Asparagus staging overrated?

[OhioBob]

I'm fairly new to the forum, but from the comments I've read so far, it seems that asparagus staging is all the rage. I'm not sold on it. Although it is highly efficient, it is also very expensive. Although this may not be a concern in sandbox mode, it is in career mode. The most important consideration in career mode should be cost per ton of payload. In this regard I think solid rocket boosters are a better option. An SRB delivers about four times as much impulse per unit cost versus a liquid stage of comparable size.

For example, let's consider a stock vehicle that comes with the game – the Kerbal X. This is an all-liquid launcher using asparagus staging that includes a 16.12 t payload.

For comparison, I started with the sample basic vehicle and kept its payload and first stage intact. I stripped away the six liquid strap-ons and replaced them with three S1 SRB-KD25k solid strap-ons. I also added a second stage (I consider the existing upper stage part of the payload) consisting of a Rockomax X200-16 fuel tank and a Skipper engine. I also replaced the tail winglets with four Vernon engines to give me better control. Just to give it a name, let's call this revised version the Kerbal X-A.

Both of these launchers have almost exactly the same lift capacity and performed equally well in delivering the payload to orbit. As expected, the Kerbal X has a higher payload fraction – 0.123 versus 0.103 for the Kerbal X-A. However, the total cost of the Kerbal X is 72,420 while the Kerbal X-A costs only 60,116. That's $4,493 per ton of payload versus $3,729 per ton for the X and X-A respectively. Using cheap solids reduced the cost by 17% compared to asparagus staging.

I understand that there may be situations, say for a very large super-heavy launch vehicle, where solids may not be practical. However, from a cost perspective, solids seem to be the way to go in situations where they can be effectively used.

Please enlighten me if there's something I'm not considering here.

Edited November 13, 2014 by OhioBob
changed to "Answered"

[Pecan]

...it seems that asparagus staging is all the rage. I'm not sold on it. Although it is highly efficient, it is also very expensive. Although this may not be a concern in sandbox mode, it is in career mode. The most important consideration in career mode should be cost per ton of payload. In this regard I think solid rocket boosters are a better option...

1) Asparagus has always divided the KSP community because it is not practical in real-life. In KSP, however, it is the mass-optimal way to stage. If you aren't optimising for mass then don't use it.

2) SRBs are heavy and have almost no recoverable value when empty. They are of only of limited use as a first launch stage, liquid fuel (inc. nuclear) and ion engines beat them in all other cases.

3) Using a stock vehicle as an example of anything except how not to do things isn't a good idea - they are all designed to be "slightly wrong".

4) If you care about money you should work on, in order of difficulty and cost-effectiveness; i) reusable SSTO rockets, ii) VTVL jets, iii) spaceplanes. Jets are immensely more cost-effective than rockets in all cases. spaceplanes make it easier to land where you want (increasing recoverable value) than is generally possible with VTVL designs.

[Tokay Gris]

Please enlighten me if there's something I'm not considering here.

Lots of things.

First, this is not the real world.

Asparagus tends to be rather large in diameter. Which in the real world means higher drag. Quite a lot higher drag.

Which is not the case in stock KSP.

Second, Asparagus is expensive. Which does mean nothing in Science or Sandbox games. There the only thing that counts is "Can it be done?"

Since the contract-version of KSP came out, I changed my lifting vessels a lot. But I still have no idea how to lift my biggies to space cost-wise.

I had an asparagus lifter that was able to get 200 tons to orbit. With cost effectiveness I can get 100 tons to orbit for just above 60.000.

My two cents: Asparagus is good, until you need to figure in cost. Which just a few months ago, we didn't need to. Just build it, get it up and thats it.

Now, we need to keep in mind how much that costs.

Thats the whole difference as I see it.

[The_Rocketeer]

Well, for starters the Kerbal X is notorious for being an example of how asparagus staging should NOT be done. Like most stock craft, it's a deliberately bad demonstration of how things can be done in KSP, but it's not an efficient craft and certainly isn't optimised for cost.

[DerekL1963]

However, from a cost perspective, solids seem to be the way to go in situations where they can be effectively used.

