Placement of decouplers - matter of life and death?

[Atanar]

Just some thought-food:

This craft can launch and make orbit without problems:

Whiles this craft repeadetly explodes during first stage:

(yes, the only diffrence is the placement of the radial decouplers)

Why is that? Can anyone explain?

I am going to builld my asparagus like the first one from now on.

[AndrewBCrisp]

By "first stage", do you mean the vehicle explodes when the boosters separate, or is the vehicle exploding on ignition and liftoff?

[purpletarget]

I'm going to assume you mean it explodes upon staging the boosters loose from the rocket.

You can probably get some detail from the flight log by looking at where the first failure or collision is.

However, my guess would be this.

radial decouplers should kick a little when they disconnect.

When they're low on the booster, it'll cause the booster to rotate on it's COM, so the engine will get kicked away, and the top will go toe in towards the central body. If you're not clear enough, the top of the booster is probably hitting the middle or bottom of the central booster as it powers past, and there's explosions and shrapnel.

When the decoupler is high, the top of the booster is pushed out, and the engines will pivot in a bit, but probably not quick enough to hit the skipper before it powers itself out of the way, which is why the top design carries on unimpeded.

If for whatever reason a bottom decoupler is required and there's toe in issues, a couple cleverly mounted sepratrons can usually help clear the problem.

[Atanar]

No, the craft explodes because of structural failure somewhere between launch and separation of stage 1.

[Aragosnat]

Oh. I have had that problem. It is called you are to top heavy and there is not enough support to support the weight above the barrels of fuel.

[Anglave]

I'm having trouble believing that it just explodes on its own during ascent, before staging, and the only difference is the placement of the decouplers. Try building a fresh copy of #2 from scratch?

[Clouds]

I think it might be a bug.

[KevinW42]

Does it start spinning before it explodes? I sometimes have a problem with asparagus-staged craft where some imperceptible twisting causes some off-axis thrust, eventually tearing the rocket apart.

[Atanar]

Here you get an example. Does not happen every time but about half the launches.

And yes, I build copies with the same setup and they do the same. You should be able to replicate it yourself.

[michaelphoenix22]

more struts

[Brotoro]

What's the point of the long angled struts?

[Vanamonde]

What's the point of the long angled struts?

Triangle arrangements are very stable, and it's helping to distribute the stresses. From looking at the recording, I think a couple of lateral struts to either side of the decoupler would fix the problem. You've got all the forces of a Mainsail shoving up and the mass of 1.5 orange tank-equivalents dragging down on that decoupler, whereas in the other arrangement in your screenshots, the decoupler is much higher, less of the mass of the tanks is riding directly on it, and some struts are helping distribute the loads.

[entaran]

it's not a bug. It's him not firing the centre motor at the same time, so the mainsails on the outside apply excessive force outside of the COM so the thrust doesn't apply in a parallel vector to launch. Putting the radials at the top will stop the top of the stages "coming together" and causing the fuel tanks to snap the struts/touch/explode. You can get around this SLIGHTLY by leaving gimballing active, but then you get other issues.

Basically your problem is the low power motor in the middle not firing so you have to heavily reinforce the outer stages.

Also learn to strut in the direction of the force applied. In this case, apply the strut from a lower section of the outer booster to around the middle of the central stack.

[Atanar]

I think most people are missing the topic. I was trying to point out that the placement of the decouplers matters in stability. That placing them on the top fuel tanks is better and requires less struts. I wasn't asking for advice, of course I would need more struts on the second design.

Putting the radials at the top will stop the top of the stages "coming together" and causing the fuel tanks to snap the struts/touch/explode.

Could be true, buty why does the log note the structural failure of the decoupler first?

Edited June 8, 2013 by Atanar

[Anglave]

I think most people are missing the topic. I was trying to point out that the placement of the decouplers matters in stability. That placing them on the top fuel tanks is better and requires less struts. I wasn't asking for advice, of course I would need more struts on the second design.

Thanks for the video. I see your point and appreciate your taking the time to point it out.

Maybe this should go in some kind of basic "how to" section -- put your decouplers higher up for less explodey rockets!

[Giggleplex777]

more struts

That makes it even worse.

[SpaceSphereOfDeath]

I'm not sure, as I always use separators

Although it may be some part clipping issue?

[Fel]

I think most people are missing the topic. I was trying to point out that the placement of the decouplers matters in stability. That placing them on the top fuel tanks is better and requires less struts. I wasn't asking for advice, of course I would need more struts on the second design.?

You're just talking about basic "Trussing" issues here. The "Decoupler" is an anchor, but is a weaker anchor than structs.

STRUCTS do not HOLD the rocket together, they just transfer energy from one part to another (The "Decoupler" holds the rocket together).

Hence the decoupler failure causes the failure of the structs. As for the delayed response... "Drag Forces"

Thanks for the video. I see your point and appreciate your taking the time to point it out.

Maybe this should go in some kind of basic "how to" section -- put your decouplers higher up for less explodey rockets!

It is a generalized point, not a good point... the decouplers should be closer to a function of "CoM" and "CoT", he just placed them in the worst possible spot which caused them to fail.

[Atanar]

You're just talking about basic "Trussing" issues here. The "Decoupler" is an anchor, but is a weaker anchor than structs.

STRUCTS do not HOLD the rocket together, they just transfer energy from one part to another (The "Decoupler" holds the rocket together).

Struts can't phisically break in ksp?

So basically the stress at the top is lower that it breaks less likely there? Why? That contradicts all other assumption of this thread that the stress at the top must be higher because of the unactivated center rocket pulling the rocket together on the top.

I must admit that I was so fascinated by my first discovery that I haven't even looked into how well the decoupler does in the middle.

Edited June 8, 2013 by Atanar

[palker]

Simple enough the decoupler acts like a pivot for the lever as the center of thrust of the the radial booster is offset to the side from it. It pushes the fuel tanks towards the center of the rocket.

[Fel]

Struts can't phisically break in ksp?

So basically the stress at the top is lower that it breaks less likely there? Why? That contradicts all other assumption of this thread that the stress at the top must be higher because of the unactivated center rocket pulling the rocket together on the top.

I must admit that I was so fascinated by my first discovery that I haven't even looked into how well the decoupler does in the middle.

Struts can break... they're just "really really strong"

But the "attachment node" is what you have to worry about breaking... it is MUCH weaker than struts.

The tanks at the top are empty at the time the "decoupler" snaps, right?

Tanks empty DOWN, so that means that the majority of the mass (hence force) is going to be at the bottom. When it is fully empty, they'll be taking the full force of the engines.

*** And yes, that gets into the CoM issues. Too much top mass and you have issues... to little and you have issues... hence "issues".

*And yes, pivots are good points as well, since the decoupler is closer to the force... Trussing isn't my field, but it doesn't seem like KSP is wrong here.

Edited June 8, 2013 by Fel

[Atanar]

Simple enough the decoupler acts like a pivot for the lever as the center of thrust of the the radial booster is offset to the side from it. It pushes the fuel tanks towards the center of the rocket.

Now, would putting the decouplers in the middle better than having them on the top?

[palker]

The thrust will always cause torque in relation to the decoupler and it should be alway the same size as the distance between the force and pivot does not change. I guess if you put decoupler in the middle ans strut it on the top and bottom with lateral struts the load on the struts will be lower as it is in your current arrangement.