Rockets flipping over around 10-20k alt, 200-300 m/s.

[Jadis]

I solved this problem with smaller rockets by just using the T-45 engine, but with larger rockets, I can't seem to stop the rocket from just flipping over. What am I doing wrong? I thought maybe I was trying to gravity turn too fast but it happens even when I go straight up. My rocket is pretty simple it's just 3 stages. Is it just bad design? Thanks.

[ChrisHale]

My current solution is to make sure the flipping stage has a gimballed engine and tail fins at its base. Might not be the best approach but it's worked so far.

[Streetwind]

You have too much drag at the front and too little at the back, and you're going too slow.

Try placing fins at the very bottom of the stage that's burning when the flip-over happens. Fins produce little to no drag when flying straight, but as soon as the rocket tries to flip, the fins are turned into the airstream and start dragging hard. This rights your rocket back up.

Also, more boosters Between 10 and 20 km you should already be at least 45 degrees pitched over, likely more. If you're only going between 200 and 300 m/s at that point, your rocket will simply fall out of the sky like a spear thrown by a nerdy elementary school kid. Even in the old "souposphere", you'd have been too slow; now with the more realistic, less soupy aerodynamics, you're even more too slow. Try roughly doubling your speed!

[Jaeleth]

Add 4 basic fins at the bottom of that stage. Increase reaction wheels, large ones, preferably, to make it more controllable. It has worked for me.

[nikolay-spb]

Add 4 basic fins at the bottom of that stage. Increase reaction wheels, large ones, preferably, to make it more controllable. It has worked for me.

Bad idea, my testing showed that fins make it even worst. Best way is to make rocket without fins, with alot of vectoring engines and RCS thrusters on the top, and make turn very slowly.

On some rockets you might need to disable SAS or it will destroy your rocket.

I think current situation with this flip is not realistic, i think that rocket in reality is easier to fly than rocket in KSP

[KerbMav]

Adding fins to the second stage always feels weird/wrong in respect to the first stage being below it ... ?

[Looney]

Have you looked at your centre of mass on that stage? If it sits too high, try adding a fuel tank to het it more toward the engine, perhaps that will help with stability.

[rakutenshi]

For stages that will burn out in atmo, fins with control surfaces and disable the gimbal, flipping usually means topheavy/front dragheavy, fins can balance out that quite a bit, almost always works for me, also try to make your TWR at least 1.5ish at takeoff.

on larger rockets my first two stages will often have fins, stages that are going to be burning above about 35km or higher can forget the fins, though. (totally opinion based)

[Jaeleth]

- - - Updated - - -

Bad idea, my testing showed that fins make it even worst. Best way is to make rocket without fins, with alot of vectoring engines and RCS thrusters on the top, and make turn very slowly.

On some rockets you might need to disable SAS or it will destroy your rocket.

I think current situation with this flip is not realistic, i think that rocket in reality is easier to fly than rocket in KSP

It works with me... All the time, besides, it is the logical choice; of course, the size of the fins does matters, try to change that. A lot of vectoring engines is not the most efficient or reliable way, something must be wrong in your design, turn forces overlay and see where those big red arrows are coming from. In 1.0.2 a rocket should be nothing more than a cilinder maybe with a slightly bulkier section a the top to enclose the payload and, at most, with a ring of boosters... ONE ring of boosters... If it's more than this, it's not a good design. If you need an ugly mass of engines stacked sidways like a pancake, to launch a payload, redesign it so that it can be launched in parts and assembled in 0 gravity.

[nikolay-spb]

- - - Updated - - -

It works with me... All the time, besides, it is the logical choice; of course, the size of the fins does matters, try to change that. A lot of vectoring engines is not the most efficient or reliable way, something must be wrong in your design, turn forces overlay and see where those big red arrows are coming from. In 1.0.2 a rocket should be nothing more than a cilinder maybe with a slightly bulkier section a the top to enclose the payload and, at most, with a ring of boosters... ONE ring of boosters... If it's more than this, it's not a good design. If you need an ugly mass of engines stacked sidways like a pancake, to launch a payload, redesign it so that it can be launched in parts and assembled in 0 gravity.

In reality large fins should be able to keep rocket steady and on course very easy in thick atmosphere (1-20km on earth).

Vectoring engines do fine job also, but dont depend on the air density, because of that they are choice on all modern rockets.

Helper engines are introduced to keep course even more steady.