The problem is, because of their relatively short burn time - there's not many situations where they *can* be effectively used.

Also designs using solids have to navigate between two subtle performance robbing traps... The first is a T/W 'hole', where the vehicle doesn't have sufficient performance to accelerate against the existing drag when the solids burn out and vehicle actually slows down until it reaches the equilibrium point. The second is terminal velocity, where you have enough acceleration to exceed maximum velocity for a given altitude - but since that's physically impossible, the thrust of any liquid engines is thus wasted until the vehicle either slows down (due to the 'hole') or reaches an altitude where terminal velocity is above the current velocity. Avoiding these two traps is not impossible, but it's not easy either. (I've deleted the stock vehicles, so I don't know if your version of the Kerbal-X encounters either of these.)

Cost is only one of the many variables to optimize around. And frankly, unless you're flying a lot of very expensive off-contract missions, I haven't found cost to be a major limiting factor even in pure stock (I.E. no contract mods) career games.

[DrD]

It was more popular before they released some of the extra-large engines and tanks.

Before then the only way to get a heavy payload up was to use multiple rockomax parts.

Because the joints are somewhat delicate it's not practical to build multiple stages up, so one's forced to build sideways.

In that scenario you may as well use asparagus to get rid of as much mass as quickly as possible.

For smaller rockets I tend to use SRB also. You need to make sure TWR is not too high or you'll waste fuel fighting the atmosphere.

I use KER and if necessary tweak the thrust of the SRB in the VAB. Sometimes is easier to just stage my liquid fuel engine to fire after SRB's are done rather than at the same time.

[LethalDose]

I'm fairly new to the forum, but from the comments I've read so far, it seems that asparagus staging is all the rage.

Since you're new to the forums, I'm kind of surprised you got the impression that asparagus was "all the rage"... At least by the traditional meaning of the phrase. It actually can be somewhat rage-inducing on the forums.

Anyway, the reason I say I'm surprised is that, at least IMO, asparagus staging has become much less common than it used to be due to the addition of funds/budgets/contracts/etc. It used to be almost everyone used it, but while it may be mass efficient, isn't cost efficient at all. Before things cost money, mass efficiency to lift into LKO was the primary way people compared... I guess success? I feel like "skill" isn't the right term. Now, cost efficiency is paramount for many people playing in games where funds are somewhat limited. Again, IMO funding really wasn't much of a limitation in 0.24 but is much more so now with the difficulty sliders.

You've really figured all this out for yourself with your experiments above, I'm just providing a bit of history to put what you've seen into context. It's going to take some time for play styles to adjust and asparagus to recede from it's current prominence. And players who stick to sandbox and science mode are probably still going to use it since it's a very easy formula for getting a lot of mass to orbit, and it's definitely easier than designing efficient payloads and upper stages to get the same job done.

[OhioBob]

1) Asparagus has always divided the KSP community because it is not practical in real-life.

I agree. One of the reasons I've avoided asparagus staging is because I don't see any real-world analog to it. It looks to me like something that would be very difficult to engineer in real life.

[Pecan]

Asparagus tends to be rather large in diameter. Which in the real world means higher drag. Quite a lot higher drag.

Asparagus is no wider and creates no more drag - in stock or any other aerodynamic model - than any other radial-staging strategy, including radial SRBs. "Pancake" was always an argument used by the anti-asparagus lobby but just because someone can build bad designs doesn't mean they have to.

My major objection to SRBs is, in fact, that you need to strap so many of them onto anything but the smallest payloads in order to make much difference that you end-up with something aerodynamically and structurally far less sound than all but the worst excesses of bad asparagus designs.

Neither are SRBs particularly cost-effective - Tech-0 lets you SSTO and return with an LV- T30 engine and 12 FL-T200 fuel tanks (payload = command pod, girders and parachutes for recovery). Adding two RT-10 SRBs lets you remove 3 fuel tanks and save 675 funds in construction. This is heavier than the 'pure' T30 design though and recovery value is so much less that per-launch costs are 246 funds more. In other words, it's only cheaper if you're throwing it all away anyway, not if you are aiming to recover it. Three RT-10s is worse (only removes one more tank) and 4 almost impossible to get into orbit. Staging - throwing away the empty SRBs - changes this and other SRBs have different performance but 'moar boosters' is still never a good automatic choice ^^.

