Rocket design tips?

[CkGordon]

So I'm starting to have a problem. I can send a 1 man lander somewhat easily to the mun or orbit and sent a 3 man lander to Eve (99% luck. matched it's orbit around Kerbol and waited a couple years for an encounter) with a NERVA, but lately I'm finding that anything remotely heavy I have tons of trouble even getting into orbit, yet alone the Mun or another planet.

I downloaded the DEMV rover which weighs about 4.5 tons, probably closer to 6 after I fitted it out with some stuff and can't even get the thing in orbit. Is it because I'm using larger tanks and engines to account for the extra weight? Small rockets I have no problem with but the second i start slapping large parts like 3 man landers and large fuel tanks, it's feel like I'm trying to launch a boulder over a house with a rubber band. I need suggestions.

BTW my standard orbit procedure is to use up my SRBs, wait til my second stage gets me to 300 m/s, turn 45degrees along the 90degree til my height gets to 95km, then tilt to the horizon and burn til I reach a stable orbit. Maybe I'm using too much fuel? What tanks/engines do most of you use to launch sizable payloads long distance? I have several parts packs including novapunch, deep space mission pack, and a couple other random ones (nothing cheaty tho).

It feels like my progress was a steady increase: First orbit, first Mun landing, first escape of Kerbin, first Eve landing... then I started falling backwards to struggling to reach orbit with larger payloads.

So I turn to the forum for advice.

[EndlessWaves]

If you post an example attempt it would probably be easier to point out where you're going wrong. Generally though you want to be carrying a minimum of dead weight at all times for maximum efficiency. Ideally dump fuel tanks as soon as each one is dry and don't add RCS/SAS/Winglets etc. etc. unless you actually need them.

If you want a quick fix I find a 3200L tank with an aerospike on the bottom to be a good modular system for small to medium loads up to a few dozen tons. Each unit (tank + spike) will take approximately 2.5-3 tons directly to orbit using an efficient ascent profile and you can form them into a raft for heavy loads with not much bracing required.

The central section using small tanks and nuclear engines is interplanetary, it doesn't fire at all on ascent

The kerbal engineer redux mod shown on the left in that picture is also dead handy for getting an idea of how a rocket performs. It won't calculate the extra gained from drop tanks and other more complicated scenarios but for basic stages I find the DeltaV and thrust to weight readouts invaluable.

[Stevenator1]

The single best rocket design tip I have ever used is ejecting used tanks as soon as they are empty. One fantastic show of efficiency is DEADBEEF's http://imgur.com/a/Myh5Y#0 . That little thing can get the pod all the way to Laythe. It ejects the stages as soon as they are finished. Those 4 radial tanks can get you all the way into orbit, and that single tank on the NERVA is plenty enough for any basic orbital maneuver. I use that system to get satellites into orbit all the time, it works really well for low payloads.

[CkGordon]

This is the lander I'm trying to launch for Mun/Interplanetary missions. I'd preferably like to get it into a low orbit with a little fuel left for a last burn away from kerbin before having to rely on my NERVA for course adjustments but I'm running out of fuel before I even reach 85km. The tanks on the side feed the center first and I discard them on the way up but it's not enough.

[ATTACH=CONFIG]34002[/ATTACH]

[ATTACH=CONFIG]34003[/ATTACH]

I'm afraid to add a ton of parts for lag/rocket stability. I'm hoping I'm making some simple noob mistake that someone will spot and point out.

Edited October 1, 2012 by CkGordon

[AmpsterMan]

The jump from one seat landers to three seat landers is quite large. The 1 seat has a mass of 0.8, whereas the 3 seat has 4. That's 5 times heavier. Now because of the logarithmic pattern in which delta-v functions, your three seat launchers are usually going to be a fair bit heavier than your one seat designs.

That being said, the only general tips I can give you are these:

1: Always build your upper stages specifically for the mission you are planning. A mission that is only going to be orbitting Mun is vastly different than one that will be landing.

2: Plan for each step. Munar landing missions have at least 5 steps, or stages. 1st, get out of of lower atmosphere. 2nd , get into parking orbit. 3rd, transfer wherever you want to go. 4th, land. 5th, return. Just because there are five steps does NOT mean there should be a stage (on your ship) for each one, but it is a good starting point.

3: Every mission has the same start, getting out of Kerbin's Atmosphere. Therefore, I find it vert helpful to already have designs (either saved or in your head) that can get certain payloads into a parking orbit of your choice.

4: Calculate how much delta-v you need for each step in the mission. Note that this does not have to be pnipoint perfect. Just a rough estimate.

5: Always carry a little bit of extra fuel with you. It is better to bring too much then not enough.

Hope this helps!

[CkGordon]

The jump from one seat landers to three seat landers is quite large. The 1 seat has a mass of 0.8, whereas the 3 seat has 4. That's 5 times heavier. Now because of the logarithmic pattern in which delta-v functions, your three seat launchers are usually going to be a fair bit heavier than your one seat designs.

