[1.3] Kerbal Joint Reinforcement v3.3.3 7/24/17

[Zander]

Correct. I need to set tanks to update node position OnLoad(). I will get to it as soon as I can.

Also Procedural fairings seems to have the same problem with the expandable fairing base.

[Zander]

This is how serious it is

You can see those are all Heavy struts from the b9 and kw rocketry mod yet the procedural fairing base is still that far off the stretchy fuel tank

I even did this and the ****ing thing still flies like a slinky. This is the most frustrated ive ever been in this game XD

Edited December 7, 2013 by Zander

[Starwaster]

This is how serious it is

You can see those are all Heavy struts from the b9 and kw rocketry mod yet the procedural fairing base is still that far off the stretchy fuel tank

I even did this and the ****ing thing still flies like a slinky. This is the most frustrated ive ever been in this game XD

ok, the known issues aside, you've got something else going on because I use those same parts in something I whipped up yesterday and I didn't put a single strut on. Not a single one. I even stuck an interstage upside down on the bottom to give the engine a fairing (for FAR). Stuck fins on it and everything. It flies straight as an arrow.

[Guest]

For some reason, I've found struts to actually weaken KJR joints. For example, when testing my Energia rescale, I strutted it extensively, and it kept falling apart. First I piled struts on for quite some time, but after a while, I decided to remake the test article, this time I didn't put any struts on. And it worked! No wobble, parts falling off, nothing. It seems that struts are bugged in KJR, or something.

[Zander]

ok, the known issues aside, you've got something else going on because I use those same parts in something I whipped up yesterday and I didn't put a single strut on. Not a single one. I even stuck an interstage upside down on the bottom to give the engine a fairing (for FAR). Stuck fins on it and everything. It flies straight as an arrow.

Ohhh can you try my ship and see if you have wobbling? It normally starts at around mach 1 and gets more and more severe until the whole thing bends like a banana and the top snaps off

http://s000.tinyupload.com/?file_id=74288547311650535620

It uses nathans modular fuels mod with real fuels. and of course stretchy tanks and stretchy srbs KW and B9. i assume you already have those.

if anybody could figure out what the problem is that would be great

[ferram4]

@Zander: Yeah, the g forces and control inputs aren't helping there, and adding more than one strut between parts the same two parts really doesn't help; it's just the way the joints are coded.

Look, just try adding "StretchyTanks" to the list of exempt part modules in the config.xml and see if that helps (it'll probably do the opposite, but just to be sure). Then, fly with the smallest amount of control inputs you can; I've only seen things like that happen when I've gone overboard with control inputs.

The only reason you're having the problem with the procedural fairing base is that each and every one of them is attached to a stretchy tank; it's a source attribution error. If you used a different connection between them you'd get the same result.

The problem looks like it's a known issue with Stretchy Tanks (as has been said many times) and the high loads you're placing on the rocket.

@Dragon01: Yeah, over-constrained systems tend to have structural problems in physics simulations. You could get that to happen in stock KSP too, it was just slightly more difficult.

[Starwaster]

Ohhh can you try my ship and see if you have wobbling? It normally starts at around mach 1 and gets more and more severe until the whole thing bends like a banana and the top snaps off

http://s000.tinyupload.com/?file_id=74288547311650535620

It uses nathans modular fuels mod with real fuels. and of course stretchy tanks and stretchy srbs KW and B9. i assume you already have those.

if anybody could figure out what the problem is that would be great

Sure I'll check it out when I get a chance. I think though that some of what's being said though re: the struts is correct. Sometimes struts can work against you if you put too many even in stock without KJR. I've had rockets that just kept going to pieces so I put MOAR STRUTS! MOAR on them but they'd get worse until I just ripped them out and started fresh with just a few struts in key places. Sometimes it also jhelps to keep them short depending on what's being reinforced

[jrandom]

An interesting feature might be to dynamically raise the breaking tolerances very high as the ship settles on the launchpad and then scale back down to the settings in the config file. Might help with the insta-breaking problem I get occasionally without having to have unrealistically-strong joints during flight.

[ferram4]

Actually, there's already a version of that in my dev build; it scales the breaking forces and torques up to 10000 times what they start as for some set number of physics frames (currently 30) while locking control input so it can't be used to start launching monstrous things. It happens anytime a craft comes off rails, so it also helps with RSS-timewarp-related spontaneous explosions and some landed craft issues.

[jrandom]

Is the dev build available/playable? (Looks like I'm currently running 1.4.21)

[ferram4]

No, because I'd prefer to keep my clumsy mistakes to myself, thank you. It's not polished enough yet.

[jrandom]

Juuuust checking.

[jrandom]

Here's a weird one: In the rocket below, it flies nice and stable until I stage out the SRBs, then it goes all wobbly as hell. At first I thought this happened as I hit around 400m/s (maybe it's drag!), but if I throttle down the main engine so I stage the SRBs before I hit that speed, I get SuperWobble. If I leave the SRBs attached after they run out (moving the decoupler stage for them above the stage-out for the first engine and tank), it remains stable. And then when I stage out that first engine and tank, the stages above it go super-wobbly. I don't see any damage in the F3 nor the log file... is there perchance something that KJR does during a stage event that might cause this issue?

(Also, anyone happen to know which mod spams the log file with "Minimize status is: True"?)

Launches at 1.21 TWR. There are small rings of struts that connect the stretchy tanks to what's above them and what's below them to minimize the ST node attach OnLoad() bug. Out of curiosity I tried with no struts -- same behavior except it's noticeably more wobbly on launch.