ETA:

I agree. One of the reasons I've avoided asparagus staging is because I don't see any real-world analog to it. It looks to me like something that would be very difficult to engineer in real life.

*grin* In that case you will also want to avoid any other use of fuel-lines, rapier, LV-N and ion engines, any type of spaceplane (actually any SSTO), SAS, KSP's magnetic docking ports and probably 90% of the other parts too ^^.

Do you have a problem with radial staging? If so see ESA launch vehicles. If not, the only difference is that asparagus, on a core+2 (liquid fuel) booster design, just means adding fuel lines from the boosters to the core so a) the boosters can be used and jettisoned as early as possible, the core still has a full fuel load and can burn longer. While it's not technology that has been used yet on Earth, it's usually a mistake to limit KSP's engineering to our limited historical experience. I agree that multiple stages and, especially, layers of asparagus are almost certainly never going to happen but I'm at least as happy to accept the possibility of core+2 as I am LV-Ns, etc.

Edited November 12, 2014 by Pecan

[OhioBob]

Since you're new to the forums, I'm kind of surprised you got the impression that asparagus was "all the rage"... At least by the traditional meaning of the phrase. It actually can be somewhat rage-inducing on the forums.

I guess I haven't read enough of the forum. My initial impression came from another thread in which I got a few replies referring to asparagus staging as if it were the standard that every body uses. My apologies if that's not the case.

[Red Iron Crown]

Asparagus is the second most mass-efficient staging method and the most mass-efficient commonly used method (the more efficient twisted candle staging is trickier to set up well). As others note, it isn't overly cost-efficient.

Furthermore, it isn't overly player time efficient. Onion staging delivers nearly the same efficiency and is much easier to construct because full on symmetry can be used. I play mostly sandbox so cost isn't a factor for me, but even so I've switched to mostly onion staging as my play time is finite.

[LethalDose]

I guess I haven't read enough of the forum. My initial impression came from another thread in which I got a few replies referring to asparagus staging as if it were the standard that every body uses. My apologies if that's not the case.

No need for apologies! I've only been around for about a year and I still see stuff I don't understand because it pre-dates when I started following the forums and the game.

PS Welcome to the forums.

Asparagus is no wider and creates no more drag - in stock or any other aerodynamic model - than any other radial-staging strategy, including radial SRBs. "Pancake" was always an argument used by the anti-asparagus lobby but just because someone can build bad designs doesn't mean they have to.

My major objection to SRBs is, in fact, that you need to strap so many of them onto anything but the smallest payloads in order to make much difference that you end-up with something aerodynamically and structurally far less sound than all but the worst excesses of bad asparagus designs.

I respectfully disagree with this statement.

IME, it's very uncommon for rockets to need more than 2 SRBs to boost LFO cores into space, if those SRBs as their name implies: as boosters. A small number of SRBs (usually 2, sometimes 4) can be used very efficiently (in terms of drag, cost, and complexity) to increase TWR off the pad to get an initial boost of speed, which a central sustainer stage that has a lower TWR by itself pushes the upper stage to orbit. The SRBs are usually ditched early, but are still there in the thick lower part of the atmosphere.

In contrast, asparagus staging almost always use 4 or more radial cores (I'm not sure I'd call only 2 cores cross feeding the center Falcon Heavy-style as "asparagus staging", but definitions may vary), and typically 6 cores, and the cores tend to be around longer than SRBs. Both these issue contribute to more drag compared to properly used SRBs.

For what it's worth, I'm also a die-hard FAR fan, so this may also flavor my opinions/experiences. I also use DRE, and don't think recovering stages is typically worth the effort or $$$.

Edited November 12, 2014 by LethalDose

[r_rolo1]

Anyway, the reason I say I'm surprised is that, at least IMO, asparagus staging has become much less common than it used to be due to the addition of funds/budgets/contracts/etc. It used to be almost everyone used it, but while it may be mass efficient, isn't cost efficient at all. Before things cost money, mass efficiency to lift into LKO was the primary way people compared... I guess success? I feel like "skill" isn't the right term. Now, cost efficiency is paramount for many people playing in games where funds are somewhat limited. Again, IMO funding really wasn't much of a limitation in 0.24 but is much more so now with the difficulty sliders.

