Help on lifter design

[luchelibre]

I've been searching for ideas on lifter design, and nearly everything I've come across is for heavy lifters (>50 tons).

I'm trying to build a line of standard lifters for 10, 20, and 30 ton payloads. But I must be overbuilding these things. Am I? Or does this look about right?

Here is my 20 ton lifter (The top tank fuel isn't burned and just gives weight.): 20 Ton Lifter

Surely I'm just being stupid and can't figure out a better way to make the lifter more efficient.

So far, my building rules are: stock parts, no asparagus staging, no jet engines, and ~4300m/s total delta-v in KER under atmospheric stats. Frankly, I've been trying to eliminate crossfeed altogether, but it's hard to do without my rockets getting out-of-hand size-wise.

If anything, I'd appreciate a pointer to a thread or page with some smaller lifter designs.

Edited July 28, 2013 by luchelibre

[SkyHook]

If your rules include no asparagus staging, then get rid of the outer stacks, and add SRBs instead. The current stacks that you have only add a massive weight penalty to the design, and probably don't even lift themselves. Happy launching

[luchelibre]

Each outer stacks have a TWR of 1.48.

I'll what I can do with boosters.

[calkurne]

why no asparagus?

having the outer fuel tanks feed directly into the center will give you >200m/s ÃŽâ€V (its not asparagus)

[espm400]

I know this one uses KW Rocketry, but I've built nearly identical stock craft with largely similar results.

It does use fuel crossfeed, but given that in a short time, so will real rockets, I don't think it breaks your 'no asparagus staging' rule.

As for stats; the lower two stages (core and boosters) add up to ~4600 m/s dV (unloaded), with that top stage adding another 3300 m/s dV (vacuum). It will easily lift 20-25 tonnes to anywhere in the Kerbin system and 30+ tonnes into LKO. As I said, it does use KW, but if you relpace the tanks with Jumbo-64s and the first and second stage engines with Mainsails, and the upper stage with a Skipper (or whatever you deem fit), the performance should be largely similar.

As for a thread with 'non-asparagus' designs, here you go.

Hope this helps.

[luchelibre]

why no asparagus?

having the outer fuel tanks feed directly into the center will give you >200m/s ÃŽâ€V (its not asparagus)

Because if I don't learn how to make efficient non-aspy rockets before realistic aerodynamics come along (or I install FAR first), then I'm just going to have to build all my rockets over again.

[Nurph]

I built a close copy of your lifter and flew it to a 100 km orbit just fine: https://www.dropbox.com/s/sqvhucjs1y73es8/22T%20No%20Asp_%202.craft

I also made an edited version, adding some winglets, more fuel, fixed Mainsail overheat, and used less RCS. This one made it to a 100 km orbit with fuel to spare: https://www.dropbox.com/s/l6cgbb0frm0qe6u/22T%20No%20Asp_.craft

Your design is actually pretty good and it's definantly not over-engineered. What does your ascent profile look like?

[luchelibre]

As for a thread with 'non-asparagus' designs, here you go.

That thread helps a great deal. Seems my rocket isn't really as overbuilt as I thought it'd be.

[luchelibre]

I built a close copy of your lifter and flew it to a 100 km orbit just fine: https://www.dropbox.com/s/sqvhucjs1y73es8/22T%20No%20Asp_%202.craft

I also made an edited version, adding some winglets, more fuel, fixed Mainsail overheat, and used less RCS. This one made it to a 100 km orbit with fuel to spare: https://www.dropbox.com/s/l6cgbb0frm0qe6u/22T%20No%20Asp_.craft

Your design is actually pretty good and it's definantly not over-engineered. What does your ascent profile look like?

Thanks for tinkering with the ship. Forgot about the anti-heating measures. I wish Squad would adjust that little glitch.

I'm left with about the same fuel as you get. Maybe a bit less due to not being as good a pilot.

[FenrirWolf]

Because if I don't learn how to make efficient non-aspy rockets before realistic aerodynamics come along (or I install FAR first), then I'm just going to have to build all my rockets over again.

That's a silly way to think about it. Your goal should be to learn to design things efficiently. And with Asparagus being the most efficient design method, there's no reason not to at least have experience with that paradigm. It's not like making Asparagus rockets will render you incapable of building anything else...

When the time comes that Asparagus is obsolete, go right ahead and learn whatever new paradigm emerges after that.

