Decoupler Snapping

[RocketBlam]

I've got a problem on one of my rockets. I'm using this big Decoupler here, and... it just keeps snapping half way through launch.

Anyone got an idea how to resolve this? As you can see I'm trying to put in some reinforcing struts, but nothing seems to make any difference. This thing just snaps randomly in the middle of flight. I know it's the problem because after it happens, I can still select it, but I can't select anything below it.

[turkwinif]

Those big orange tanks are notorious for being unstable. Try putting those two big white tanks in place of the orange tank instead. (don't forget the struts)

[NecroBones]

The way you're doing it with struts is exactly how I reinforce those things. It usually works for me, but I also do a lot of asparagus staging around it, and strut those to the upper stage as well. Usually I also add struts to brace those girder pieces to each other (make a nice square around that orange tank).

If you install Kerbal Joint Reinforcement, it solves this for you pretty well. I just started playing with that tonight, and it tightens up those joints quite a lot.

[Jouni]

When does the decoupler usually snap? Already on launchpad? Or maybe some time after liftoff, just before booster separation?

That connection looks perfectly fine by itself. Can you show a picture of the entire rocket on launchpad, so we can figure out the source of the problem? And please switch the MechJeb delta-v window to show all stats, so we can see how high TWR goes in each stage.

[Vanamonde]

How much is piled up on top of it, and what is the G load on the ship when it happens?

[lincourtl]

I've never had a need for that level of strutting. What I usually do is place a cubic octagonal with four fold symmetry above the decoupler, and below the decoupler, then use a single strut with four fold symmetry to connect the two. If you get Kerbal Joint Reinforcement though, you probably won't even need that. It'll be interesting to see how this plays out with the new joints in 0.24.

[RocketBlam]

This is the whole rocket. It's the biggest one I've ever built because the payload weighs a crap-ton.

Now trying with two white tanks instead of one orange, and no reinforcements. I'll let you know.

[RocketBlam]

Well, I think the orange tank was the culprit. I replaced it and now the decoupler at the bottom is failing - it is also attached to an orange tank. now fixing.

[Jouni]

The problem seems to be that you have a tall top-heavy rocket, which probably makes it more than a bit wobbly. The best way to fix the problem is to try to make the rocket shorter.

Orange tanks are large single parts. That makes them rigid, which is not a bad thing in itself. In a wobbly rocket, the connections next to large parts often experience the greatest stress, so structural failures tend to happen there. Those structural failures can be sometimes avoided by making the rocket more flexible. By using stacks of smaller fuel tanks instead of large fuel tanks, you create more joints in the rocket. However, that only fixes the problem, but not its root cause.

Based on the numbers, I assume that you have 7 Skippers in stage 5, and a single Mainsail in stage 4. It might be better to combine those stages: use a Mainsail in the center and 6 Skippers in the boosters, and drop the boosters when their fuel tanks are empty. This makes the lifter stages shorter and more rigid. On the other hand, TWR may grow a bit excessive before booster separation, so it may be wise to reduce throttle at that point.

[RocketBlam]

Well, I made some progress and lost some progress. My save file bugged out so I had to rebuilt the rocket. Now, once again, that same decoupler is failing, even with two white tanks instead of an orange one. I guess I'll just have to launch this in two pieces.

[Claw]

Well, I made some progress and lost some progress. My save file bugged out so I had to rebuilt the rocket. Now, once again, that same decoupler is failing, even with two white tanks instead of an orange one. I guess I'll just have to launch this in two pieces.

If it's happening part way into your flight, throttle the engines down a bit as you burn off fuel. It might be because it's a weak link (large SAS and couplers seem weak for their size). If your rocket is accelerating too hard and snapping it, throttling back might help. (Although your TWR doesn't seem too high...)

If you're okay with MechJeb helping you out a little, here's how I have used MJ for stuff like this.

Pull up the "vessel info" and "utilities" tabs.

In the Vessel Info tab, there's a "Current acceleration" parameter.

Fly your rocket however you normally do. Watch in the current acceleration parameter.

Where ever it breaks, take off a couple m/s^s and put that in acceleration limiter in the utility tab.

Hopefully you'll still have a descent TWR after throttling down. If you fly it by hand, don't forget to throttle back up after staging.

Edited March 15, 2014 by Claw

[SRV Ron]

It's probably happening during orbital turn with that top heavy load. You do, however, need to brace both ends of the decoupler or it can come loose, braces or not.

[Sirrobert]

I asume the heavy top of the rocket causes the entire thing to wobble, and at that point it will start to break. Is this correct?

You might be able to reduce the top heavy payload by breaking it up in 2 parts (with a docking port between) and sending them up in 2 different launches. Than assamble the 2 parts in orbit