Is asparagus the best staging system? (might contain science)

[Tankbuster32]

My greatest problem with asparagus is TWR. I either have way too much at launch and sufficient after dropping the tanks, or I have the prefect TWR at launch (as I prefer to have) and way too little after dropping the tanks. This makes Asparagus impractical for realistic-ish heavy-lift launch vehicles.

[Caelib]

I've done quite a bit of experimentation with rocket design and found that in the vast majority of scenarios, asparagus is just a band-aid for inefficient design or to compensate for trying to launch an excessively oversize payload.

You would be amazed at how much you can do with so little if you're efficient -- I would highly recommend using Kerbal Engineer mod to show you the delta-V for each stage so you can experiment with different configurations to see how different combinations of parts affects your design.

I would only recommend using radial engines/boosters for the first/launch stage. In my opinion, fuel lines are a hack

[Pecan]

This is a very old thread with a very simple conclusion: good staging improves payload ratio and asparagus is the optimal staging strategy. On the other hand it is fiddly, uses a lot of parts and, since funds became significant, it also tends to be expensive in all the engines you're throwing away. If using an aerodynamic mod like FAR it is important to note that drag increases with the width of the rocket, so no pancakes - more than one 'layer' around the core is more trouble than it's worth in any case.

One of the main things to remember when discussing asparagus staging on these forums, however, is to remember how many people can't tell the difference between asparagus and any other radial-staging strategy.

Personally, I've finished with asparagus since the engines were rebalanced and it's so easy to SSTO any payload I'm interested in (<= 100t, hardly ever over 40t) with simple, low-part, builds and full recovery of parts so launches only cost fuel. Without FAR's nerfing of jets it's even fairly simple and low-part to SSTO 40t with a spaceplane so you don't even use much fuel in the first place.

[magnemoe]

This is a very old thread with a very simple conclusion: good staging improves payload ratio and asparagus is the optimal staging strategy. On the other hand it is fiddly, uses a lot of parts and, since funds became significant, it also tends to be expensive in all the engines you're throwing away. If using an aerodynamic mod like FAR it is important to note that drag increases with the width of the rocket, so no pancakes - more than one 'layer' around the core is more trouble than it's worth in any case.

One of the main things to remember when discussing asparagus staging on these forums, however, is to remember how many people can't tell the difference between asparagus and any other radial-staging strategy.

Personally, I've finished with asparagus since the engines were rebalanced and it's so easy to SSTO any payload I'm interested in (<= 100t, hardly ever over 40t) with simple, low-part, builds and full recovery of parts so launches only cost fuel. Without FAR's nerfing of jets it's even fairly simple and low-part to SSTO 40t with a spaceplane so you don't even use much fuel in the first place.

Yes, with the 3.75 meter parts and funds I mostly launch single core rockets with SRB support, this is enough for up to 100 ton even with reuse of core, more if you discard the core stage and circulate with the nuclear engines, majority of +100 ton payloads is interplanetary ships.

If you need aspargus as in more in 3 cores and crossfeed+ srb you are into real Wachjobian designs.

[Red Iron Crown]

This is a very old thread with a very simple conclusion: good staging improves payload ratio and asparagus is the optimal staging strategy.

I believe twisted candle staging can actually deliver a greater payload fraction, but that's a bit of a special case and doesn't scale very well.

Personally, I've finished with asparagus since the engines were rebalanced and it's so easy to SSTO any payload I'm interested in (<= 100t, hardly ever over 40t) with simple, low-part, builds and full recovery of parts so launches only cost fuel. Without FAR's nerfing of jets it's even fairly simple and low-part to SSTO 40t with a spaceplane so you don't even use much fuel in the first place.

Same here, I only use asparagus now if I'm making an Eve ascent rocket which needs every last drop of efficiency. For everything else, onion staging is almost as efficient and much quicker and easier to construct.

[Frozen_Heart]

Though asparagus pancake rockets are technically superior, I simply prefer building linear staged rockets due to aesthetics. I can still manage 10% payload or greater with them.

[Doc]

One reason asparagus isn't used in 'real' rockets is this:

The fuel flow need to run say, a Saturn F1 engine cluster is about 1/6th the flow rate needed to drive one of the main turbines at hoover dam - a *tremendous* amount of liquid, running in 2 distinct systems.

