Flipping out on Gravity turns

[Bishop149]

I'm trying to launch my first really BIG payload at 120 tons and I just can't get it to orbit

Many problem is that every time I try and start a gravity turn at 10-15 km up I just loose control of the bloody thing at it just falls all the way over.

Have now been through umpteen design iterations mostly involving increasing thrust, more reaction wheels, more gimbleing or control surfaces . . . . . nothing makes the slightest difference.

Best I've managed is not starting the turn until about 20-25 km up and then due to the massive inefficiency of the maneuver running out of fuel and thrust just short of orbit.

Its possible the payload is too unaerodynamic. . . but there is nothing I can do about that, faring isn't an option payload is too big.

Any tips appreciated, starting to tear my hair out

[Alshain]

If your payload is unbalanced or unaerodynamic you need to wait till about 20km and just bring more fuel. It seems like you figured that out already. We need a picture of the craft to tell you more.

[Mikki]

Maybe you are going to fast, try your assembly launching with a TWR around 1.5, and don`t exceed 1.5 - 1.8 probably untill you reached 20000km, then slowly turn orbit prograde... now you can throttle up as much as you like... not very efficient but safer.

Sometimes i do it so, gigantic things take their time.

[ForScience6686]

I would suggest you're being too aggressive in your gravity turn. I always start my gravity turn right off the pad, and for bigger payloads I wait until above 30k before any aggressive maneuvers.this way I'm still gaining horizontal speed. Another method I've used is excessive fins, I've used large wing segments and control surfaces to give more control authority. The bigger the payload the greater the aero forces effect it, even high up in the atmosphere.

[The_Rocketeer]

The problem is probably because your payload's drag force is too high compared with the rocket's drag force plus steering force (which could be reaction wheels, control surfaces or engine gimbals, or combination of these. Top drag is unstable, whereas bottom drag is stable (like a thumb tack standing on it's flat end vs on it's point). It goes away at higher altitude because the drag falls off as the atmo gets thinner.

I'd recommend, instead of standing the payload on top of a big rocket, try slinging it between two slightly smaller rockets. That way the drag force becomes much lower relative to the CoT and steering force, which give it much less leverage to tip you over. It's a bit like the difference between sitting on a unicycle vs sitting on a swing.

[Slam_Jones]

Personally, I'd start my gravity turn much earlier. I usually start it around 1,000m and continue til reaching 45 degrees at 10km, then 30 degrees at 30km, flat at 45km, 5 degrees below horizon at 60km. Keeping my TWR between 1.3 and 1.4 til at least 30km, then upping to 1.5 - 1.6 or so.

[The_Rocketeer]

Personally, I'd start my gravity turn much earlier. I usually start it around 1,000m and continue til reaching 45 degrees at 10km, then 30 degrees at 30km, flat at 45km, 5 degrees below horizon at 60km. Keeping my TWR between 1.3 and 1.4 til at least 30km, then upping to 1.5 - 1.6 or so.

Yes, the issue in the OP is that turning earlier results in the rocket flipping. TWR might be part of the issue, but I think drag is more likely.

[FlyingPete]

Have you considered more unconventional ideas such as forward fuel tanks? The idea being to position the troublesome payload further aft, with fuel lines connecting nose-mounted drop tanks to the main engine. As tanks become empty, they are jettisoned as normal (you'll need sepratrons to move them out of the way). Might make the thing a bit more controllable but will need more struts to brace the fuel tanks above the payload.

[Empiro]

If you can't put a fairing at the top, then keeping your rocket from flipping will be difficult, as all the drag is concentrated on the front. You can try turning off SAS as that might cause wobbles to amplify. Too much TWR can also cause your rocket to flex and then flip.

[ghpstage]

@Bishop149,

Could you post some pictures of your rocket?

Yes, the issue in the OP is that turning earlier results in the rocket flipping. TWR might be part of the issue, but I think drag is more likely.

The OP mentions failure when starting a gravity turn at 10-15km, which is far from early, and not losing control when done at 20-25km where the atmosphere is practically non-existent. Added to the fact that the rocket can get to that altitude while travelling straight up without flipping strongly suggests an overly aggressive gravity turn is at least part of the problem. Edited September 8, 2015 by ghpstage

[The_Rocketeer]

The OP mentions failure when starting a gravity turn at 10-15km, which is far from early, and not losing control when done at 20-25km where the atmosphere is practically non-existent. Added to the fact that the rocket can get to that altitude while travelling straight up without flipping strongly suggests an overly aggressive gravity turn is at least part of the problem.

Based on personal experience I have to disagree. Excessive top-hamper (mass/drag) on a rocket means that vertical ascent (and I mean NO deviation) is the only viable flight profile until at altitude where the atmospheric forces have decreased. Even slight deviations result in a flip. Aggressive gravity turns could have a similar effect, but considering the payload here I think the issue is probably as I've described.

