Why is my rocket so cost-inefficient?

[tsotha]

Right off the bat those x4s are horribly overpriced. I never use them for Kerbin launches unless I've got a test contract.

[tsotha]

The 48-7S and the Mainsail don't belong in there. If you consider vacuum Isp, the Pareto frontier is formed by the LV-N, the Aerospike, the KR-2L, and the O-10. By atmospheric Isp, the optimal engines are the Aerospike, the KS-25x4, the LFB, the KR-2L, and the O-10.

You can't judge an engine purely on ISP. The 48-7S is ridiculously light, so if you're building something small it's the only way to go.

And as others have pointed out, atmospheric ISP doesn't matter that much except on the first stage of an Eve return rocket. The aerospike isn't entirely useless, but it's mostly useless given its paltry thrust.

[Jouni]

You can't judge an engine purely on ISP. The 48-7S is ridiculously light, so if you're building something small it's the only way to go.

The 48-7S was the engine with the highest TWR until about 7 months ago. Now it's beaten by the KS-25x4 (higher TWR, higher Isp), the LFB (higher TWR, lower Isp), the KR-2L (significantly higher TWR, higher vacuum Isp, lower atmospheric Isp), and the O-10 (arbitrarily high TWR, lower Isp). At the same time, the ion engine, the Mainsail, the Skipper, and the Poodle have been buffed, replacing the 48-7S in many applications, where it once excelled. Now it's a reasonably good all-around engine like the LV-T30, with its small size as its main benefit. It's the obvious choice only in those relatively rare situations, where you need a high TWR in a small ship.

[capi3101]

Got some good news and bad news for you, OP. Good news is that you can adjust the rocket the way I suggested. The numbers (by which I mean KER) say you should make it. I did go ahead and replace the big RCS tank with a couple of roundified tanks (really, nobody ever truly needs the big tank except maybe for docking maneuvers) and that helped matters a bit. As much fuel as I was able to take out of the S3-7200 tanks, you might consider replacing them with S3-3600s. Seriously.

Now for the bad news: when I went to test fly the design, I noted the x4 in the center was dangerously overheated pretty much the entire way up. When the third stage boosters separated, the x4 immediately went critical and exploded. Why it did this I don't know; I'll have to investigate further. Might've just been a fluke.

[deepspacecreeper]

Okay...so we're trying to isolate the cause of the TWR bottleneck for 150 tonnes, right? That's easy - your core is too weak for the amount of fuel in it.

Temstar's Supernova - in fact his entire Zenith series - was based on a set of design formulas and philosophies which his booster set follows. Based on that philosophy, we'd design an asparagus staged rocket as follows:

1)150 tonnes is the desired payload. Assume 15% payload fraction - that means our final vehicle will be around 1000 tonnes (150/0.15 = 1000)

2)Shoot for a TWR between 1.6 and 1.7 at liftoff. So we'll need between 15,680 - 16,660 kN of thrust (1000*9.8*1.6=15680, 1000*9.8*1.7=16660).

3)Shoot for 22% of your launch thrust in your core and divvy the rest among the boosters. That works out to 3449.6-3665.2 kN in the core (just multiply the values time .22) and for three booster pairs, each booster should have between 2038.4-2165.8 kN ((15680*.78)/6 = 2038.4, ((16680*.78)/6=2165.8).

4)Subtract the mass of the engines you need and the mass of the payload from the estimated weight. The remainder is your estimated fuel mass.

So let's ignore the cost factor for a moment. Thrust wise, a KS-25x4 is what you should have in the core for the payload you want. For boosters, I'd recommend a single KR-2L set to about 87%. There's your thrust requirement. Now, the total mass of your engines with that configuration is 9.75+(6*6.5) = 48.75 tonnes. So our rough fuel mass should be 1000-150-48.75 = 801.25 tonnes. Divide that evenly among seven stacks; you get 114.46 tonnes each.

Looking at your design - you have an S3-14400 and an S3-7200 in each booster stage. I assume you've also got that in the core, plus another S3-7200. If they're full tanks, that's 123 tonnes of fuel in the boosters and 164 tonnes in the core. And if they're not full, the assumptions of the design might not hold because the mass ratio of the tanks you're using isn't 9:1 - it's 8.2:1. Your options are to change your tank setup - three orange tanks and an X200-16 will give you what you need - or to reduce the amount of fuel in the S3-7200s to about 90% (that should be, if I'm doing the math right, 2916 units of liquid fuel and 3564 units of oxidizer each) and take out that topmost S3-7200 in the core. This is assuming you swap out the dual Mainsails with a single KR-2L in any case (two mainsails are 12 tonnes, a single KR-2L is 6.5 tonnes, so the KR-2Ls fall in the "more efficient" category in this case).

I'll tell you what - I'll go ahead and replicate your booster with the tweaks I've proposed, to see if just adjusting the fuel will do the trick or not (I'm concerned about hitting an early delta-v bottleneck).

Bear in mind that Temstar's designs generally wind up with a 1.3 TWR in the core once all the boosters are separated. If you've got that, you're golden.