Draining 2+ separate tanks, and keeping everything nice and balanced, at that depletion rate requires some really good (read - large and expensive) pumps and valving solutions. Further, the rocket needs to be able to switch between 'booster' and 'main stage' fuel sources without inducing so much as a bubble or 'boom' goes the rocket.

The added cost in $$$, extra weight, and increased risk has made the gains just not worth it. Rocket fuel is pricey, but not *that* bad, compared to the cost of say, an engine (the biggest cost of which is the pumping machinery).

It's the same reason we don't use variable-bell nozzles on all our rockets; yes, they are 'better' in theory, but the cost of implementing them is almost as high as what you will save.

[Red Iron Crown]

One reason asparagus isn't used in 'real' rockets is this:

The fuel flow need to run say, a Saturn F1 engine cluster is about 1/6th the flow rate needed to drive one of the main turbines at hoover dam - a *tremendous* amount of liquid, running in 2 distinct systems.

Draining 2+ separate tanks, and keeping everything nice and balanced, at that depletion rate requires some really good (read - large and expensive) pumps and valving solutions. Further, the rocket needs to be able to switch between 'booster' and 'main stage' fuel sources without inducing so much as a bubble or 'boom' goes the rocket.

The added cost in $$$, extra weight, and increased risk has made the gains just not worth it. Rocket fuel is pricey, but not *that* bad, compared to the cost of say, an engine (the biggest cost of which is the pumping machinery).

It's the same reason we don't use variable-bell nozzles on all our rockets; yes, they are 'better' in theory, but the cost of implementing them is almost as high as what you will save.

Yeah, I would say the most we'll see IRL is crossfeed to the center stack (planned for Falcon Heavy, possibly others). Beyond that the complexity starts getting out of hand.

[Sylandro]

I honestly prefer IRL designs, for some reason.

[Pecan]

I believe twisted candle staging can actually deliver a greater payload fraction, but that's a bit of a special case and doesn't scale very well....

Indeed, you are correct. Thank you for reminding me.

I've only used candle a couple of times, it's interesting and pretty flexible. Using several slack-tank stages gives quite a bit of room to add the 'twist' and it's easy to tune the thrust when it's time to add new engines.

[LethalDose]

This is a very old thread with a very simple conclusion: good staging improves payload ratio and asparagus is the optimal staging strategy. On the other hand it is fiddly, uses a lot of parts and, since funds became significant, it also tends to be expensive in all the engines you're throwing away. If using an aerodynamic mod like FAR it is important to note that drag increases with the width of the rocket, so no pancakes - more than one 'layer' around the core is more trouble than it's worth in any case.

One of the main things to remember when discussing asparagus staging on these forums, however, is to remember how many people can't tell the difference between asparagus and any other radial-staging strategy.

Personally, I've finished with asparagus since the engines were rebalanced and it's so easy to SSTO any payload I'm interested in (<= 100t, hardly ever over 40t) with simple, low-part, builds and full recovery of parts so launches only cost fuel. Without FAR's nerfing of jets it's even fairly simple and low-part to SSTO 40t with a spaceplane so you don't even use much fuel in the first place.

The age of the thread indicates that it was started prior to any concerns about the cost of the method in funds, so in some ways the necro was somewhat justified because the game environment changed. Asparagus staging is mass efficient, but isn't cost efficient.

You really hit the nail on the head in the last block of text: Keeping payloads to reasonable sizes (under 50 or 60 tons) vastly reduces the need for asparagus staging. And the smaller payloads are much MUCH easier to engineer when the game actually gives you the information you need (See KER, VOID, MechJeb)!!!!

[eataTREE]

eataTREE's rule of asparagus staging:

Any gains in delta-V from complex radial staging are insufficient to make up for the loss of time involved in having to revert and fiddle with the Separatrons fifty-three times before every stage separates cleanly.

[Franklin]

eataTREE's rule of asparagus staging:

Any gains in delta-V from complex radial staging are insufficient to make up for the loss of time involved in having to revert and fiddle with the Separatrons fifty-three times before every stage separates cleanly.

http://forum.kerbalspaceprogram.com/threads/61238

I hope for a day when aero's revamped and pancake asparagus rockets aren't optimal, but for now they are.

[Doc]

Yeah, I would say the most we'll see IRL is crossfeed to the center stack (planned for Falcon Heavy, possibly others). Beyond that the complexity starts getting out of hand.

