Stable Radial Decouplers

[hyperbolicj]

Well I've sent a total of 5 vessels into space now but I periodically have a reoccurring problem.

That is: When I use radial decouplers to detach my booster rockets from my main stage they tend to rotate once detached. If they do so fast enough my rockets main stage then blows up and the whole mission goes up in flames. I postulate that this is due to the radial decoupler not being inline with the center of mass of the booster rocket giving it a slight angular momentum. But I'm only human and setting my rockets as close to their COM with the radial decouplers I still occasionally have this issue. The question posed is this:

Is the position of the radial decoupler the cause of the problem?

If so is there a way to balance it out with a second radial decoupler, or is there a more elegant solution?

If there isn't how can I attach a rocket with multiple radial decouplers, because when I tried only one would be attached at launch.

Edited October 29, 2014 by hyperbolicj

[Starman4308]

Well I've sent a total of 5 vessels into space now but I periodically have a reoccurring problem.

That is: When I use radial decouplers to detach my booster rockets from my main stage they tend to rotate once detached. If they do so fast enough my rockets main stage then blows up and the whole mission goes up in flames. I postulate that this is due to the radial decoupler not being inline with the center of mass of the booster rocket giving it a slight angular momentum. But I'm only human and setting my rockets as close to their COM with the radial decouplers I still occasionally have this issue. The question posed is this:

Is the position of the radial decoupler the cause of the problem?

If so is there a way to balance it out with a second radial decoupler, or is there a more elegant solution?

If there isn't how can I attach a rocket with multiple radial decouplers, because when I tried only one would be attached at launch.

It's a known bug with how the decouplers apply force: it is particularly noticeable for FAR/NEAR users. Claw's stock bug fixes includes a patch for this: you can additionally use Tweakable Everything to set decoupler force to 0 (allowing it to basically slide off), or abuse Sepratrons to prevent collisions.

[DeathSabre]

You can't attach something by two different points. (except when docking a craft with multiple docking ports but that's another story. Can't be done in VAB/SPH)

The placement of your Decoupler is probably at fault. but it'll be its position relative to the ejected piece, not relative to the main craft. although moving the whole radial assembly down can sometimes help so that you can get out of the way before it hits your main craft.

the two easiest solutions are to

A: Use sepatrons to get your ejected assemblies out of the way of your rocket. or

B: attach your radial assemblies so their COM is bellow the Decoupler so that when you eject them their noses will fall out and their tails fall in.

for smaller craft I use method B. for bigger ones with 2.5m parts or bigger attached radially I generally use Solution A.

fly safe

[Geschosskopf]

Is the position of the radial decoupler the cause of the problem?

If so is there a way to balance it out with a second radial decoupler, or is there a more elegant solution?

If there isn't how can I attach a rocket with multiple radial decouplers, because when I tried only one would be attached at launch.

The root cause of the problem is a bug introduced in 0.24.2 which screws up decoupler forces. As mentioned by others, you can download a mod (which is actually a patch) that solves the problem. Or you can just work around it.

To work around the problem with radial decouplers, there are a few underlying assumptions and other things to bear in mind:

• The main problem is the top end of the radial booster swinging into the central stack, so you want to keep that from happening
  
• The radial boosters, except for the smallest SRBs, generally need struts to hold them on during use. However, no matter how big your radial boosters are (including tall 3.75m liquid stacks), you only need at most 2 struts per booster, 1 at the top and 1 at the bottom, connected to the central stacl.
  
• It's important to keep the number of struts at a minimum because the root cause of the problem seems to be a screw-up in how decouplers break struts. The more struts they have to break, the less well the decouplers work and the less predictable the results.
  
• Working around this issue requires putting Sepratrons on the radial boosters but you have to be careful there. You don't want too many Septratrons blowing fire on the central stack for too long or they'll explode the central stack.
  
