Kerbal Space Shuttle and rocket Economics

[The_Rocketeer]

The most efficient space shuttle design would have been Buran-style, with the engines under the main fuel tank, and dedicated on-orbit engines making good use of onboard fuel.

Well that depends what you mean by efficient. You'd basically lose your ascent engines every launch when the ET decoupled. They have some big price tags, I understand.

[Highguard]

Pure rocket SSTOs are easy to build than spaceplanes and much better able to handle wide or awkward payloads.

"Shuttle" style suffers from the aerodynamic limitations of spaceplanes and the fuel-hungry limitations of rockets. Either SSTO approach will always win, if you can get anywhere near KSC without wings and wheels.

[KerikBalm]

I'd like your comment on the shuttle I posted as it seems to deliver 42t to a 200 x 200 orbit just fine, recovering all but the ET. Though I'm not sure that SSME stand in is a stock engine.

It looks cool, but its not "close enough" functionally for what I'm looking for:

#1) That is a pre-1.0 video, where all engines have constant thrust whether at sea level or vacuum

#2) Those are gimballing adjustable thrust liquid fuel boosters, not SRBs. While the real shuttle had thrust vectoring on the SRBs, we don't have that option. I think to stay true to the shuttle challenge, you shouldn't be able to adjust the throttle.

Using LFBs also gets you much better Isp and performance.

#3) He doesn't recover the LFBs. He switches to them and deploys the parachute, but they will still despawn in the atmosphere, due to a gameplay limitation.

They actually despawn in the video if you pay attention.

This will severely impact the cost efficiency of this, as those LFBs are pretty expensive relative to SRBs.

As far as I can tell, the cost of those non-recoverable boosters is:

25,000 for the KR-2L, 2,500 for the adaptor, 3,000x4 for the LF tanks, 800 for the 2.5 to 1.25m adaptor + some chutes and probably sepratrons.

Over 40,000 for each booster.... gone

Also, I don't like part clipping, which his designs use (to each their own though)

I see on his thread, he updated it for 1.0 aero... he uses mammoth engines now, thats 39k just for the engine, and you can't really recover it.

I'm sure I could make a bigger shuttle. His version uses >6x the amount of LFO that mine uses, bit it gets a bigger payload to a higher orbit,

Neither one is going to be very economical

Edited September 14, 2015 by KerikBalm

[mikegarrison]

Air-breathing spaceplanes in KSP are way, way too optimistic. Partly it is because of the tiny Kerbin scale, and partly it is because of the completely unrealistic engines.

[GoSlash27]

Technically you don't even need to incline your engines if you balance the center of thrust by adding liquid fuel boosters connected under the external tank. Once the external tank dries up, dump it and the balancing LFBs. Maybe that would work? Anyway I've seen some shuttles that lift oranges and more to orbit on Youtube. Once again I haven't had the chance to try this so take this for what its worth, but maybe your shuttle building isn't as refined as your SSTO building?

MoreBoosters,

If I were to design an economically- viable STS style lifter in KSP... I wouldn't. STS- style launchers just plain aren't worth it for the economics and DV budget of Kerbin.

Yeah, I realize that I could build something that's not much like the actual STS and get better results. That's why I build SSTO spaceplanes instead. I chose to build a design that's as faithful as I can get it to the actual STS using all stock parts.

I *could* attach a booster to the tank, or go Ariane/ Dyna-Soar or a tank in the nose or sandwich or all sorts of other options. In fact, I've played with all of them. But those aren't faithful STS designs.

It might very well be that my STS designs "aren't as refined" as my SSTOs, but that ain't why you won't find vertical lift reusables that exceed 40% payload fraction. Nobody is making shuttles that can do that. Or for that matter, shuttles that can match the cost- effectiveness of disposables.

You point out someone who's made an economical STS in KSP, and then restate your case.

Best,

-Slashy

[Geher]

You point out someone who's made an economical STS in KSP, and then restate your case.

Well, would you still call it a STS, when it has a heatshield, a probe core, some rcs and lot's of parachutes on the external tank?

[The_Rocketeer]

Well, would you still call it a STS, when it has a heatshield, a probe core, some rcs and lot's of parachutes on the external tank?

Is this a joke? I want to laugh but it might come off as rude.

[More Boosters]

MoreBoosters,

If I were to design an economically- viable STS style lifter in KSP... I wouldn't. STS- style launchers just plain aren't worth it for the economics and DV budget of Kerbin.

Yeah, I realize that I could build something that's not much like the actual STS and get better results. That's why I build SSTO spaceplanes instead. I chose to build a design that's as faithful as I can get it to the actual STS using all stock parts.