Edited December 8, 2013 by jrandom

[wasmic]

Here's a weird one: In the rocket below, it flies nice and stable until I stage out the SRBs, then it goes all wobbly as hell. At first I thought this happened as I hit around 400m/s (maybe it's drag!), but if I throttle down the main engine so I stage the SRBs before I hit that speed, I get SuperWobble. If I leave the SRBs attached after they run out (moving the decoupler stage for them above the stage-out for the first engine and tank), it remains stable. And then when I stage out that first engine and tank, the stages above it go super-wobbly. I don't see any damage in the F3 nor the log file... is there perchance something that KJR does during a stage event that might cause this issue?

(Also, anyone happen to know which mod spams the log file with "Minimize status is: True"?)

Launches at 1.21 TWR. There are small rings of struts that connect the stretchy tanks to what's above them and what's below them to minimize the ST node attach OnLoad() bug. Out of curiosity I tried with no struts -- same behavior except it's noticeably more wobbly on launch.

I think the problem might be that you use StretchyTanks. KJR has some known problems with that mod.

[jrandom]

I think the problem might be that you use StretchyTanks. KJR has some known problems with that mod.

Yes, the joint is being calculated from the center of the tank instead of the attachment point. Something to do with the attachment node not being moved during a call on OnLoad() (in stretchytanks) or some such thing, which causes extra wobble. What's weird here is things not going wonky until a decoupler fires and I was curious if KJR recalculates anything when that happens.

[Starwaster]

Yes, the joint is being calculated from the center of the tank instead of the attachment point. Something to do with the attachment node not being moved during a call on OnLoad() (in stretchytanks) or some such thing, which causes extra wobble. What's weird here is things not going wonky until a decoupler fires and I was curious if KJR recalculates anything when that happens.

Check the log, it logs all that. I don't think I've ever seen that sort of recalculation going on but I never specifically lot watched for it either.

[jrandom]

I looked closer at another failure run, and the struts I have in place are actually keeping the stretchytanks attachments in place! The severe wobble is actually between the engines and the procedural fairing interstage part below them. And those connections don't get wobbly until I either decouple the SRBs, or keep the SRBs on and decouple the main engine/tank when it's done. Tried some more combination and the severe wobblyness never happens until that first stage-out, whatever part it might be.

I'll do a clean .log run and post that in a bit.

[a.g.]

When something is decoupled, KJR goes through the vessel removing joints that bridge two newly-created vessels, and its own decoupler strengthening joints that likewise appear to span vessels; without this the 'no decoupler force' bug happens. It might be removing too much in this case, similar to what probably happens with the launch clamps and the ground.

[jrandom]

Here's the .log from start to the decoupling and introduction of the wobblies.

Some mod is spamming the log with a few different things, making it trickier to work my way through. Time to learn how to use Notepad++'s regex find/replace, I guess..

[a.g.]

To get a useful log you need to enable debug in the KJR config file. Also, there is a different log file in KSP_data or something, which contains things like actual stack traces for NullReferenceExceptions, of which there are a few.

[jrandom]

When something is decoupled, KJR goes through the vessel removing joints that bridge two newly-created vessels, and its own decoupler strengthening joints that likewise appear to span vessels; without this the 'no decoupler force' bug happens. It might be removing too much in this case, similar to what probably happens with the launch clamps and the ground.

Interesting... so my upper stages' joint connections can actually get weaker when I drop lower stages? Ack! Is there anyway around this?

[a.g.]

Interesting... so my upper stages' joint connections can actually get weaker when I drop lower stages? Ack! Is there anyway around this?

That would be a bug. It should only remove connections that span the decoupler that is being activated, but there is a feature where it walks the vessel hierarchy further than direct children and parents to provide more strengthening, and this might cause it to prematurely detect activation of the upper stage decouplers if they happen to create 'invisible struts' to those boosters for some reason.

Edit: to stop it potentially walking too far, you can try removing all decouplerStiffeningExtensionType* entries from the config file. This would obviously reduce the number of invisible struts created by decouplers however.

Edit2: More specifically, given how things work currently, putting StretchyTanks on that list is a very bad idea that can easily lead to this weird behavior. The list should be formed so that in any situation not involving weird things like decoupler on top of another decoupler, there would be at least one part not on the list between any two decouplers. Otherwise, decoupling one of them would also remove strengthening from the other one.

Edited December 8, 2013 by a.g.

[jrandom]

KSP log and output log with KJR debug enabled. Here's to hoping the issue is both identifiable and easily-fixable! I so very much want to get past the "small rockets to orbit" stage of RSS/RO. If I can't even launch three small sats into geosynchronous orbit, how am I going to make rockets large enough to get to BigMun?

[Starwaster]

I looked closer at another failure run, and the struts I have in place are actually keeping the stretchytanks attachments in place! The severe wobble is actually between the engines and the procedural fairing interstage part below them. And those connections don't get wobbly until I either decouple the SRBs, or keep the SRBs on and decouple the main engine/tank when it's done. Tried some more combination and the severe wobblyness never happens until that first stage-out, whatever part it might be.

I'll do a clean .log run and post that in a bit.

That's so bizarre, last night I launched something with 3 of those interstates and it was so stable.

Interstages. DYAC.

[jrandom]

That's so bizarre, last night I launched something with 3 of those interstates and it was so stable.

Interstages. DYAC.

I use the interstages all the time for my pod-to-orbit launches I did to get used to the scale of RSS/RO and didn't suffer any problems. I'm just seeing this issue now with a taller rocket/heavier payload/more parts/more stages. If the entire rocket would just stay as stable as it is before I eject the SRBs, it would be perfect.

Even tried strutting the interstage base to the engine above it, to no avail. Instawobblies on first stage-out.