• The radial boosters will always have 1 end or the other swing in somewhat no matter what you do, so you need some stand-off distance between them and the central stack. This means you should only use the TT-70 radial decouplers for anything bigger than the small SRB. The TT-70 is the one with the diagonal bracing holding it out from the central stack.
  

With that said, you now have 2 main options depending on your rocket design. The easiest method solves the underlying problem (tops of boosters swinging in) by using Sepratrons to swing the boosters in the opposite direction, so their lower ends swing in. To keep them from colliding with the central stack, you want their lower ends to miss just behind the engine of the central stack. This requires you build the rocket as follows:

1. The radial decouplers must be mounted as low on the central stack as possible, so the pivot point is as low as possible.
   
2. Build the radial boosters so their lower ends are no higher than about even with the bottom end of the central stack. If they stick down below the central stack, so much the better.
   
3. Put 2 Sepratrons mounted in the horizontal plane at the top ends of each radial booster. While keeping them in the horizontal plane, drag them out outboard around the curve of the radial booster so their nozzles don't point straight back at the central stack. This minimizes Sepratron flame impingement on the central stack down to safe levels.
   
4. OPTIONAL: tweak the amount of fuel in the Sepratrons down to a very small amount while leaving their thrust alone. You need their full thrust to counter the inward swing of the top ends of the boosters but the longer they burn, the more they swing the bottom end of the booster inwards.
   

Using this approach, often the lower ends of the boosters will collide with each other just behind the central stack, resulting in harmless and entertaining explosions to make your launch more fun, while the central stack continues on undisturbed.

If mounting the radial boosters as above isn't possible, then you also need Sepratrons on their bottom ends. This makes 4 Sepratrons per booster. Usually, mounting (and optionally tweaking) the Sepratrons exactly the same way at top and bottom is sufficient. Then the boosters will fly out sideways while remaining more or less parallel with the central stack. But this is boring because you don't get to enjoy the harmless explosions when they collide with each other .

[Claw]

[*]It's important to keep the number of struts at a minimum because the root cause of the problem seems to be a screw-up in how decouplers break struts. The more struts they have to break, the less well the decouplers work and the less predictable the results.

Geschosskopf has all kinds of awesome information. I just wanted to point out that while this statement is true, it's not how struts break off that are at the root cause. It's actually a different problem, and struts are a symptom of that problem. Even without struts, if you decouple liquid fueled boosters consisting of FL-T800s at speeds between 700-750 m/s, the results are catastrophic. If you can manage to keep the decoupled boosters from colliding with the core, the decoupler heads (the part that's ejected) will still break off from the jettisoned LFBs.

I finally found a way to fix all that within stock's ability, and decouplers seem to work now with the same abilities as pre-v0.24.2. For people who prefer to go without add-ons, there are still other design workarounds that can get the job done (as Geschosskopf so deftly pointed out.)

Cheers,

~Claw

[herbal space program]

FWIW, I dealt with this issue by putting a little roll on my ship as I hit the staging button. That caused the ejected stages to move away from the rest of the ship enough not to knock its engine off. If I had already made my gravity turn, I also made sure the stages to be ejected were on the sides and not above/below the rest of the craft.

[davidpsummers]

If you have a SRB attached to a liquid booster, and you don't want to bother doing one of the proper fixes listed above, you can finesse it by cutting thrust, decoupling, and then powering back up (gently until you know how fast you can do it) to get buy the ejected boosters.

If you aren't under thrust, they won't hit the central stack enough to do damage and modest thrust will let you get clear of them safely.

I ran into this problem with parachutes on boosters (which drag the nose of the ejected booster straight down which can cause collisions as the bottom moves out). But perhaps I was blaming a more general problem on the parachutes.

[James_Eh]

[*]The radial boosters, except for the smallest SRBs, generally need struts to hold them on during use. However, no matter how big your radial boosters are (including tall 3.75m liquid stacks), you only need at most 2 struts per booster, 1 at the top and 1 at the bottom, connected to the central stacl..

Wait a second... this would tend to suggest that my general practice of 50 struts per booster is maybe not optimal...