Rocket losing control

[UnthinkableThinker]

Well for starters make sure the brakes are activated. The wheels spinning won't help. But when you perform your gravity turn at 10km how far over are you pitching? If its too much too soon yea it'll fall right back down. With a craft that large.. delay your gravity turn until about 250m/s then SLOWLY start pitching over. You say the craft tumbles over after the first stage. Is this before ( ideal ) during ( not ideal lol ) or after the pitch?

Other then that.. idk. Honestly I'm sorry to say, but that rocket looks like a nightmare. The problem could be a little bit of what everyone said. Something like that is very hard to diagnose. You may be better off starting back at the drawling board. Look at your payload by itself ( the crane ) is the mass on center? Switch off gravity in the debug menu ( alt - f12 ) and see if there's any phantom torque. Make sure the payload is tested so you can rule that out.

Then build and test the rocket a section at a time. Sub orbital tests to orbital tests and so on. Also it might not hurt to take a look in the space craft exchange. Look at some heavy lift, asparagus launchers. Just to get some ideas.

Something I just noticed... Is that the space plane hangar you built that in? Correct me if I'm wrong ( been awhile since I've been in the SPH ) but your only allowed up to 2x symmetry. NOT good for rocket builds.

I will try your suggestions, and yes, that is SPH, but I did not build it in there.

[Person012345]

Just to make sure, you don't have any fuel lines that split, then eventually lead back together do you? Because IIRC the game doesn't understand those at all.

[SRV Ron]

Another thing to consider. As the tanks are drained, more stress is generated by the engines under them. That can cause twisting which can cause spin, and in extreme cases, random failure of connections.

[UnthinkableThinker]

@Person012345: No, no fuel lines.

@SRV Ron: I actually thought about that, it's why there's so many struts on each tank. Though thanks for the heads up.

Obs: Just did the "No gravity test" suggested by Motokid600, the ship actually makes it to orbit safe and sound, so the problem is really what I had already in mind. As the rocket goes up, the first stage is ejected (Solid Fuel Boosters), once the most outer fuel tanks run out, the rocket loses control even with control surfaces (wings). And it tilts to the side, meaning the CoM of the probe itself which was placed slightly forward in relation to the center isn't the one causing the problem.

Even not performing the gravity turn, heading straight up the ship loses control just by the "loss of mass" caused by the ejection of the outer tanks.

I will be testing now the rocket system itself, without the probe.

Edited October 16, 2013 by UnthinkableThinker

[Nurph]

Okie dokie, finaly got it to load properly. Let's see.

Javascript is disabled. View full album

1. It's definatly unbalanced, the boosters under the rover are not symetrical. The whole way up it wanted to yaw to the west.

2. The rover wont work anyway because some of the decouplers are placed upside down, so they will not decouple from the rover.

3. Stage 3 decoupled all the orange boosters, you can see how well that went. Not sure if that's how it was setup or if editing out the Mechjeb part from the craft file caused weirdness.

4. Suprisingly, no structural failures that were not my fault(forgetting to throttle up, staging things I shouldn't). Good work on the struts!

5. You have an excessive amount of monopropelent, you could do with much less.

I'm going to try building my own launcher, or at least rebalancing yours. I'll post more results latter. We are here to help you!

[Chrisd857]

In Nurph's first frame, you can clearly see a bunch of RTGs that are on the left and not the right, and that is at least part of the imbalance.

[Tank Buddy]

For once I'm going to say less rocket. You could definitely create a much smaller and more effective rocket to deliver a payload of that size.

[Awass]

Hi guys, I'm new to KSP and I really want to know why my rocket is going out of control at around 10km, just when passing the first atmosphere layer, during gravity turn. The rocket is carrying my crane which is pretty big since I'm trying to land it on the Mun so I can build a Station there. The entire ship weights around 100t.

Ask Whackjob. He's an expert when it comes to monstrous abominations of rocketry.

[MisterFister]

So, you are running a legacy version of MechJeb. I run v2.0.9. Please remove MechJeb and resubmit.

[Kerbart]

Ask Whackjob. He's an expert when it comes to monstrous abominations of rocketry.

Nownownownownow. To call that a monstrous abomination is a bit heavy handed, isn't it?

One other thing to think off: even when you manage to get COM and COT aligned (and preferably in the middle of the craft) there's a good chance that once you start the gravity turn the torque is so large that the rocket simply can't control it.