[ForScience6686]

Based on personal experience I have to disagree. Excessive top-hamper (mass/drag) on a rocket means that vertical ascent (and I mean NO deviation) is the only viable flight profile until at altitude where the atmospheric forces have decreased. Even slight deviations result in a flip. Aggressive gravity turns could have a similar effect, but considering the payload here I think the issue is probably as I've described.

If you start your gravity turn immediately and follow prograde through the atmosphere you should have no problems. Increase control authority via more or larger fins to increase drag at the rear, should correct any deviation from prograde. If you can't get it up, look into reducing payload size or breaking it up into multiple launches. Or launch vertically until above 30k, but you will need much more fuel to get to orbit.

Aero forces are still relevant above 20-25k, which is why I keep my rescue craft inside a fairing until above 30k as the lift from the wings will flip it immediately.

[Tank Buddy]

You should start your gravity turn when you reach 100 m/s and tilt over ~5 degrees, following prograde vector. If you follow prograde exactly and your initial turn is good, you should have absolutely no problems. What is your initial and final TWR? Can we have pictures? Do you have fins on your rocket?

[The_Rocketeer]

If you start your gravity turn immediately and follow prograde through the atmosphere you should have no problems. Increase control authority via more or larger fins to increase drag at the rear, should correct any deviation from prograde.

You'd think so, but this isn't always the case. Top hamper has massive leverage compared with any amount of fins at the bottom of the craft - even if you do have absurd amounts of tail fin, chances are they're adding so much drag that your fuel requirements have increased significantly anyway, and there's still always the chance that your control authority just results in a craft that bends instead of one that just drifts off course. Aero-forces are still relevant right up to orbit altitude, but the magnitude of their input is hugely diminished above ~19000m. Late gravity turns are much easier to engineer than major design modifications, and from the sounds of it this is an exceptional payload.

[Bishop149]

If your payload is unbalanced or unaerodynamic you need to wait till about 20km and just bring more fuel. It seems like you figured that out already. We need a picture of the craft to tell you more.

Thank everybody, I'm pretty sure it's the brick like aerodynamics of the payload that is the issue, I did finally manage to get it to orbit by doing the above.

A little more fuel in the second stage and a slightly different ascent profile, started turn at 25km and also turned a little slower than usual (I normally just point orbit prograde and leave it there) to allow a higher apoesis and more time for the somewhat sluggish upper stage engines to burn through their deltaV.

I'd recommend, instead of standing the payload on top of a big rocket, try slinging it between two slightly smaller rockets. That way the drag force becomes much lower relative to the CoT and steering force, which give it much less leverage to tip you over. It's a bit like the difference between sitting on a unicycle vs sitting on a swing.

This was what I was going to try next, the design would have been an absolute pain though as the sides of the payload aren't straight.

Anyway pictures below, happy to call this one "Answered"!

If anyone's interested in what it is, it's composed mostly of KSP-I parts primarily 8 huge atmosphere scoops.

It's designed to sit just above the Kerbin's atmosphere and skim off nitrogen as part of my orbital fuel production operations

Edited September 9, 2015 by Bishop149

[nekogod]

Would it not have been more aerodynamic and a lot easier to fly if you'd just flipped the payload upside down?

[Bishop149]

Slightly perhaps, but the other end is still a 5m almost flat surface . . . . . And I'd have had to turn it round after it separated to apply thrust which I don't think I would have had time for, was cutting it fine for "time out of the atmosphere to circularise" for the final burn as it was.

[Sharpy]

I believe what this thing needs for balance is the Pinecone Nosecone.

You may even give it an even bigger variant on the Kerbodyne adapter.

[Clipperride]

Although I only play with a stock install, those atmosphere scoops look like a lot of drag. Can they be closed like an engine intake?

My "to try" list for payloads that flip;

Start a gentle turn at 10km - 300m/s.

Be ready with the "a" key to stop it turning too quickly.

Always try and use a aero fairing.

Then I try adding overside fins at the back (I've had good results with the space plane tail fin)

Add some vernier engines for more control (although I like the Pinecone Nose above!)

Make the stage that's firing when it flips wider, with radial tanks and engines.

I'd also be tempted to turn the payload through 180° and see if you can add a nose cone to the flat end.

Let us know how you get on and good luck. I've had a few strange payloads that I've really had to fly into orbit. Keeping the thing point more or less East as it fights to go its own way. IMO more fun than a text book launch ;-)

[Chakat Firepaw]

Slightly perhaps, but the other end is still a 5m almost flat surface . . . . . And I'd have had to turn it round after it separated to apply thrust which I don't think I would have had time for, was cutting it fine for "time out of the atmosphere to circularise" for the final burn as it was.

Even as is it would still be better than what you have, plus you could mount a nose cone to improve the aerodynamics further.

As for turning it around: Perhaps a set of thrusters on each end which are solely there to do the flip and that you can ditch as soon as you are around.

[Kryxal]

I'd probably go with a bit more delta-v to reach orbit in the "reverse" orientation, and maybe have the last stage still capable of deorbiting itself. Also use a nosecone that deorbits too.