Your rocket design philosophy

[Commander Amaro]

Maybe philosophy is the wrong word here, but what goes through your mind while designing a rocket, and how do you go about making it fulfill it\'s intended purpose?

I first try and determine how adaptable I can make the entire design and how well it can respond to unexpected (read: overlooked) problems. Then I think in terms of what each stage should do:

The first stage (aka the 'lift stage') has to be big and powerful, because the more power and thrust that I have initially, the more equipment and fuel can be in the other stages, and I\'ll have more versatility with the rest of the design. SRBs can often be used here to great effect to help overcome the large amount of inertia and to save fuel.

The second stage (or the 'orbital insertion stage') has to strike a balance between power and efficiency. After the first stage throws a large amount of equipment into the upper atmosphere, this stage helps to add a large amount of orbital velocity to the craft and give it the final push out of the atmosphere that it needs to achieve orbit. It has to have enough power to give it the necessary velocity to achieve orbit, but at the same time it has to have the efficiency and the fuel capacity to burn long enough to get the job done.

The third stage (aka the 'orbit stage' or the 'mission stage') mainly has to be fuel efficient and to have a good fuel capacity, without as much concern given to power as the other two stages. This stage is used for any correctional maneuvers that are going to be made to the orbit/trajectory, any more velocity that is going to be added, and any de-orbit maneuvers that will take place. Power is not as big of a concern, but efficiency and capacity is, because if this stage runs out of fuel while you are still in orbit, it is highly likely that you are not going to make it back to the surface of Kearth.

That is my rocket design philosophy. That is what goes through my head while designing and building a rocket, and so far it has been a rather reliable set of ideas to stick by.

So, what is your rocket design philosophy? How do you build a new rocket? What goes through your head during the design process? Insight into other people\'s design process could be very useful for anyone who is starting out or looking for new ideas.

[SkyRender]

It usually goes something like this:

1. Okay, I need a lander, so I\'ll make a small one with the ship\'s ASAS.

2. Okay, it needs to get into space, so let\'s build a tri-coupler stack below it with a decoupler between the two and launch that bad boy into space already.

3. Crap, not enough lift. More fuel!

4. Damn it, why are these engines so fragile that a soft-landing disconnects them? Struts!

5. That took forever. Man, I\'m just gonna make an orbiter next time...

[tasuit]

I usually try to keep the 'spare' parts down. If I make a mistake, then I\'ll have to relaunch.

Less is more, I always say.

[SteevyT]

1. Does it fly?

-If no, add more engines

2. Does it get where I want it to?

-If no, add more fuel

3. Does it not rip itself apart?

-If no, add more struts

[Newman]

3. Does it rip itself apart?

-If no, add more struts

Heheh, I think you meant if yes add more struts. That or 'if no add more boosters'

[CarolRawley]

1. Build an orbit stage(Usually a small satellite)

2. Build a lift stage

3. Cover the orbit stage in weapons from mods.

4. ...

5. Fulfill Order Foxtrot 520.

6. ...

7. Profit!

[SteevyT]

Heheh, I think you meant if yes add more struts. That or 'if no add more boosters'

Fixed it, but even when the answer is no, I sometimes add more struts. =P

[lebull]

A rocket\'s stability is the second most important aspect (not as important as the number of rockets, of course). If you\'re ever in a case where a rocket seems uncontrollable, I do something like following...

1. Find the guys who made found the SAS.

2. Create another rocket.

3. Put the guys who made found the SAS on the rocket.

4. Launch to the moon.

4a. If this rocket has stability issues, go to step 1.

5. 'Forget' your parachute.

[Commander Zoom]

Capsule - 'mission' stage (lander, orbiter, whatever) - upper booster with one engine - lower booster with multiple engines - ring of SRBs around the base to get the whole thing off the ground and moving.

[Millitron]

The first stage has to be big and powerful enough to lift everything else. Once I am sure it can lift the upper stages, all my focus goes into increasing burn-time. The longer I can burn for, the easier it is to steer. For example, say engine A and engine B will both provide 1000m/s delta V with the given fuel, but engine A takes 30 seconds while engine B takes 60. All other factors the same, engine B will be more stable, and thus I waste less delta-V to correction burns. Also, I connect all radially mounted engines to each other, and then to the main stack with struts, to further help stability. The new fuel-lines system is making the awesome, as I can now have proper liquid boosters.

Next stage varies based on what I\'m doing. If I\'m lifting huge payloads, I have stage 2 be more big engines, to help get into orbit. If the payload is smaller, I skip this and just start my orbital stage now.

My orbital stages focus on fuel conservation. I always take the longest burning engines I can, since they seem to provide the best fuel-efficiency. I rarely use RCS, I find it unnecessary for most situations.

My lunar landers consist of legs made out of radial decouplers and wings, and usually do not have a dedicated descent stage, I bring the whole lander back to Kerbin.

[Newman]

Yep, once you\'ve already reached orbit, I find slow burning engines are better from that point on.

[Johno]

1) 'Hey, that part looks really cool. Wonder what it does?'

2) Try out resulting rocket on launch pad.

Repeat until you get the desired result.

[Technical Ben]

Does it look cool? If no, then add more of everything!

[Mostlyharmless]

I usually consider the payload i\'m taking up first. Top down design. My cycle goes something like this.

Make an orbiter with maneuver fuel, rcs, and SAS

Test launch it.

Does it work? Great, moving on.

Build the lift stage.

Is it stable?

if no,Add ASAS to the bottom of the payload stage.

if yes. Get ye into space.

Did they come back alive?

if no, remember to add parachute to the command module.

if yes. SUCCESSFUL MISSION.

[Switchblade88]

I usually consider the payload i\'m taking up first. Top down design.