Well, we could say that the devs made sure that asparagus would not be cost efficient, and TBH they used some very crass methods to do so, like making decouplers costing the same as a ICBM capable rocket or a individual fuel line costing more than some rocket engines. In other words the OP, due to lack of historical context, saw the situation completely backwards ( it is not his fault, though ,like no one would blame people that saw the kind of ships we used in 0.16/0.17 to think that we were crazy to spam so much aerospikes ... ): asparagus is costly because the devs made it costly on propose , to make other aproaches less handicapped in comparison.

That said:

a) the Kerbal X is a badly designed ship, as others pointed, and it is like that on propose. Using the words of one of the workers of SQUAD, the stock ships are supposed to be "almost good" and all of them have a flaw of some kind ( to give beginners the chance of tinkering with it and make it better, like the OP did ). The Kerbal X flaw is bad management of the fuel in hand and some dubious staging, as the OP discovered

As DerekL1963 said, SRBs have their own trappings. Besides the ones (s)he mentions, the lack of variety of SRBs in game at this point is also a big issue, making that for some loads it is hard to get a satisfactory solution at a good price and it is not that difficult that a simple asparagus solution can be cheaper for some payloads than using SRBs ( it already happened to me ). More, unlike RL, where you can somewhat control the thrust per time profile of a SRB by fuel shaping, it is completely impossible to have a varying thrust SRB in game, that makes that a SRB solution will always be quite short of perfect in low atmosphere ( and , due to their specs in game, somewhat useless out of it ) ...

[RizzoTheRat]

Also being fairly new to the forum I'd got the impression that asparagus had fallen slightly out of favour since the introduction of career mode, however there are methods/mods to allow you to recover stages that parachute back down hence making it a lot more cost effective

Asparagus is the second most mass-efficient staging method and the most mass-efficient commonly used method (the more efficient twisted candle staging is trickier to set up well).

I had to look that up, wow, it's a thing of beauty. http://forum.kerbalspaceprogram.com/threads/55615-0-22-0-23-0-Payload-Fraction-Challenge?p=968265&viewfull=1#post968265

[LethalDose]

Well, we could say that the devs made sure that asparagus would not be cost efficient, and TBH they used some very crass methods to do so, like making decouplers costing the same as a ICBM capable rocket or a individual fuel line costing more than some rocket engines. In other words the OP, due to lack of historical context, saw the situation completely backwards ( it is not his fault, though ,like no one would blame people that saw the kind of ships we used in 0.16/0.17 to think that we were crazy to spam so much aerospikes ... ): asparagus is costly because the devs made it costly on propose , to make other aproaches less handicapped in comparison.

I agree that the costs of the decouplers (and, honestly, many many parts) are just screwed up currently, but I have to have faith that it will be fixed, by the devs or the community. But even if the situation were reversed to a more intuitive situation (cheap decouplers, expensive engines), I think the SRBs would still come out more cost efficient than the asparagus.

I'm also fine with asparagus being expensive since it's so easy and effective. /shrug. I think the game is in a better state with it being nerfed.

[MarvinKitFox]

blah blah blah blah...... Using cheap solids reduced the cost by 17% compared to asparagus staging.

Your statement is incorrect.

No, worse. It is FALSE!

You are comparing a low-tier, badly designed stock asparagus liquidfuelled rocket, with a very high-tier design using Solids that is optimised by yourself.

THEN you go and invent fallacious criteria to compare them!

Hint:

Anyone that is serious in career mode used recovery of launchers.

I can put that same 16.5 ton cargo of yours in orbit for 830 kerbucks, by using a jet-based VTVL rocket.