[luchelibre]

That's a silly way to think about it. Your goal should be to learn to design things efficiently. And with Asparagus being the most efficient design method, there's no reason not to at least have experience with that paradigm. It's not like making Asparagus rockets will render you incapable of building anything else...

When the time comes that Asparagus is obsolete, go right ahead and learn whatever new paradigm emerges after that.

I can build asparagus-staging just fine, but when you're as slow designing rockets as I am, it's better to shift and standardize my rocket design paradigm now (and get efficient at it) to prepare for the future instead of waiting for it and then hitting a design wall.

[luchelibre]

So I've got three 20-ton lifters to choose from. All have just about 4300 m/s of delta-v under atmo conditions and a TWR of about 1.7 off the pad.

1) The compact version I have already: 14400 units of fuel and costs 87670

2) A more natural looking rocket based off a post in the non-asparagus thread: 16920 units of fuel and costs 96620

3) A no-crossfeed single column: 12240/6800 units of liquid/solid fuel respectively and costs 83750

Which one would you recommend to use? Is the single-stack one non-scalable? It seems to be pretty much at the limit of stability for just 20 tons.

Finally, I don't know if this question can be answered, but between the first and second one, would 2500 units of fuel be necessary to overcome the aerodynamic difference should realistic drag be implemented?

Thanks for your indulgence.

[Wesreidau]

My 20-ton Munshot lifter is as such...

Mainsail / orange / orange / nosecone / drogue

Radial decoupler / crossfeed v / space tape

Mainsail / orange / orange / explosive stack separator / payload to LEO ~20Kt (Kerbotons)

Radial decoupler / crossfeed ^ / space tape

Mainsail / orange / orange / nosecone / drogue

Lock outer Mainsail gimbals and you're good to go. Drop the outer rockets when they run dry. Basically the #2. I understand your concern about wild asparagus staging and ignoring drag, but there's less drag with three stacks than five and the #3 looks unstable and inefficient. Engines that aren't burning are dead weight, and so is fuel higher up the stack. Remember, perfection isn't when nothing can be added, but when nothing can be taken away. Simplify.

[espm400]

My 20-ton Munshot lifter is as such...

Mainsail / orange / orange / nosecone / drogue

Radial decoupler / crossfeed v / space tape

Mainsail / orange / orange / explosive stack separator / payload to LEO ~20Kt (Kerbotons)

Radial decoupler / crossfeed ^ / space tape

Mainsail / orange / orange / nosecone / drogue

~snip~

I think your rocket fell over...

@Luchelibre: I would go with either the first or second of your three designs. Personally, as you can see by my pic on the previous page, I usually go for a core with two boosters and a third smaller stage. The reason being that the two booster design I use, as I said will get 30+ tonnes to LKO or 20 to the Mun. However, if I need to lift a bit more, all I have to do is go to the VAB, click on the radial decoupler connecting the booster stages and use 4x, or even 6x symmetry, and instantly I have a bigger rocket capable of lifting more. The way it's set up, I don't even have to add a fuel line or an extra sepratron.

The reason I would advise against the third design is simply on the wobble factor. It may be stable now, but if it ends up being your primary lifter, you're eventually going to want to upgrade it to lift more (as I just detailed above). The fact of the matter is that it's just too tall and skinny. That second design of yours, is the closest to how a rocket should widen outwards as it gets taller.

What can I say? I just like having my rockets look like they'd fly if they were build IRL.

Edited July 29, 2013 by espm400

[lammatt]

my general way way to build lifters is that i make sure i can afford a twr of ~1.5 to 2.5 for the 3-4km/s dV

so that i dun drop back to ground because of an unsufficient thrust to work against gravity and atmospheric drag

[metaphor]

The third design is not scalable since there is no more powerful engine in the game than a mainsail, that you could put on the bottom. The second one looks pretty realistic, like the Delta IV Heavy. The first one is asparagus which you said you don't like, but it's more efficient. From my experience a rocket's mass on the launch pad should be about 6-10 times its payload.

[TMS]

Where's this idea come from, that an aerodynamic model overhaul would mean abandoning asparagus staging?

A little presumptuous?

[Pursuedtank]

Where's this idea come from, that an aerodynamic model overhaul would mean abandoning asparagus staging?

A little presumptuous?

Yes, I am wondering this too. I have used Far before and asparagus staging has worked fine, so why would an overhaul remove the ability to asparagus.