Now, SSTO transporter spaceplane might be an option as well. DocMoriarty has put out what I consider to be the definitive guide to transporter spaceplanes and his guidelines have proven themselves most useful in my own fights with flight. A good rule of thumb for a transporter spaceplane is a 25% payload fraction - so for a 150 tonne payload you'd want to shoot for a 600 tonne spaceplane. His guide says that the max take-off weight per RAPIER engine is 13 tonnes...so you'd need 47 engines. All I can say there is "good luck" if you wanted to try that route...

Thanks for the advice. My core stage only has 1 S3-7200 and -14400 each.

Got some good news and bad news for you, OP. Good news is that you can adjust the rocket the way I suggested. The numbers (by which I mean KER) say you should make it. I did go ahead and replace the big RCS tank with a couple of roundified tanks (really, nobody ever truly needs the big tank except maybe for docking maneuvers) and that helped matters a bit. As much fuel as I was able to take out of the S3-7200 tanks, you might consider replacing them with S3-3600s. Seriously.

Now for the bad news: when I went to test fly the design, I noted the x4 in the center was dangerously overheated pretty much the entire way up. When the third stage boosters separated, the x4 immediately went critical and exploded. Why it did this I don't know; I'll have to investigate further. Might've just been a fluke.

:0.0: (Both Bill and Bob come in and scream in agony as x4 engine explodes, and they reach for the abort switch, but stopped by Jeb )

Well maybe I can half the amount of RCS in the big tank by tweakables (i added RCS to compensate for sluggish steering). I think I should switch the mainsails for the KR-2L, like you said earlier. May solve the problem. When I have time I will try to modify my rocket again:

-Mainsail below boosters --> KR-2L

-Tweak fuel level to 90%

-Change S3-7200 in core 2nd stage to -3600 (again, note that I only have 1 -7200 and 1 -14400 in the core)

-Play around with the fuel levels further

-(seperatrons, tweak RCS level, struts and other small tweaks, only for final rocket)

[capi3101]

Got some good news for you:

So, pre-0.23.5 when the biggest tanks were the Jumbo64 and the strongest engine was the Mainsail, there was this bug in the game that would make a Mainsail explode if you placed it directly underneath an orange tank. It had something to do with the proximity of their respective centers of mass to one another and the way the game calculated heat dissipation from the engines. I had the thought that this could be what was happening with your booster, just scaled up to the next set of record-holders for large parts (ie. the S3-14400 and KS-25x4). The solution in the olden days was to set a smaller tank between the Mainsail and the J64, so I did an analogue - I put an S3-3600 between the S3-14400 and the 25x4, in the core, set up the rest of the booster like the last time, and flew it. The boosters got hot, but the core stayed cool all the way up. I reckon if you do the same with the boosters - put the big tank on top, the small tank in the middle and the engine on the bottom - the engines wouldn't overheat at all.

In any case, the redesigned booster makes orbit. Final problem remaining: I haven't launched this sucker yet where it didn't destroy the launch pad in the process. But I think that's my issue, not yours; I just wasn't careful to make sure the engine bells were sitting directly on the ground.

Final configuration. This is what made orbit. For the hell of it, I did go ahead and tweak the mass of the NREP testing weight upwards to see how far it'd go...

That I believe is the maximum practical payload mass for the revised booster design - 180 tonnes. At that point your margin of error for delta-V losses is a little narrow, and the core TWR is very low (though not yet below the 1.1 threshold). You'd have to fly it perfectly to make orbit with that much mass. One thing to note - with the 180 tonne shot, the boosters are still set to 87% thrust, so you could tweak them to full power and probably get the same TWRs you had at 150 tonnes.

So hopefully that addresses the issues you were having with your bottlenecks. We've knocked out a little bit of the cost in the process too...the booster is still expensive, but I think that's going to come with the territory of a massive payload launch like this one.

Best of luck in whatever it is you're doing.

Edited November 3, 2014 by capi3101

[Pecan]

Hmmm, this is the best I could put together for a 150t payload so far. I thought I'd be able to make it cheaper, but no luck so far. The core has 2 Mainsails clipped below the Kerbodyne fuel tank which is there to add width. Decouplers attach to that tank, boosters need strutting and the launch-clamps are neccessary to stop the launch-pad exploding.

832.9t including payload, cost 287,206. Launch TWR 1.47.

[Wanderfound]

Hmmm, this is the best I could put together for a 150t payload so far. I thought I'd be able to make it cheaper, but no luck so far. The core has 2 Mainsails clipped below the Kerbodyne fuel tank which is there to add width. Decouplers attach to that tank, boosters need strutting and the launch-clamps are neccessary to stop the launch-pad exploding.

http://i.imgur.com/J3wVGazl.png

832.9t including payload, cost 287,206. Launch TWR 1.47.

Swapping some tank capacity on the radial boosters for a ring of SRBs would probably bring the price down a fair way.

[capi3101]

Hmmm, this is the best I could put together for a 150t payload so far. I thought I'd be able to make it cheaper, but no luck so far. The core has 2 Mainsails clipped below the Kerbodyne fuel tank which is there to add width. Decouplers attach to that tank, boosters need strutting and the launch-clamps are neccessary to stop the launch-pad exploding.

http://i.imgur.com/J3wVGazl.png

832.9t including payload, cost 287,206. Launch TWR 1.47.