Even the center-crossfeed option is risky, since those decouplers need to be *dry* or you risk a catastrophic detonation. IMHO a modification of a Delta-IV type would be 'better'. On take off, run all three boosters 'hard', then scale back the center engine stage to keep an optimal thrust profile. Use the 'slow burn' time on the main stage to refuel from the two boosters. You can significantly reduce the fuel crossfeed rate, and you can 'shutdown' the transfer much earlier, as well as balancing across the load.

I'm just not that convinced that it's worth it, from a dollars and cents point. The Saturn F1 engine *cluster* was under 0.2% of the fully fueled weight of the 1st stage, and around 30% of the *dry* weight of the stage - that's not that terrible. It also represented 5% of the total mass lofted to stage 1 separation.

You can do something similar with the J-2 engine cluster for stage II. The J-2 cluster weighed ~9000 lbs, out of the total 1.500.000 for everything after stage I. That means that the J-2 cluster represented about 0.67% of the fuel to get up to stage I separation, or, about 32,000 lbs of fuel/oxidizer.

LOX is cheap, and RP-1 is ~$5 a pound - running a back-of-the-napkin of 1:2.5 fuel/ox ratio, it works out to about 10,000 lb of RP-1 , or about $50,000 in fuel costs to lift the fricking Saturn V 2nd stage engine cluster.

The reason we used fuel crossfeeding on the Space Shuttle was to recover the engines & pumps intact - they are the expensive part of the rocket. I am betting that this is part of the reason they are doing cross-feeding on the 9H as well, so that they can simplify the recovery process on the 9H main stage (ie, recover the center stage and get the biggest part of the launch cost back).

Unrelated: Look up youtube videos for 'Delta IV on fire' for some bizzare 'real world' weirdness with rocket launches. There is a great one of a Delta-IV flying, with the entire lower rocket structure on fire. A good example of why you don't want to drip any fuel from your cross-feed connections.

[Red Iron Crown]

Those are good points Doc, I guess we'll see whether SpaceX can make the crossfeed economically viable.

Not quite related, but we need a name for this brilliant bit of single engine staging (not mine, but possibly the most kerbal thing I've ever seen):

If you look closely, the tanks are emptied bottom to top. As each one is emptied, sepatrons are fired to destroy the tank and bring the engine up for docking to the next tank. Increased efficiency through more explosions!

[magnemoe]

One reason asparagus isn't used in 'real' rockets is this:

The fuel flow need to run say, a Saturn F1 engine cluster is about 1/6th the flow rate needed to drive one of the main turbines at hoover dam - a *tremendous* amount of liquid, running in 2 distinct systems.

Draining 2+ separate tanks, and keeping everything nice and balanced, at that depletion rate requires some really good (read - large and expensive) pumps and valving solutions. Further, the rocket needs to be able to switch between 'booster' and 'main stage' fuel sources without inducing so much as a bubble or 'boom' goes the rocket.

The added cost in $$$, extra weight, and increased risk has made the gains just not worth it. Rocket fuel is pricey, but not *that* bad, compared to the cost of say, an engine (the biggest cost of which is the pumping machinery).

It's the same reason we don't use variable-bell nozzles on all our rockets; yes, they are 'better' in theory, but the cost of implementing them is almost as high as what you will save.

Yes, also as the TWR of real life engines and the dry mass of tanks are less its far less benefit. Saw some data on falcon heavy and the benefit was far less than in KSP.

Falcon heavy is especially suited for crossfeed as boosters and core is the same design.

You could modify the Soyuz rocket with it four liquid fuel boosters but this would require redesigning the tanks with larger boosters and smaller core.

And you would not use the 2+2+1 asparagus we use but rater drain two boosters first and then to second pair.

Last most uses solid fuel boosters, unlike in KSP they can have variable trust as you design the hole in the core to give a trust profile you want: kick at start then slowly reducing trust to keep TWR within limits.

[Drethon]

Even the center-crossfeed option is risky, since those decouplers need to be *dry* or you risk a catastrophic detonation. IMHO a modification of a Delta-IV type would be 'better'. On take off, run all three boosters 'hard', then scale back the center engine stage to keep an optimal thrust profile. Use the 'slow burn' time on the main stage to refuel from the two boosters. You can significantly reduce the fuel crossfeed rate, and you can 'shutdown' the transfer much earlier, as well as balancing across the load.