I *could* attach a booster to the tank, or go Ariane/ Dyna-Soar or a tank in the nose or sandwich or all sorts of other options. In fact, I've played with all of them. But those aren't faithful STS designs.

It might very well be that my STS designs "aren't as refined" as my SSTOs, but that ain't why you won't find vertical lift reusables that exceed 40% payload fraction. Nobody is making shuttles that can do that. Or for that matter, shuttles that can match the cost- effectiveness of disposables.

You point out someone who's made an economical STS in KSP, and then restate your case.

Best,

-Slashy

It seems the main problem is the anemic recovery system, necessitating the use of mods.

You (and KerikBalm) came up with the whole "honest to STS" thing,the thing was hardly efficient in real life and there's no way it can compete in a toy planet where SSTOs are possible without super future tech. The main idea always was recoverable payload fairing w/ engines launched vertically.

As for people making an economical STS in stock KSP, it may not be possible, but I'm not sure about mod usage. Even then, I still haven't had the chance to fool around with designs myself.

[Warzouz]

Does RSS changes something ?

[More Boosters]

Does RSS changes something ?

Well a user on the first page claimed that the only way to get a spaceplane to space is a shuttle-like assembly there.

[Nansuchao]

In stock Kerbin you can reach orbit easily with a spaveplane, orbital speed 2.230 m/s more or less. Try to imagine doing the same thing in a planet 10 times Kerbin, with 130 km of atmosphere, with an orbital speed of 8.000 m/s more or less. How many spaceplanes you can build with more than 10 k of Delta/V? What payload can they bring to orbit?

And, due to the fact that in RSS you have DRE, how easy it will be to reenter the atmosphere at more than 8.000 m/s?

In that case, a shuttle style vehicle is more useful, if you can manage to recover some empty stages.

[GoSlash27]

It seems the main problem is the anemic recovery system, necessitating the use of mods.

You (and KerikBalm) came up with the whole "honest to STS" thing,the thing was hardly efficient in real life and there's no way it can compete in a toy planet where SSTOs are possible without super future tech. The main idea always was recoverable payload fairing w/ engines launched vertically.

As for people making an economical STS in stock KSP, it may not be possible, but I'm not sure about mod usage. Even then, I still haven't had the chance to fool around with designs myself.

More Boosters,

The main problem (at least in stock) is that the orbiter itself is dead weight and weighs more than the payload. You have to build a stack that will lift all of it into orbit instead of just the payload itself. That's a whole lot more fuel, structure, and boosters than you need for the job and when it's all said and done you've spent more than the cost of 3 engines in order to save them.

I'm sure RSS changes the situation because SSTO spaceplanes are no longer viable, but even then I doubt an STS would be worth it. It was never competitive with disposables in real life.

Best,

-Slashy

[More Boosters]

It was never competitive with disposables in real life.

Best,

-Slashy

But the point of this thread is to see how effective it would be if what made it non-competitive, the high maintenance costs, didn't exist. Now it's probably not competitive at stock KSP, but we still haven't heard details about RSS. Assuming RSS doesn't toy with the recovery system and you still get a full refund for a successful landing on the runway, it may be sensible there.

[GoSlash27]

But the point of this thread is to see how effective it would be if what made it non-competitive, the high maintenance costs, didn't exist. Now it's probably not competitive at stock KSP, but we still haven't heard details about RSS. Assuming RSS doesn't toy with the recovery system and you still get a full refund for a successful landing on the runway, it may be sensible there.

More Boosters,

I haven't played with RSS; I run 100% stock. But I doubt it could ever break even with a disposable due to KSP's cost balancing. Engines are too cheap and fuel is too expensive.

Say your shuttle weighs 3 times the payload it carries. That means you need 4 times as much fuel and structure, 4x the boosters, and 4x more powerful main engines vs. a simple disposable. The whole thing exists in order to recover the cost of the engines. Expending 1/4 the fuel, boosters and structure and just wasting the weaker engines is cheaper overall.

Best,

-Slashy

[KerikBalm]

It seems the main problem is the anemic recovery system, necessitating the use of mods.

Well, the boosters aren't that expensive, but yea...

You (and KerikBalm) came up with the whole "honest to STS" thing,the thing was hardly efficient in real life and there's no way it can compete in a toy planet where SSTOs are possible without super future tech. The main idea always was recoverable payload fairing w/ engines launched vertically.

As for people making an economical STS in stock KSP, it may not be possible, but I'm not sure about mod usage. Even then, I still haven't had the chance to fool around with designs myself.

Well, many of the economic issues present in real life are not present in KSP, as stated in the first post.

Also, I'm not trying to compare it to SSTOs.