Controlling a rocket is very much alike to balancing a broomstick on your finger, with thrust vectoring (and winglets -- maybe you should add a dozen or two) being the same as moving your hand around. If you've ever done this balancing trick then you know that you'll need to keep the stick fairly upward; once it's past a certain point you will not get it upright. I wouldn't be surprised if this monster launch vehicle displays similar behavior.

Some well meant advice: if you're just starting, work your way up to larger vehicles. To me, this looks like going from couch potato to running the New York marathon. Why not start with something slightly less ambitious?

[Awass]

Nownownownownow. To call that a monstrous abomination is a bit heavy handed, isn't it?

Sorry. I should use the politically correct term: "massive, convoluted spaceship." Well that describes this rocket at least. I still maintain that Whackjob's creations are abominations that threaten the very existence of the Kerbal universe.

[MisterFister]

Incidentally, it's listed in my craft list (inaccessibly, mind you, due to a MechJeb version mismatch) as havign 779 parts. Seven hundred seventy nine. Even my computer would struggle to run that, and I have a pretty high-end machine that can run Crysis at full settings with 16xAA.

[Nurph]

Huzzah! The rover is on the Mun! Pictures and info to come tomorrow, I need sleep.

[UnthinkableThinker]

@Nurph: First of all thank you for trying it out and pointing out the flaws and especially for sending the pics. Like I've said previously, I'm pretty new to KSP, but nevertheless, that doesn't limit my imagination and the will to always want to go bigger each time .

Regarding your observations: 1) You're correct, they are not symmetrical, should be causing the problem;

2) That was my fault, haha, didn't notice I had commited that silly mistake ;

3) Ah, yes, now that you've said it, I thought that the "version" that I've submited had the stages already balanced out correctly, again, my fault;

4) Thank you, once you look at it on the hangar it seems like a lot, but really, they're all important;

5) Regarding the amount of RCS, as you could see, I've put extra RCS tanks on the rover itself so I could bring its CoM slightly forward, since I will be lifting heavy stuff from the back. And the RCS tanks on the rockets themselves is mainly for the purpose of trying to balance it out.

For the other guys, thank you for trying to help me so far, we are making progress.

[Steve-S]

One thing I see right off is the outer most orange tanks are the only ones with gimbeling motors. Once you drop those what is in control of the crafts heading? Is it enough? I usually turn off gimbles on the outer most engines and leave it on only on the center one. I've read, and found, that works well.

[UnthinkableThinker]

One thing I see right off is the outer most orange tanks are the only ones with gimbeling motors. Once you drop those what is in control of the crafts heading? Is it enough? I usually turn off gimbles on the outer most engines and leave it on only on the center one. I've read, and found, that works well.

All of the engines used have thrust vectoring.

[DRIVENES]

Anything in the flight log showing what failed?

[UnthinkableThinker]

There isn't a structural failure (things falling apart), so no.

[boomerdog2000]

Are you using Mechjeb for the ascent? And if so are you using its ascent mode or the SASS?

[Tassyr]

sweet mother of GOD that's a lot of parts. Your computer must be impressive to not just choke on that. As for the roll- I was having that problem, albeit with more... 'rocket' like rockets, and found that reducing the angle of ascent in the mechjeb ascent curve to about 65%75% seemed to fix it.

[NephilimNexus]

I can't really be sure from the picture, but are all of your engines perfectly level with each other? With so many on there it would be easy for a uneven engine placement to slip past. Even a tiny imbalance in thrust can grow exponentially into a complete spinout very, very quickly.

[UnthinkableThinker]

@boomerdog: Tried both things, manually first, then Mechjeb, but the problem persists.

@Tassyr: Yeah, my computer does struggle with that . In other hand, the problem isn't really the angle or anything, the main problem is that it is losing control right after the most outer tanks get ejected.

@Nephilim: I will double check the simmetry, but I'm sure they're all correct.

[antbin]

The rats' nest of struts in the middle is probably to blame. Somehow it's flexing under load and twisting.

This would be okay except that your outside orange tank mounts are flexing too much for the mainsails' thrust gimbaling to correct your spin. This happens to me all the time - mainsail thrusts left to correct your roll, decoupler flexes left and the effects cancel out.

You can fix it by re-working your strut design on the outer rockets to keep them from flexing. Never strut across like band-aids. - - - or | | |. Always strut diagonally, to make triangles. X es are best. Always put struts symmetrically or you'll make the spin worse.

See the last example in this post for a picture of what good strutting looks like.

[UnthinkableThinker]

@Antbin: Ok, I'll try re-doing the struts as you've suggested. Thanks!