how to make space shuttle?

[Snikersnee]

Oh, I guess i was wrong.

[Zephram Kerman]

Getting the Space Shuttle to work right is an extreme challenge. When you finally do accomplish this, it will be just as satisfying as your first successful Mun landing! The system requires deep understanding of several concepts that are not intuitive. So keep researching, reading, asking, striving! This stuff is fun to know, and you can use it at parties to impress your friends with how nerdy you are.

I got 80% through making a Space Shuttle tutorial in 0.90, until I took an arrow to the knee. (Version 1.0 came out, and all the laws of physics changed.) I might get around to it someday. But there are plenty of shuttle tutorial videos and threads already.

In addition to all the great advice in this thread, here are a couple more images that might help:

Like Slashy's vector diagram. In this one, the yellow box shows how the combined center of mass moves as the fuel is consumed. The vector sum of all five engines must pass through the combined center of mass. Once the boosters separate, the vector sum of the three SSMEs continues to do the same. The three SSMEs were each canted outwards from each other; in case one or two of them failed during ascent, the thrust vector would still be aligned.

The main issue is the Skippers and Mainsails don't have nearly as much gimbal range as the SSMEs. (That's right, the real life engines are more OP than the game ones!) So to compensate for this with all stock, I suggest it's not cheating to hide three of the big SAS in the external fuel tank.

From WinterOwl's tutorial. His brilliant idea was to cluster seven SRBs on each side. The clustered components had four different thrust limiter settings, so that they would burn out at different times. I think they were 100%, 90%, 80%, and the middle one was 70%. My version had the center one throttled way down, so that it had just enough thrust to propel the empty cluster along during separation. That made the boosters separate in formation with the rest of the ship, just like real life. It looked very cool.

Finally, like any spacecraft, design the thing in reverse order! Build the orbiter stage first, and get it to fly right for the landing. You can launch it on the runway using Infinint Fuel cheat, for testing purposes only of course. Once you've got the orbiter locked down, you can design the external tank and boosters around that.

[ajburges]

It's funny, it took me a couple decades and dozens of shuttle launches before I realized that there was no engine on the "big orange thing" and that it was just a big fuel tank. I guess I just assumed it had one, I never scrutinized the space shuttle design close enough. I love the space shuttle design, I find it very fascinating what they came up with and how well it worked for so many flights.

OTOH the Buran had engines on every stage. I thought "add moar boosters" was America's (and KSP players') hat.

Problem is, from a utility standpoint, you want to maximize recovery fraction. That means maximizing the proportional cost of your orbiter. That tends to lead to a larger orbiter. A larger orbiter also mean a larger displacement vector. Payload has a greater displacement of CoM.

Hmm...

Perhaps another big strike against STS replicas is the scaled down nature of Kerbin. The low orbital speed means you need less dV to orbit, but the planet mass still demands similar TWR to Earthen craft. This make a very disportionate balance between tank size (your source of dV) and orbiter size (your source of thrust). This unbalance means that your dry CoM is shifted towards the orbiter and your CoM migration increases. Ergo, trying to mimic Earth launch stages either gives you excess dV, wonky performance, or the parts were tweaked to make it work similar to Earth systems under new conditions.

[GoSlash27]

*snip*Problem is, from a utility standpoint, you want to maximize recovery fraction.

Ideally, yes, but IRL there was a caveat that went ignored: except where it's cheaper not to. Another subject for another day...

Perhaps another big strike against STS replicas is the scaled down nature of Kerbin. The low orbital speed means you need less dV to orbit, but the planet mass still demands similar TWR to Earthen craft. This make a very disportionate balance between tank size (your source of dV) and orbiter size (your source of thrust). This unbalance means that your dry CoM is shifted towards the orbiter and your CoM migration increases. Ergo, trying to mimic Earth launch stages either gives you excess dV, wonky performance, or the parts were tweaked to make it work similar to Earth systems under new conditions.

All true. STS/ Buran style launchers are pretty pointless in KSP. Vertical rockets can get the same payload into orbit much more cheaply even when you don't bother recovering the pieces. SSTOs easily outperform them in all ways and are 100% recoverable to boot Not to mention easier to make...

The only reason to make one in KSP is the challenge of it.

Best,

-Slashy

[The_Rocketeer]

Pointless isn't really a relevant term is it? After all, we're playing a videogame... the only point of any of it is entertainment value, or arguably a little bit of intuitive knowledge gain.

Anyway, one thing that building a STS replica does do is teach you that not all rockets need to be phallic missiles - they can be practically any shape you want, and as long as you understand the physical practicalities and design accordingly you can make absolutely anything and get it into space.

[GoSlash27]

I knocked together a very crude demonstration of an offset lifter with no RCS, vernors, or reaction wheels.

Not good for anything other than a learning aid. It will eventually flip out once the center of mass falls behind the center of lift, but it's reasonably stable to well over Mach 1.

Download link to follow once Wikisend decides to act right.

Best,

-Slashy

[ajburges]

Damn, no wonder it's retired

Not sure if serious or...

The reason it's retired was a lack of good abort modes. On a traditional stack, you have a LES which you can fire at the first sign of things going wrong. STS abort modes were much more limited.

Kerbals don't sped a lot of time developing safe abort modes. The closest we have is to bail without a parachute and hope the engines don't cook them.

[Mikki]

Hi OP

I grabbed a older Pic from 1.0, still works fine.

You can see the crew is fine in this space shuttle!

I solved the problem of the moving CoM with attaching the fuelline inline with the average CoT line (Mammoth engine, somewhere under the LFO tank behind the MK3 Commandcockpit)... no need for much SAS or Vernorspam... This Shuttle had 2x3 Kickbacks attached at takeoff. Orbital Manouvering with the smaller engines attached to the Mammoth, inline with Shuttle CoM alone. Stockparts, no big deal to manage, takes some time nevertheless.

Making this was one of my best KSP experiences

Oh, this shuttle moves slightly sideways after takeoff, but its allways pointing radial to the surface... making Gravity turn like a rocket, throttlecontrol to keep stable heading.

Edited August 23, 2015 by Mikki

[Geher]

The trick is really about the engine gimbal.

I've built a pretty good "almost stock" MK3 Shuttle. I copied the skipper engine and doubled the gimbal. Now I have 2 Skippers: The Normal Skipper Engine and the Shuttle Skipper Engine.

The biggest SRB's in stock KSP are also not powerful enough. Use 3 or maybe even 4 on each side.

But don't think launches will get cheaper, just because you can reuse the engines. It is much more expensive than an easy expendable rocket. Here's the reason why:

You need lot's of extra fuel to get not only the payload, but also the orbiter itself into orbit. This extra fuel costs funds.

You need a bigger external tank to carry the extra fuel. Fuel tanks are surprisingly expensive in KSP.

You won't be able to reuse the SRB's in stock KSP.

So what I did to make Shuttle launches actually cheaper than expendable rockets is:

Install a mod for stage recovery (for the SRB's).

Add a heatshield, a probe core, a battery, some rcs and lot's of parachutes to the external tank. After detaching it from the orbiter it is almost in orbit and by adjusting the speed just a little bit I can choose to land it wherever I want. I use mechjeb's landing guidance to get the landing near KSC. After I achieved orbit with the Shuttle I change vessel back to the external tank and watch it as it falls back to earth.