I'm wondering if it competes with disposable rockets... along the lines of Red's post:

Recovers the pricey rocket engine, all it discards are the SRBs and fuel tanks. Still, it is not the most economically viable, for that I would order them so:

1. Airbreathing recoverable SSTO (fuel cost only)

2. Pure rocket SSTO (fuel cost only, but higher than airbreathing)

3. Shuttle style (fuel cost + tanks and SRBs)

4. Rocket with disposable stages (fuel cost + tanks + engines, etc)

-fwiw, in my case, the orbiter itself does have some LFO storage in the form of the mk3 to 3.75m adaptor, thus some tankage is recovered (also, the OMS tanks).

Anyway, I question if 3 is really cheaper than 4.

I know it will not beat 1 or 2, thats not the question.

So in the case of my design, I calculated roughly 1,500 per ton to orbit.

http://forum.kerbalspaceprogram.com/threads/134299-Launch-Vehicle-Optimization-Test-Results

That guy is getting between 1200 to 1500 funds per ton to orbit with a 2 stage disposable system.

I *might* be able to significantly improve my cost per ton to orbit with the STS style... perhaps down to 1,000, if I increase the fuel tankage of the orbiter I could save some funds that way (I'm even envisioning recovering 1 of the two orange tanks inside the mk3 bay after the payload is released).

The SRBs will cost 10,800, of the >26,000 launch cost, and I don't think I can recover that cost.

Recovering the orange tanks would save over 5000, but I think I'll only recover/eliminate 1 of them max.

-Nothing to be done about fuel cost.

[GoSlash27]

KerikBalm,

You actually spent $27,000 to place 50t in orbit. 18t of that was payload.

Had you treated the entire thing as payload and sacrificed the engine, it would've dropped your launch cost to $1,000/ tonne.

But of course you wouldn't have needed such a big ungainly stack or a Rhino just for an 18t payload, so your final cost would've been much lower.

I think there's also some confusion about part recovery and stock KSP. You're not going to recover the boosters without mods and I don't know if you're counting the tank as recovered.

If you're using a mod to recover parts for your shuttle, there's no reason why you can't also recover spent stages on a traditional lifter as well.

I did figure out where the discrepancy was between our designs, though:

-Your shuttle, stack, and therefore payload are twice the mass of mine.

-You're calculating the dropped bits as recovered and I'm not.

Best,

-Slashy

[Nansuchao]

You can have some useful shuttle style vehicle in RSS, but not following the STS way to launch. The advantage of shuttles, is that they can deliver payloads and crew together, and with KIS/KAS this becomes interesting. With the Recovery Stage mod then, you loose just the fuel.

[Red Iron Crown]

The reason shuttles can't keep up with disposables is the tyranny of the rocket equation. Most of what's placed in orbit is not payload, so the stack is huge in comparison. All the extra fuel and oxidizer erases any savings you get from returning the engines.

I think of the tyranny of the rocket equation as the exponential growth of propellant mass with increasing dV requirements. Changing payload mass has a linear relationship with total propellant mass for a given dV requirement, so it's not quite as bad. For that matter it's not as if the disposable staged rocket puts only payload in orbit, the circularization stage must enter orbit as well. The penalty is the additional mass from a larger engine and the recovery equipment vs a disposable circularization stage, which I'm pretty sure will be significantly less than the payload mass.

Anyway, I question if 3 is really cheaper than 4.

I know it will not beat 1 or 2, thats not the question.

That was very much a from-the-hip estimate. I should try to further optimize a shuttle design and see how low I can get the costs.

[Captain H@dock]

The advantage of shuttles, is that they can deliver payloads and crew together

I don't see how that's different from any of the others 3 design families listed by Red Iron Crown...

As long as your payload/part-that-reaches-orbit has a crew compartment, you get this no matter what the vehicle lifting it to orbit is...

And as far as crew compartment parts go, cockpits aren't the most weight effective ones*. So that's yet another feature of the STS that is not helping with efficiency.

* in order: mk3 crew (400kg per seat), mk2 crew (500kg per seat), mk1 can (860kg per seat). 1.1 might change that a bit with the mk1 crew cabin. I think all cockpits are above 1Ton per seat.

[The_Rocketeer]

Duh. Ignore this post.

Actually, I found a better use for it.

Can anyone offer an actual costed design for a disposable rocket with identical or better mission capability/specifications to an STS replica?

By this I mean:

1. Reach orbit in 2 stages or less (counting SRBs as a stage).

2. Large cargo bay for deployment AND recovery of modules to/from orbit.

3. Space for at least 4 kerbanauts.

4. Ability to land safely with or without payload.

5. While doing both of 4, ability to land consistently on the runway at KSC without sick pilot skills (gliding in isn't THAT hard).

