Why does it seem like I can't get radial decouplers to play nicely anymore?

[hamez]

Ever since the most recent update nearly every rocket I build has terrible collision problems when I try to use radial decouplers in the design. Rockets that would have worked perfectly before now seem to collide with parts upon ejection. It's aggravating and no amount of sepratron use seems to help.

[AlexinTokyo]

This is a known issue with the latest couple of updates.

You can download a fix from Claw's stock bug fixes thread.

[WuphonsReach]

SpaceY mod has decouplers with integrated Separatron rocket motors. Useful for pushing the SRBs away from the core stack.

Alternately, make sure that your SRBs are positioned properly on the radial decoupler. If they tilt inward on separation, move the SRB downward. If they tilt too far outward on separation, move the SRB upward slightly. Once you have adjusted the SRB positioning as far as possible, add (4) separatrons to each SRB (2 up top, 2 towards bottom), pointed outwards away from the central stack.

You'll have to do repeated tests to get the Separatrons properly balanced.

[Dr-Drunk]

I'm using KW rocketry so I've got a different set of boosters and sepratrons but the mechanics are largely the same. I had some issues with boosters falling back in towards the spacecraft as well but I ended up putting sepratrons on the top of the booster facing up. If you adjust their thrust limiter and ammount of solid fuel you can set them up to provide just enough push to drag the nose of the booster away from the craft. Creates a nice Korolev Cross effect as well that way.

You could also go for the more brute force method described by Wuphons though. Basically just yank the booster away once it's spent. Also... If you're using tt-70 radials, don't use the offset tool. This seems to place the decoupler inside the hull of the rocket wich basically always ends up in the decoupler exploding once fired.

[Claw]

As AlexinTokyo pointed out, radial decouplers are bugged. So no, you aren't going crazy.

Any of the above suggestions can work, depending on your preferred route. There is also some tweaking you can do inside the .cfg file that helps a bit. If you want to try that, I can give you details.

If you want to do the design/sepratron route, it's good to know this tip-in effect gets much worse as you approach 750 m/s. At 700-750 m/s, it can be bad enough that the decouplers can rip themselves off. Above 750 m/s, the game changes a bit and decouplers behave less bad. So if you are going to separate decouplers, it's best to be lower in speed, or faster than 750 m/s (mixed with the sepratron suggestions above).

I might be a bit biased, but my add on fix will make the decouplers act as they did pre-v0.24.2. So if you aren't opposed to add-ons and this bug is causing problems, I recommend giving it a try.

Cheers,

-Claw

[Alshain]

Just curious. Do we actually know for sure this was a bug and not an intentional change to force people to use sepatrons? Has anyone from Squad commented on it or is it just our assumption that it is a bug?

[wibou]

Ah! So I wasn't the only one having problem with that... although I didn't think it was a bug, just that they were totally underpowered.

I always use the much more poweful "Hydraulic Detachment Manifold" instead of any radial decouplers to avoid trouble, usually quite a lot of them on the same part.

As example, when I use SRB-KD25k, I make sure they are held by 3 HDM each; it's generally sufficient to throw the part to a safe distance. I also find that stopping other engines prior to firing the decouplers provides extra safety (for some reasons) at the cost of some speed.

The only downside from using that much force to decouple (outside the cost) is that anything attached to the part usually break from it. So any winglets or nose cones WILL be sent flying and sometime collide with the rocket core... but they are light enough not to cause damage, so it's good enough for me.

[Jarin]

Just curious. Do we actually know for sure this was a bug and not an intentional change to force people to use sepatrons? Has anyone from Squad commented on it or is it just our assumption that it is a bug?

It's inconsistent physics, with the decoupler velocity dropping to 0 and thus pulling the entire SRB inwards on separation, so if it was intentional, it was a bloody weird choice. Also, the issue completely vanishes above 750m/s.

[RW-1]

I never thought about it as a bug, until seeing this, but it did force me to use the sep's (which blows when you are in early career).

If you place two sepatrons right about where the decoupler is, I've found that to work ok (instead of four, two at top and bottom of the stage to be jettisoned) after numerous trial and errors, mainly errors hehehe.

[Claw]

Just curious. Do we actually know for sure this was a bug and not an intentional change to force people to use sepatrons? Has anyone from Squad commented on it or is it just our assumption that it is a bug?

It took me many months to convince the appropriate people that something was actually wrong. This bug was introduced in v0.24.2 when attempting to fix a bug with stack decouplers in v0.24.1. I think in the case of this bug, it became a lesser of two evils between broken stack (that didn't work at all) or broken radial decouplers.

Ironically stack decouplers are actually broken too. It's just that it is much less noticable because it doesn't result in catastrophic failure like the radial decouplers.

Cheers,

-Claw

[Alshain]

It took me many months to convince the appropriate people that something was actually wrong. This bug was introduced in v0.24.2 when attempting to fix a bug with stack decouplers in v0.24.1. I think in the case of this bug, it became a lesser of two evils between broken stack (that didn't work at all) or broken radial decouplers.

Ironically stack decouplers are actually broken too. It's just that it is much less noticable because it doesn't result in catastrophic failure like the radial decouplers.

Cheers,

-Claw

I remember that. Only the stack decouplers were only originally broken in 64-bit, which is massively broken since 0.25, so you'd think they could go back on that now.

[RainDreamer]

well, you can shut down engine, then spin really fast before decouple, and those thing would fly away without having to use seperatron, if you are limited by part counts.