[r_rolo1]

@LethalDose

I'm also OK with asparagus being somewhat handicapped, but IMHO the way devs did it was ... unrefined, to say the least. I always thought that the best way of dealing with asparagus would be either part failure ( a thing that the devs strongly dislike, for reasons I'm not 100% in agreement with ) and/or limiting the fuel flow of fuel lines ( something that would probably need more code, but given that Harvester apparently spent some days working on the fuel lines some weeks ago, I was with some faith on that for 0.90 ). Those would be atleast realistic sounding reasons ... while decouplers and fuel lines cost costing fortunes probably does not.

But that is OT . On the topic, the cost and science structure of KSP in 0,24+ is strongly against asparagus staging, or broadly, against anything that uses extensively fuel lines, decouplers or even cubic struts. And it is like that by design, for balance reasons ...

[Pecan]

I respectfully disagree with this statement...

Actually, I don't think we'd disagree too far. My point was about the inadvisability of relying on an SRB lift stage - for additional short-lived, disposable, launch thrust they are certainly unequalled - but for anything more they really aren't worth it. On the other hand, in FAR you don't want more than about 1.2 TWR at launch, do you? If your core isn't supplying that on its own what do you do after ditching the SRBs? (Genuine question, I have too many memory problems to create another one with FAR so haven't tried it much).

FWIW I'm also fine with nerfing asparagus - I'm not in love with it any more than I am spaceplanes, ions or LV-Ns that are completely unrealistic compared to real-life performance. I engineer within the constraints of the system and tools available; that used to be asparagus for mass, now it's almost anything else for cost, no worries.

Asparagus is the second most mass-efficient staging method ...

Doh! Please kick me every time I forget twisted candle :-(

Edited November 12, 2014 by Pecan

[John FX]

I use clusters of SRBs. Small medium or large but with different thrusts so they burn out at different times, which levels out the TWR and means you don`t get a really short burn with too much TWR at the end but a much longer more controlled one that ends with you doing a much more appropriate speed for your height. If for example you had a central SRB with three pairs of two surrounding it then you could tweak your TWR to 2 by having the inner one at 100% then then next pair at 80%, the next 60% and the last 40% which would reduce your TWR at points in your launch and also extend your burn time. Each craft has different values. I try to get about 2500Dv so my Ap is well outside of atmosphere then circularise then.

I either jettison burnt out boosters in pairs when they empty (can cause explosions) or keep hold of them (wasting fuel) and jettison as a bunch but either way I have chutes attached to them so that StageRecovery can recover them.

If I jettison them in pairs then It is similar to saparagus although I don`t use that as I use FAR and DRE.

The cost difference between using them and other methods is not one I even notice in normal game play if I am honest. Jets just take ages (and are really hard in FAR) and liquid boosters have a high purchase cost IMHO

Horses for courses of courses.

EDIT :

For what it's worth, I'm also a die-hard FAR fan, so this may also flavor my opinions/experiences. I also use DRE, and don't think recovering stages is typically worth the effort or $$$.

I use DRE and FAR and also found it a pain to recover stages. Since I got Kerbal Construction Time, which has stage recovery built in, I just slap some radial chutes on anything I want to recover and it just happens (as long as it does not burn up within 2.5km)

The construction time added a lot more to my game than I thought it would too.

Edited November 12, 2014 by John FX

[OhioBob]

You are comparing a low-tier, badly designed stock asparagus liquidfuelled rocket

Since my experience with asparagus staging is practically nill, how was I supposed to know that that was such a bad design? I assumed, apparently incorrectly, that the stock versions were reasonable good designs.

with a very high-tier design using Solids that is optimised by yourself.

Actually I didn't optimize it. I forced myself to use the existing first stage from the Kerbal X. Had I optimized the design from scratch I would have likely downsized the propellant load of the first stage and transferred some of that to the second stage.

Your statement is incorrect.

No, worse. It is FALSE!

Can you give me examples of comparable launch vehicles that demonstrate the cost effectiveness of asparagus staging vs. SRBs?

Edited November 12, 2014 by OhioBob

[Pecan]

...The cost difference between using them and other methods is not one I even notice in normal game play if I am honest. Jets just take ages (and are really hard in FAR) and liquid boosters have a high purchase cost IMHO

Horses for courses of courses.