That being said, the only general tips I can give you are these:

1: Always build your upper stages specifically for the mission you are planning. A mission that is only going to be orbitting Mun is vastly different than one that will be landing.

2: Plan for each step. Munar landing missions have at least 5 steps, or stages. 1st, get out of of lower atmosphere. 2nd , get into parking orbit. 3rd, transfer wherever you want to go. 4th, land. 5th, return. Just because there are five steps does NOT mean there should be a stage (on your ship) for each one, but it is a good starting point.

3: Every mission has the same start, getting out of Kerbin's Atmosphere. Therefore, I find it vert helpful to already have designs (either saved or in your head) that can get certain payloads into a parking orbit of your choice.

4: Calculate how much delta-v you need for each step in the mission. Note that this does not have to be pnipoint perfect. Just a rough estimate.

5: Always carry a little bit of extra fuel with you. It is better to bring too much then not enough.

Hope this helps!

I've got the upper stage pretty well off. The 2 engines with lander engines on the side are for descent and landing, then they detach and the lander has another more powerful engine and fuel tank underneath for return to kerbin, and the NERVA should have enough fuel to handle the actual trip between parking orbit and destination. It's just getting the damn thing into a parking orbit without resorting to some huge lag monster that I can't fly and have to rely on mechjeb (I like to avoid autopilot at all costs) to get me through the trip. Would more lifting engines help? The one I'm using is from a add on parts pack and puts out 1550 thrust at 7.5 mass and thats the sole engine i'm using for lift to parking orbit before switching to NERVA (atleast if I could get to parking orbit).

[Vanamonde]

What sort of a problem are you having on launch? Not enough thrust, or stability? I'm asking because, while I'm not familiar with the mod parts you're using, that narrow waist between the stages looks like something that's going to flex an awful lot and throw the ship around violently, resulting in both steering chaos and possibly damage. Struts between the stages would help a great deal, but I also like to build my ships so that the stages overlap vertically, keeping the rocket shorter and providing more connection points for struts. The shorter a ship is, the less it flexes.

[Ziff]

Here are some suggestions. If you like Mods, use MechJeb to track things like your current T:W ratio and Delta-V expended during launch. Download the Kerbal Engineer Redux mod. This mod will be able to tell you the current Delta-V of each stage and the T:W ratio. Once you start building bigger rockets it actually takes a bit of math and design to get some really decent rockets. Read this Wiki, even if you don't understand the math, pay attention to the part about delta-v maps. Also, you can use this handy Delta-V Chart to plan your missions. Just add up the delta-v numbers along the path to figure out what is required.

The best advice:

1. Have a well defined mission goal so you can calculate what your required Delta-V for that mission is. Going to the Mun/Minmus are nearly identical in Delta-V requirements. However, going to Duna and back is completely different from Eve, or a moon around Jool.

2. Start with the smallest possible lander required to get the mission accomplished. Having final stages that are too large just compound problems for your rocket as each stage gets progressively larger, and in effect, less efficient.

3. Radial stages are by far more efficient than vertical stages. Have fuel lines feed from outer stages to the inner ones. There is a balance between how high you can stack fuel tanks before they get too heavy for lift off vs having not enough fuel to reach orbit. Start with 1 large engine and then try to support it with higher ISP rated engines (in radial stages) even if they have less thrust.

This is why the Kerbal Engineer Redux is such a great mod. It figures out the delta-v automatically for each stage. Using this in combination with a delta-v map will allow you to know, without testing, whether you have a shot at pulling off your mission with your current design. If you want to land on the mun and don't have about 700m/s of delta-v in your lander stage, chances are, you ain't making it. If you dont have about 4500-4700m/s of delta-v total between your launch stages, you won't reach orbit.

I threw this simple design together as an example of what you might want to start with. Edit: It's important to note that there are 2 sets of radial stages (each with 3 engines) , they have fuel lines that feed inwards and detach separately when each set empties. This is the efficient way to do it.

Liftoff!

Orbiting the Mun.

Edited October 1, 2012 by Ziff

[Peppe]

Thrust just gets you moving. You need to sustain the thrust over time to get into orbit. How much fuel do you have for that engine?

First guess, it sounds like you need more fuel in your first stage rather than adding any more thrust.

For stage one, a general target is 1.5-2 thrust to weight ratio. More thrust than that will often be wasted in atmosphere if you over throttle (you accelerate into your terminal velocity).

Without doing a lot of math you can ballpark your thrust to weight by comparing the thrust / 10, example for your stage one it is 155. compare that to your total mass. I don't recognize most of your parts, but looks like around 40-50 tons. So you are probably in the 3:1 to 4:1 TWR range. In game should be able to feel it as not needing much throttle to get moving.

