Irrecoverable tumble at 29000m-35000m

[nitevision92]

Basically what the title says.

I just built a rocket that's supposed to carry a rover to Duna, but every time the rocket reaches 29000m (I do a 45-degree gravity turn at 10000m) it starts to pitch back up to 0, then it goes past 0 and starts to tumble. Sometimes it keeps at the 45-degree line but does this weird thing where the flight direction moves along the line. I've looked at the ASAS- turns out it wasn't his fault; the fins were at full deflection and it was impossible to control. I have some screenshots if it helps.

In the VAB (Just in case something is wrong with my CoM/CoT)

Gravity Turn

Thrust Vector starts to wander off and the tumble starts......

....and what it did to the (originally) equatorial orbit.

Past the point of no return-tumbling and impossible to control

[7499275]

Have you tried turning SAS off and controlling the gravity turn and beyond manually? As that might have a part to play in it.

[matt_cain]

I notice that your tumble begins after you leave the atmosphere. Presently your using those tail fins as your control, correct? If so, that won't work after you leave the atmosphere. Consider a SAS wheel for torque, adding RCS, or (and I can't tell if you're doing this already) using thrust vectoring engines. Hope this helps and good luck.

[Syhrus]

At that height your control surfaces don't have much effect on the ship. It may have something to do with ASAS trying to use them instead of torque wheels.

EDIT: I just noticed you're going full throttle in all these pictures. Sometimes when I get a spin at high thrust, I find that throttling back a notch or two helps keep the rocket under control.

DOUBLE EDIT: It could also have to do with the positioning of your engines on the bottom of the tank (I thought you were using a mod, but those are just LV's right?) Look straight down the rocket in the VAB and make sure the COT is exactly in line with the COM

Edited June 20, 2013 by Syhrus

[Razor235]

As Syhrus said, control surfaces are fairly ineffective at that height. Adding an RCS system on will help drastically.

[UmbralRaptor]

That only has LV-T30s on the bottom? Nothing with TVC? You're critically short on control authority once you're out of the lower atmosphere...

[Pbhead]

You are using 3 bicouplers to make a 4 engine mount, yes?

I dont think the bicouplers are perfectly perfect. I have tumbling issues when I try to use them

[Vanamonde]

Try rebuilding your launch vehicle so that the upper stage is not a mass of fuel tank sitting at the end of a boom made up by the length of the NERVA. As you have it now, when that last orange tank begins to run low, you've got an inverted pendulum.

[stupid_chris]

That only has LV-T30s on the bottom? Nothing with TVC? You're critically short on control authority once you're out of the lower atmosphere...

This is the size of LV-30s, he spotted your problem. Replace the engines on your core stage to LV-45s and your problem is probably going to be fixed. RCS is expensive while launching so I don't recommend relaying on this. Else your Launcher looks perfect for the job, just switch for gimballing engines.

Consider a SAS wheel for torque

SAS wheels don't add torque for now. They can help you kill your rotation when you activate them but that's all, they don't help you turning faster or give more control authority. The only parts that give torque are for now command pods and probe cores. According to this week's Weekly KSP, those are being recoded/changed for the next update by C7 and actual reaction wheels will be added to the game and command pods will stop giving torque

[nitevision92]

Thanks for the help guys, here's what I have done so far:

1. Changed to 4x LVT-45 TVC in the center (I have been using the radial thingies with 4-way symmetry and the subassembly loader). This has mainly solved the problem, however I notice the craft still wants to tumble out of control (only ASAS managed to keep it at 45deg).

2. Changed the NERVA to a Rockomax Poodle in an attempt to solve the inverted pendulum problem noted by Vanamonde. No observable change- might just be me and my eyes though.

3.Checked CoM/CoT. Both are aligned.

In both cases the prograde vector still wanders, and the thing is almost impossible to orient towards the horizon at 100,000km Ap to start the orbit burn. Should I go with RCS thrusters now?

P.S one thing I noticed is that I never had any of these problems with my Mun rocket- and that one had a lander with big fuel tanks and all the shiny bits like crew ejection systems and whatnot, plus an even bigger version of the main lifter. I don't understand this.

Edited June 20, 2013 by nitevision92
spelling

[stupid_chris]

Yeah once actually in space, RCS finds it's use. You're going to have trouble orienting if your engines aren't burning once up there, so your best bet is an RCS system