Take a look @ design of Russian Proton-M rocket, it is most dependant rocket in world, there were sutuations when center of gravity was alot moved respectivelly to center of thrust and rocket did it job because of 6 vectoring engines and helper engines.

There should be really harsh conditions in order to rocket to flip in reality, i say that because flip model in KSP at moment is nowhere near to the reality.

Edited May 8, 2015 by nikolay-spb

[Looney]

How does one view the TWR before or during flight?

[Streetwind]

You cannot, in stock KSP. Either you install a mod like Kerbal Engineer Redux, or you calculate it by hand:

<total thrust> / <mass of rocket> / 9.82

For example, this early career rocket I used in 1.0 masses 9.8 tons and is propelled by a Reliant engine, which has roughly 201 kN of thrust on the launchpad:

201 / 9.8 / 9.82 = ~2.09

During flight, because of the rapidly changing mass and thrust, you probably don't want to bother calculating by hand.

[Old Foxboy]

For your motivation. look at 17:30 or 21:25 The pro's can do it, too.

What many don't seem to know: Gimbaling cannot only help against, but can cause flipping, too, since the centre of thrust moves out of line with the centre of mass. (This has caused many of the failures in the video seconds after ignition)

It's better to control the rocket with control surfaces as much as possible. Add more fins.

[Jaeleth]

In reality large fins should be able to keep rocket steady and on course very easy in thick atmosphere (1-20km on earth).

Vectoring engines do fine job also, but dont depend on the air density, because of that they are choice on all modern rockets.

Helper engines are introduced to keep course even more steady.

Take a look @ design of Russian Proton-M rocket, it is most dependant rocket in world, there were sutuations when center of gravity was alot moved respectivelly to center of thrust and rocket did it job because of 6 vectoring engines and helper engines.

There should be really harsh conditions in order to rocket to flip in reality, i say that because flip model in KSP at moment is nowhere near to the reality

The advantge of the fins is that the counter force grows as the rocket tries harder to flip, and the response is immediate, unlike vectoring, which is limited to a small angle and is constant force and depends on the PID used, read below.

To worsen things kerbal PID controller is not that good, many times it induces oscillation instead of stabilizing, having more reaction wheels may help counter this effect though, or flying with no SAS.

This is the result of too much vectoring. In fact, in kerbal, too much vectoring engines may destabilize the rocket instead of stabilizing it, I have some (poorer) designs on which I had to switch off most of the gimbals so that the damn thing could go straight up.

As one example from a rocket with both fins and vectoring thrust we have the biggest of them all so far, and yet never failed, saturn v

- - - Updated - - -

Adding fins to the second stage always feels weird/wrong in respect to the first stage being below it ... ?

Yes it does... But then again, placing a "light bulb" shaped payload, packet with tons of gear at the top of a candle-like rocket doesn't looks good either, and many kerbals do it... So... We have to add fins on second stage when needed... we kill one wrong with another... Eheh

[KerbMav]

Yes it does... But then again, placing a "light bulb" shaped payload, packet with tons of gear at the top of a candle-like rocket doesn't looks good either, and many kerbals do it... So... We have to add fins on second stage when needed... we kill one wrong with another... Eheh

My design department will find a way ... :cool:

[Jouni]

It's better to control the rocket with control surfaces as much as possible. Add more fins.

It's almost always better to rely on active guidance of an unstable rocket than to increase the mass and the drag of a rocket to make it stable.

Fins are like training wheels. They can help you pilot a rocket, if you can't fly it well enough otherwise. But if your piloting skills (or autopilot) are up to the task, fins just reduce the overall performance. Real rockets generally don't use fins these days, unless they're intended for low-altitude flight, because the guidance systems have become good enough.

[Wanderfound]

Echoing others:

You need to keep your CoL below/behind CoM throughout the atmospheric portion of the flight in order to maintain aerodynamic stability. Heavy bits at the front, draggy bits at the back: think of a dart or an arrow.

Adding fins to the bottom and removing draggy things from the top will lower the CoL. You may also be having CoM shift problems; tanks drain from the farthest first, so a rocket made of many small tanks will shift its CoM downwards during flight. Using the right-click menu to lock off a few of the upper tanks in the early stages of the flight may be helpful.

Second stage fins can be helpful. For example, this rocket has fins on both first and second stage. If they were on the first stage only, it would lose stability after dropping the boosters; if they were on the second stage only, the CoL would be above CoM during the booster burn, again causing a loss of stability.