For 100km orbit you need about 4700 m/s deleta-v for average control. Very good control could get it down to like 4500 m/s. This is a little harder to calculate, so to ballpark just add more fuel tanks if you come up short in a stage. Its more efficient if you can drop empty fuel tanks on the way up, but if not you can always add more fuel and more engines (thrust) to keep your thrust to weight positive (above 1, but again target 1.5 - 2 for ascent).

Once you are in space you can get away with much lower thrust. If you plan to land on anything with a lot of gravity though you may need to adjust your engines for that stage to allow an ascent.

[Markus Reese]

The elimination of dead weight is key to lifting large rockets. Endless system works, but isn't so fun in my opinion, and doesn't work so well if you want to send larger returnable launchers etc. Ideal design I use I have seen called the "asparagus" setup. Dunno why. It entails the following. Once you have your travel and lander stage set up and working, you then set about designing your launch and orbit stage. For this, I have a central single large engine (or cluster for small rockets). I mount another set of engines around it.

When in space, this thrust output will get you into orbit quickly since you want the thrust to capture orbit without wasting on fighting gravity. Around this, I have a ring of engines that fuel feed into the core one. (fuel lines) This is the high atmosphere stage. To get out of the high air resistance, I have a tertiary booster set around that secondary ring and feed into it.

I ignite the seconary and tertiary on launch, with srbs for initial thrust if needed/desired. In terms of mass/tank height and reduce initial weight, I always have my per stack size of the tertiary>secondary>primary. Mainly is since your thrust reduces, you want to reduce mass to maintain a thrust to weight ratio. Or even increase it since you have less mass=<efficency. By using the fuel feed method, you maintain a fuel supply for you next stages, but continue supplied boost to prevent dead weight. When tertiary jettison, you ignite your primary. Your secondaries have full fuel tanks because they got fuel from tertiary. Same run for your primary. When your secondary go out, your primary is still full fuel.

To give a simple but pretty clear example of how powerful this setup is, I have my standard planetary travel setup that carries around landers. The power is so great when I start getting out of the air resistance that I need to turn down the throttle else the g forces break the thing apart. In fact, when I am capturing orbit, I only need to use about 10-20% throttle.

[EndlessWaves]

I don't recognize most of your parts, but looks like around 40-50 tons. So you are probably in the 3:1 to 4:1 TWR range. In game should be able to feel it as not needing much throttle to get moving.

The identifiable stock parts on the upper stages come to 25 tons so that's a very low estimate for the entire rocket. I don't use those parts either, but assuming the central section is twice the weight of a 3200L tank (36 tons) and the outer ones are the same weight as 4x400L tanks (9 tons each) I'd put that rocket at over 100 tons.

[cardgame]

This is going to sound cliche, but my serious answer to you is moar boosters. My 3-crew munar lander is no paragon of effiency, but it only uses 3 large tanks and one half-size for the lander. The rest of it is boosters.

Edited October 1, 2012 by cardgame

[maltesh]

The elimination of dead weight is key to lifting large rockets. Endless system works, but isn't so fun in my opinion, and doesn't work so well if you want to send larger returnable launchers etc. Ideal design I use I have seen called the "asparagus" setup. Dunno why. It entails the following. Once you have your travel and lander stage set up and working, you then set about designing your launch and orbit stage. For this, I have a central single large engine (or cluster for small rockets). I mount another set of engines around it.

<efficency. by="" using="" the="" fuel="" feed="" method,="" you="" maintain="" a="" supply="" for="" next="" stages,="" but="" continue="" supplied="" boost="" to="" prevent="" dead="" weight.="" when="" tertiary="" jettison,="" ignite="" your="" primary.="" secondaries="" have="" full="" tanks="" because="" they="" got="" from="" tertiary.="" same="" run="" secondary="" go="" out,="" primary="" is="" still="" fuel.

Ed Kieth's Innovative Asparagus Stalk Booster is described in </efficency.>Tom Logsdon's'Orbital Mechanics: Theory and Applications and pretty much nowhere else. As described in the book, its seven equivalent liquid fuel rockets stacked with six surrounding the central stack, like so.

If you don't get the fuel lines right, you'll wind up with an unbalanced craft that will tumble. If you don't get the staging right, you'll wind up popping off a still burning stage with generally catastrophic results. If you strut down the boosters (and you probably should), and you don't attatch the struts symmetriclaly, it will probably will spin out of control.

If you do get everything right provides a pretty efficient lift stage that gives you maximum thrust off the pad when you need it most, and gets rid of the empty tanks quickly, so that they aren't weighing down your spacecraft, cranking up your full mass/dry mass ratio each time you dump the empty tanks and resulting high delta-V. It's become a standby in my spacecraft design stable to get things into orbit.

(edit)

As for why it's called an Asparagus Stalk Booster...see Ziff's post below.

Was wondering why the spoiler wasn't working. Thanks.

(end edit)

Edited October 2, 2012 by maltesh

[Ziff]

Maltesh, spoilers don't work when you put an ' anywhere in the spoiler text name.