Same here.

1. I\'ll start with a payload (lander, satellite, Mun base etc) then get greedy and add them all at once.

2. Test in current setup, making sure satellite ejects BEFORE Mun base does... :-\

3. TMI stage. No test, she\'ll be right mate.

4. Orbital stage, test w/ boosters to see if it\'ll do the job by itself

5. Launch stage, find all the instabilities and rip itself apart repeatedly... rebuild better.

6. Actually get to Mun to see if it\'ll work as described

7. Return for profit. Or funeral insurance payout. (Still profit? )

[Ivan Ivanovich]

Basically it\'s a 'top down' design. What should my final stage accomplish, and what steps get me there?

A moon landing ship is designed this way:

1. I need a returner. A capsule, a parachute, a decoupler below it.

2. I need to get the returner back from the moon. I need enough thrust to take off from the moon, circle around it once, get into a TKI, insert into Kerbin orbit and deorbit. Experience tells me that 1-2 tanks plus a gimballed engine serves that purpose.

3. I need a lander to land on the Mun. I need an ASAS, a fuel tank, a gimballed engine, RCS for balance and some fins to land on. This can be combined with the parts in step 2, I just need a neat way to separate them to jettison dead weight.

4. I need something that gets me from Kerbin orbit to the Mun, for Mun insertion and deorbit. 2-3 tanks should do the trick, plus an engine.

5. I need to get into Kerbin orbit. That stage has no requirements but the ability to lift everything above into a 70-100k meters orbit.

[Spoot Knight]

I\'m one of simple protocol.

It must look good, and then it must perform well.

If it doesn\'t perform well, then defy logic and make it work anyway.

[rdfox]

2. I need to get the returner back from the moon. I need enough thrust to take off from the moon, circle around it once, get into a TKI, insert into Kerbin orbit and deorbit. Experience tells me that 1-2 tanks plus a gimballed engine serves that purpose.

Y\'know, Ivan, you can save quite a bit of weight by doing what they did in the 60s (both US and Soviet programs!) and just make a direct re-entry instead of trying to reinsert into LKO before deorbiting! Just make a midcourse correction once in Kerbin\'s SOI and drop to a perikee of about 30km, and you\'ll get a nice direct reentry... with entirely survivable g-loading, even! 8)

[Jack Slinger]

MORE BOOSTERS

[NovaSilisko]

MORE BOOSTERS

Sorry, but that got old a while ago

[707-Engineering]

Sorry, but that got old a while ago

sorry but it just won\'t get old as long as people will use non modded parts, its my first though when a craft doesn\'t go as fast at a certain altitude or to a certain altitude.. and My first craft to carry 1.6 fuel to a transpher orbit mun is one based on a booster brick..

1. making a lander or orbiter

2. Orbit or/and transfer stage

3. speed enhancement stage

4. VLFFE (Very Large Fuel Fed Engines)

5. booster block to get the thing up the ground

&. in between booster attachments to speed up, keep up, or simply hope it does something right.

oh and if things go wrong? the RSLALSG maneuver

[rebelcommando1807]

I usually try to build them working the first time.. Or quickly learn. Due to me having rubbish piloting skills, I tend to have an ASAS module on the ship somewhere.. And, as of late, I\'ve been finding that I try more interesting designs (Including a tripedal design with fuel feeding into the center stack).

It may not be very Kerbal but, unless I\'m using mods, my rockets don\'t tend to explode anymore.

I think about what I want to get up, what\'s worked in the past, and how rockets behave in the latest version. For example..

RCS training craft, designed to get into a low orbit (Spaceworthy X-2)

Payload: More RCS fuel than is needed

2 tanks of fuel with gimballing engine

ASAS for stability

Generally go payload first, make cool lifter next. Generally, though, it ends up being purpose-based. Get into orbit? A three-stack large tank arrangement, leaving waaay too much fuel for orbital \'fun\'. Land on the Mun? Haven\'t ironed that one out yet, but I\'ve seen an interesting design or two that would be easy to adapt.

[Monger]

Well, usually it works like this for me:

- Create a mission: where to fly to, and how?

- What - from the feeling of my guts - is probably an efficient way to do this? How many stages do I need?

- Test: find the simplest configuration of each stage, and try to test them seperately

- Scale: If it works fine, try to find the most efficient scaling of each stage (more tanks/less tanks)

- Finish: put them together, improve stability, and then go.

Most of my ships are still rather small. My feeling is, that if I build a big rocket without knowing why it works, I have learned nothing. And testing rockets with a lot of stages is really, REALLY time consuming.

[fishjam]

my poor attempts so far start like this

1st. set goal ( atm get to orbit, minimal explosions

2nd. build basic rocket that has a chance of lifting off, usually with ASAS and boosters for inital launch.

3rd. hope it doesn\'t explode, lose control

and thats it.. so far mine all crash and burn.. so.. i think i need to remember to turn on sas before i click launch..

[khyron42]

Like many others, I tend to design the upper stages and then build & test lower stages on it, until it meets the goal.

1. Upper Stage + payload - here\'s where I play with a lot of variation in the design.

2. Fly it, see how it does with just the upper stage from the pad. Does it explode? flip over? come apart? work with 50% throttle? Do I think it will do ok once the lower stage(s) dump it into space? If so...

3. Figure out a next stage.

4. Flight test combined vehicle.

6. If needed, repeat 3.

6. Once lower stages work right:

a. Upper stage fails spectacularly once delivered to orbit/near-moon/solar orbit. Redesign with wildly different payload.

b. Upper stage barely fails. tweak upper stage or lower stages.

c. Design works perfectly. Fly it once or twice, then get bored with it and try a new crazy idea.