This is something I've considered with SRBs now that they are cheaper but just haven't tried it yet. Setup an asparagus style cluster of 6 SRBs around a central liquid fuel booster. Now vary the thrust limit of the SRBs (I think this can be done by placing them in pairs rather than 6 at once) so that one pair is 100%, one pair is 66% and one pair is 33%. Now you can use 100% from the central booster at launch and all 6 SRBs boost at launch to get up to terminal velocity. Throttle back the central as appropriate and drop off the SRBs as they dry up.

I suspect my example requires fine tuning of the thrust (either more use of SRBs at lower altitude or possibly even less use of the second and third pair to avoid excess thrust) but it seems like this could provide efficient boosting without dealing with the fuel transfer.

Just my bent $0.02.

[Doc]

This is something I've considered with SRBs now that they are cheaper but just haven't tried it yet. Setup an asparagus style cluster of 6 SRBs around a central liquid fuel booster. Now vary the thrust limit of the SRBs (I think this can be done by placing them in pairs rather than 6 at once) so that one pair is 100%, one pair is 66% and one pair is 33%. Now you can use 100% from the central booster at launch and all 6 SRBs boost at launch to get up to terminal velocity. Throttle back the central as appropriate and drop off the SRBs as they dry up.

I suspect my example requires fine tuning of the thrust (either more use of SRBs at lower altitude or possibly even less use of the second and third pair to avoid excess thrust) but it seems like this could provide efficient boosting without dealing with the fuel transfer.

Just my bent $0.02.

I've built a few rockets that used multiple SRB thrust limits like that, it's kinda fussy getting them all just right, but it works out pretty well. I mostly use them to deal with FAR + BTSM or FAR + early career, when 'too much thrust' is a *very* bad thing....

[LordFjord]

Those are good points Doc, I guess we'll see whether SpaceX can make the crossfeed economically viable.

Not quite related, but we need a name for this brilliant bit of single engine staging (not mine, but possibly the most kerbal thing I've ever seen):

http://giant.gfycat.com/ConfusedVibrantAndalusianhorse.gif

If you look closely, the tanks are emptied bottom to top. As each one is emptied, sepatrons are fired to destroy the tank and bring the engine up for docking to the next tank. Increased efficiency through more explosions!

I am speechless. Whoever came up with that is a genius. Who made that? I'd say honor the inventor by naming it <Name>-Staging.

[Red Iron Crown]

I am speechless. Whoever came up with that is a genius. Who made that? I'd say honor the inventor by naming it <Name>-Staging.

No idea. Someone linked it in IRC but they didn't know who made it either.

[magnemoe]

I am speechless. Whoever came up with that is a genius. Who made that? I'd say honor the inventor by naming it <Name>-Staging.

Yes, its insane, however he would get the same result with radial drop tanks.

I have used docked boosters from time to time, they are very useful if you have heavy ships, can refuel in orbit who have a large docking port in bottom and radial LV-N, use the booster engine to increase TWR, you want to reduce trust of it over time to improve ISP, if it have its own probe core and fuel tank you can even disconnect in orbit and land it for recovery.

[Franklin]

Those are good points Doc, I guess we'll see whether SpaceX can make the crossfeed economically viable.

Not quite related, but we need a name for this brilliant bit of single engine staging (not mine, but possibly the most kerbal thing I've ever seen):

http://giant.gfycat.com/ConfusedVibrantAndalusianhorse.gif

If you look closely, the tanks are emptied bottom to top. As each one is emptied, sepatrons are fired to destroy the tank and bring the engine up for docking to the next tank. Increased efficiency through more explosions!

Holy. I've done this by accident a few times and thought it was funny, but never did I think, "Let's see if I can do this on purpose, it seems economical."

[Mtshaw113]

So In the 2nd picture of that one post, it has the orange tank center stage with the tail connectors, for some reason the engines on mine wont start up

[Teutooni]

If asparagus staging is more mass-efficient than stacked or parallel, and twisted candle even more so than asparagus - how efficient is this:

?

[Red Iron Crown]

If asparagus staging is more mass-efficient than stacked or parallel, and twisted candle even more so than asparagus - how efficient is this:

?

That's asparagus in an odd shape.