6. Reach/exceed STS max circular orbit altitude (Slashy, give us a figure for this!)

7. Carry equivalent payload (dimensions and mass).

8. Rendezvous and dock (with suitable cargo-bay module for docking) with a station.

9. Possibly others (as suggested).

It seems to me that unless a rocket can outperform the STS at its own game, this conversation is moot. Perhaps the STS was/is the best way to do what it can do without going to the expense of developing a new vehicle for every new mission.

Edited September 15, 2015 by The_Rocketeer

[Red Iron Crown]

No, but don't forget any increase in propellant mass causes a sympathetic increase in dV requirement anyway (for a given payload).

It does not. The dV to orbit does not change for a rocket of given efficiency (similar TWR, staging, Isp, aero), no matter how large or small it is. That's how payload fractions work.

A simple example: Imagine you have a rocket that can put 5t into LKO. Strap two of those rockets together and it will put 10t into LKO.

[The_Rocketeer]

It does not. The dV to orbit does not change for a rocket of given efficiency (similar TWR, staging, Isp, aero), no matter how large or small it is. That's how payload fractions work.

A simple example: Imagine you have a rocket that can put 5t into LKO. Strap two of those rockets together and it will put 10t into LKO.

I think you misunderstood me - I was talking about increase in dV costs for a more massive rocket with additional fuel but the same engine thrust, not doubling everything - but in any case I realised you were discussing a scaling payload mass rather than a fixed one, hence retracted post.

[Warzouz]

I think you misunderstood me - I was talking about increase in dV costs for a more massive rocket with additional fuel but the same engine thrust, not doubling everything - but in any case I realised you were discussing a scaling payload mass rather than a fixed one, hence retracted post.

Yes I noticed that fuel fraction increases with payload increase. BUT, dry mass fraction is reduced. Adding that, the price per ton goes (slighlty) down as payload mass increases. In the end mass efficiency increases with high payload.

That's what I've noticed with my Cygnus SSTO rockets. Check my table I posted before.

[KerikBalm]

KerikBalm,

You actually spent $27,000 to place 50t in orbit. 18t of that was payload.

Had you treated the entire thing as payload and sacrificed the engine, it would've dropped your launch cost to $1,000/ tonne.

Had I sacrificed the engine, it would have just dropped back down after the SRBs burned out.

Or, perhaps you mean I don't recover the engine... now I need to add that KR-2L's cost to the cost of the launch:

25+27k = 52k for the launch now.

I would also need to add the cost of the internal LFO tank on the orbiter that is used to orbit.

But of course you wouldn't have needed such a big ungainly stack or a Rhino just for an 18t payload, so your final cost would've been much lower.

Yea, that's my feeling.

I think there's also some confusion about part recovery and stock KSP. You're not going to recover the boosters without mods and I don't know if you're counting the tank as recovered.

If you're using a mod to recover parts for your shuttle, there's no reason why you can't also recover spent stages on a traditional lifter as well.

I was not counting the external tank, nor the SRBs as being recovered.

I just looked at the cost in the VAB, and then took off the external tank and its attacked SRBs, and subtracted that cost from the previous cost to get the launch cost.

I did figure out where the discrepancy was between our designs, though:

-Your shuttle, stack, and therefore payload are twice the mass of mine.

-You're calculating the dropped bits as recovered and I'm not.

I'm not calculating the dropped bits as recovered.

I'm not sure my stack is twice the mass of yours.

I use twice the SRBs, yes + 48 tons

- but 2 orange tanks are the same as one large kerbodyne tank, and your external tank is even bigger than that

Based on the picture I see:

1x Kerbodyne S3-14400 Tank -> even mass with my external tank

1x Kerbodyne S3-7200 Tank: -40 tons

1x mk3->2.5m adaptor: -14.3 tons

1x C7 Brand Adapter - 2.5m to 1.25m: -4.5 tons

3x skippers are the same mass as 1x KR-2L, so that is even.

As I look at it, my design should be ~11 tons lighter than yours.

Did you get my design confused with the one from that video of a STS with liquid fuel boosters (that seemed to count the boosters as recoverable)?

Edited September 15, 2015 by KerikBalm

[Gargamel]

Statistically, it was the most dangerous vehicle ever flown into space,

This I think I disagree with. It may be true, but we cannot know as the rest of the other systems never flew as many missions as the STS did (Soyuz maybe?). Apollo had one fatal accident and one near miss, out of what, 11 manned missions (plus SkyLabs)? Mercury had 6 manned, Gemini 10 (correct me if those are wrong). Most of the programs never flew enough missions to get a good enough sample size of reliability. I'm not saying your wrong, just that we can never know just how reliable those other programs really were. The STS didn't have it's first fatal incident until STS-51, it's 25th manned flight, and then STS-107 (113th manned).