clustering techniques. Just use one turbopump for them all.

The RD-180 uses one turbopump. So does/did the RD-170. Those were large engines, and are typically counted as only one engine. I personally think it's how they managed to weasel their way into getting the most powerful liquid engine. But it really is an ingenious technique, and should be used more often.

[ThinkOutsideTheHangar]

Asparagus staging is the only reason SSTOs that can carry 500 tons is possible. Period.

[KASASpace]

Asparagus staging is the only reason SSTOs that can carry 500 tons is possible. Period.

It wouldn't be a SSTO if it was asparagus........

[Red Iron Crown]

Let me ask you this N_las:

How many separation events occur on the Soyuz Launch Vehicle?

First it launches

Then after it burns out the fuel in the boosters, it jettisons them

Then the core goes on alone

Then the upper stage is activated after core separation.

At least 2 for the first/second stage alone. Yet it's one of the most reliable LVs ever created. How much of an impact would adding some limited crossfeed capability do? You can use all of the present pumps, you just need extra piping. That adds at most a few hundred kilograms. But the payload increase would be very worthwhile.

If it is so easy to asparagus the Soyuz to make it more efficient while maintaining its stellar reliability, I have to wonder why the Russians haven't done so. They're not exactly dolts when it comes to rocket science.

Maybe their analysis, with full access to all the flight data, schematics, actual hardware to test, and staff of engineers and rocket scientists, has reached a different conclusion.

[xcorps]

If it is so easy to asparagus the Soyuz to make it more efficient while maintaining its stellar reliability, I have to wonder why the Russians haven't done so. They're not exactly dolts when it comes to rocket science.

Maybe their analysis, with full access to all the flight data, schematics, actual hardware to test, and staff of engineers and rocket scientists, has reached a different conclusion.

If only the Soviets allowed access to the KSP forums.

[KASASpace]

If it is so easy to asparagus the Soyuz to make it more efficient while maintaining its stellar reliability, I have to wonder why the Russians haven't done so. They're not exactly dolts when it comes to rocket science.

Maybe their analysis, with full access to all the flight data, schematics, actual hardware to test, and staff of engineers and rocket scientists, has reached a different conclusion.

They haven't considered because they're satisfied with its current performance.

"If it ain't broke, don't fix it."