You guys have got to link me to these awesome mods that make your rockets look like pencil sketches...

I actually did the math for a booster like the one you designed, Pecan. Turns out that two J64s and an X200-32 are more expensive than the S3-14400/S3-3600 combo by about √2000. Over six stacks, that comes to √12,000. Still...that booster's a hundred large cheaper than mine. Definitely worth considering.

Might see what the SRBs can do. Just for the hell of it.

[Red Iron Crown]

You guys have got to link me to these awesome mods that make your rockets look like pencil sketches...

Kronal Vessel Viewer. Don't use the OP's links though, go here to download.

[Pecan]

@ Wanderfound: self-imposed rule of no more than one booster 'layer', so I'd have to stack SRBs as a first stage ^^.

@ capi: that mod's beautiful, isn't it. I only found out about it a couple of days ago too, RIC has given you the link. The S3 series tanks have a higher structural-mass per fuel-unit. Despite the fact that such figures are exactly what I'm working on at the moment I haven't worked out whether that extra mass offsets any cost advantage. I shall probably mess around some more this evening ...

[deepspacecreeper]

Got some good news for you:

Final configuration. This is what made orbit. For the hell of it, I did go ahead and tweak the mass of the NREP testing weight upwards to see how far it'd go...

According to this picture, capi3101 fixed the TWR bottleneck and designed a 150t launcher with a relatively low cost and the rocket does not hit the bottleneck and waste the rest of the delta-V margin. I use 4700m/s for each launch (not 4500 cause I get very bad results by hand) and according to the picture, the rocket has some safety margin for me to deorbit the core stage, and theoretically I think I can make it.

Not only thanks to capi3101, but also thanks to EVERYBODY who gave me suggestions and explained why my rocket was expensive and TWR bottlenecked. I learned a lot from all of you who posted in this thread. Thanks again and "fly safe". :cool::cool:

More effective is to use SRB in the start they are cheaper than liquid fuel and let you lift off even if core has an TWR less than 1.

Has an rocket with a 24x4 and four large SRB it can lift 40 ton to orbit

Thanks for the suggestion. made a 60t rocket with 8 srbs and costs 20000 less than a rocket I designed with liquid fuel boosters.

Then I realized when I decouple the boosters and fire sepratrons, the middle engine blows up.

Then...

So, pre-0.23.5 when the biggest tanks were the Jumbo64 and the strongest engine was the Mainsail, there was this bug in the game that would make a Mainsail explode if you placed it directly underneath an orange tank. It had something to do with the proximity of their respective centers of mass to one another and the way the game calculated heat dissipation from the engines. I had the thought that this could be what was happening with your booster, just scaled up to the next set of record-holders for large parts (ie. the S3-14400 and KS-25x4). The solution in the olden days was to set a smaller tank between the Mainsail and the J64, so I did an analogue - I put an S3-3600 between the S3-14400 and the 25x4, in the core, set up the rest of the booster like the last time, and flew it. The boosters got hot, but the core stayed cool all the way up. I reckon if you do the same with the boosters - put the big tank on top, the small tank in the middle and the engine on the bottom - the engines wouldn't overheat at all.

It worked! the sepratrons fired and the boosters seperated and the middle engine was intact.

Back to my 150t. I deleted the nosecones because of his suggestion.

Thanks for the link. After messing around with things a bit:

1) Temster's Supernova isn't *that* cheap. 393769 is better than your 547468, but the difference is rather less dramatic.

2) Deleting nosecones and adapters make for an easy way to save funds. 20900 in your case.

3) As others have mentioned, rockomax tanks have better mass ratios than kerbodyne tanks. The S3-14400 is the only one worth using, and that is because it is almost 21% cheaper than it "should" be.

4) Also as others have brought up, Mainsails are quite cheap for the thrust that they provide, while 4x KS-25s are rather expensive. There is a TWR difference, but I think it ends up not increasing the payload fraction enough.

4) You have too many engines for this. In particular, the vertical staging (keeping the KR-2L as dead weight for most of the ascent). The best thing would probably to use a KR-2L as a core stage with a "normal" amount of tankage (say, 1 S3-14400), and other engines around it. Above 3500 m, the KR-2L is likely to have better Isp than anything else you were considering. Also, getting rid of an entire engine plus switching to cheaper tankage should save you ~38400 funds

I also considered spaceplanes but I thought it would be more difficult to design and in some way not as reliable as traditional rockets. I only use SSTOs for lightweight probes and crew ferrying to/from space stations (like the one I made, link at signature).

Seriously is this a joke?

Nope, Ruedii.

Edited November 4, 2014 by deepspacecreeper

[MarvinKitFox]

... The entire rocket uses NASA parts....

Could anyone tell me why my design is so expensive?

Try fabricating your engine from something other than solid-gold.

the NASA engines are very very good, as long as price is not a factor.

[capi3101]

I'm glad to have been of assistance; good luck in your endeavors.