Now that's the sort of thing I'd expect to be really expensive so do you have some examples I can wander off to look at? LF engines have a high purchase cost but since you're using StageRecovery you'll get nearly all of that back. In contrast I wouldn't have thought empty SRBs were worth the effort. You say you haven't noticed but ... would you mind checking?

[Geschosskopf]

Please enlighten me if there's something I'm not considering here.

People use asparagus because their payloads are so heavy that they need multiple copies of their most powerful engines to get it off the ground. There are 2 main reasons for such heavy payloads. Either the people way over-build the payload, or they're just afraid of docking several smaller payloads together. You can do the entire tech tree never having more than about 15 tons of payload and you can create massive sandbox empires never lifting more than about 30 tons at once (at least if you're not afraid of docking). Of course, some people just like to see what sort of monster they can get off the ground, and that's perfectly fine. Whatever makes them happy. But for the most part, asparagus is unnecessary.

[LethalDose]

... but IMHO the way devs did it was ... unrefined, to say the least.

/Fistbump. I would just call it "lazy".

Actually, I don't think we'd disagree too far. My point was about the inadvisability of relying on an SRB lift stage - for additional short-lived, disposable, launch thrust they are certainly unequalled - but for anything more they really aren't worth it. On the other hand, in FAR you don't want more than about 1.2 TWR at launch, do you? If your core isn't supplying that on its own what do you do after ditching the SRBs? (Genuine question, I have too many memory problems to create another one with FAR so haven't tried it much)

My LFO core lifting stages are often just sustainer stages, with relatively low TWR, at least for most of the ascent. In fact, I usually thrust limit the SRBs to give the lift stage a TWR of just below 1 when the main engine isn't lit, so I still control TWR during the ascent with the main throttle. The LFO core then just maintains the initial boost speed until higher up out of the atmosphere, where I get a much better TWR since most of the fuel is burnt up and I'm losing less energy to atmospheric drag. The SRBs just give me the extra speed and dV I need to get the AP up to ~ 70km, where I want to spend my dV.

Another reason I don't prefer asparagus is that I usually launch very small payloads, at least compared to what I see a lot of other people launching. I'd rather reduce the upper stage/payload as far as possible, instead of building up the lifting stage.

I'll agree that it's not very effective to use solid rocket-only lifters/stages. My biggest issue with them is a lack of control. A BACC booster + 3 AV-R8 winglets costs as much and and lifts less than an LVT-45 + T800. I MAY see how that scales with a tri-coupler sometime...

Doh! Please kick me every time I forget twisted candle :-(

Everyone forgets twisted candle.

Edited November 12, 2014 by LethalDose

[Pecan]

Since my experience with asparagus staging is practically nill, how was I supposed to know that that was such a bad design? I assumed, apparently incorrectly, that the stock versions were reasonable good designs...

Not just you :-(

Squad have this view that they don't want to give newbies something so good they won't have to make their own designs. Nothing tells you that though so loads of people assume the stock ships are "the way" it should be done, struggle like mad to make them work and then complain that KSP is really hard. It annoys me that unless you read (or watch) a lot of tutorials and accidentally trip-over that nugget of information your early experience with KSP can be ruined, to the extent there may be no later experience with it.

Thank you for reminding me - I must add a note about that to my tutorial some time.

As to comparing asparagus and SRBs there aren't any that I can think of. Launch-vehicle design is something of a speciality as it is - most people don't bother optimising them - and there are/were a few distinct 'schools' amongst those who do anyway. As a very broad generalisation (which is therefore almost as wrong as otherwise):

1. "Not today, thank you" - I have a launcher that can put 100t into orbit. It's fine for this 1t satellite too.
   
2. "Moar boosters" - SRBs and lots of 'em, what problem?
   
3. "FAR" - single-stack 'til I die
   
4. "Realism" - if it hasn't been done on Earth it doesn't exist
   
5. "OCD" - asparagus is 0.01% more mass-efficient in this design
   

Basically, anyone who made launch-vehicles in one style rarely wanted to make them in another as well.

[sal_vager]

I'm not all that bothered about the arguments for or against asparagus, but after using it for so long it just gets old, and traditional staging just feels better