[TMS]

I would imagine that realistic drag coefficients might make rockets with several "layers" less efficient than they are currently (or perhaps cost-prohibitive in career mode), but I don't get how one concludes that it would scupper the entire design.

[sturmstiger]

May I suggest taking a look at my Modular ELV Family (see my signature) for some ideas? The medium lift configurations, e.g., M102, M122 and M142 are simple and reliable non-cross feeding designs which lift 15, 20 and 25 tons to LKO. They are very similar to real world rockets such as configurations of the ESA Ariane 4 and the Chinese CZ-2 and CZ-3. I was using KW parts but they can be easily duplicated with stock parts.

From my experience, efficient, low part count, non-cross feeding designs should be quite easy for payloads less than 50 tons, but for anything above ~25ton to LKO, a pure serial (stock) design would make likely be very tall or involves a large number of parts (e.g. most Saturn V recreations).

One general suggestion I have for efficient (high payload ratio) rocket is to keep a good TWR: takeoff TWR of 1.2 to 1.4 and upper stage TWR of 0.7 to 1.0 seemed to be sweet spots for my experiments, and also typical for real world lift vehicles. Very high takeoff TWR means you're wasting your engine's power - you can either put more fuel or use less powerful but higher ISP engines.

[ASnogarD]

I used this...

... to get a 40 tonne payload easily to a rendezvous with a station at 250 km orbit with plenty of fuel to de-orbit the 2nd stage and leave no debris floating in orbit.

It doesnt use asparagus (it could I suppose but didnt need it), the main trick of this design is how it keeps the payload low inside the lifter by making the center stack short and feeding that stack with fuel on the outer stacks. I used 4 x LT30 for each outer stack and a mainsail on the center stack and 4 large radially mounted engines on the 2nd stage (technically 3rd as the lifter drops the 4 outer stacks when the fuel is out).

It flies relatively stable, just some minor fighting with the SAS during the gravity turn but nothing crazy.

It is made of stock parts but the payload has Aviation Lights Mod lights and Mechjeb, I also use Chatterer but since that mod has gone part-less I dont need to attach any part to use Chatterer .

[luchelibre]

Double Post

Edited July 29, 2013 by luchelibre

[luchelibre]

The first one is asparagus which you said you don't like, but it's more efficient.

It's parallel-staged like the second design.

From my experience a rocket's mass on the launch pad should be about 6-10 times its payload.

This kind of stuff is what I'm looking for. Though, that is a bit of a wide range.

May I suggest taking a look at my Modular ELV Family (see my signature) for some ideas? The medium lift configurations, e.g., M102, M122 and M142 are simple and reliable non-cross feeding designs which lift 15, 20 and 25 tons to LKO. They are very similar to real world rockets such as configurations of the ESA Ariane 4 and the Chinese CZ-2 and CZ-3. I was using KW parts but they can be easily duplicated with stock parts.

From my experience, efficient, low part count, non-cross feeding designs should be quite easy for payloads less than 50 tons, but for anything above ~25ton to LKO, a pure serial (stock) design would make likely be very tall or involves a large number of parts (e.g. most Saturn V recreations).

One general suggestion I have for efficient (high payload ratio) rocket is to keep a good TWR: takeoff TWR of 1.2 to 1.4 and upper stage TWR of 0.7 to 1.0 seemed to be sweet spots for my experiments, and also typical for real world lift vehicles. Very high takeoff TWR means you're wasting your engine's power - you can either put more fuel or use less powerful but higher ISP engines.

Thank you very much for the info. I'll test to be sure, but it was my understanding that the KW boosters were quite a bit more powerful than stock.

[sturmstiger]

Thank you very much for the info. I'll test to be sure, but it was my understanding that the KW boosters were quite a bit more powerful than stock.

You're welcome. The ones I'm using are the 1300KN engine for 2.5m core (replaceable with the 1500KN mainsail), the 230KN engine for the 1.5m core/booster (replaceable with the 215KN LVT-30), and the 120KN engine for the upper-stages which is indeed quite a bit better than any stock engine, but for the medium launchers can be replaced 2 to 1 by the Poodle. The slightly lower ISP of the stock engines can be compensated by adding more fuel to the 2.5m core thanks to the more powerful mainsail. And my design was just to show some more ideas of on all-liquid designs - you should be able to optimize for your specific needs.

[AstroDoc]

See my new thread 'Lifters, light, non-heavy, mainly-SRB and pancake, a method'