Sooo, why exactly rockets are flipping?

[0x7be]

Flipping rockets and one of little annoyances of 1.0. Why exactly this is happening? I'am talking about physics of that process.

Is this considered a bug or not?

[CaptainKipard]

No. You're probably forcing it to to turn too far from your prograde vector.

[Randazzo]

Flipping rockets and one of little annoyances of 1.0. Why exactly this is happening? I'am talking about physics of that process.

Is this considered a bug or not?

The draggiest end of the rocket will always flip behind the heaviest end. It's not a bug.

Go try and throw a dart backwards and report your results.

Of course without seeing the design you're talking about, you're likely only to get general reponses.

[sal_vager]

Looks like a gameplay question so I'll move it.

[Kerr_]

Not a bug.

Long cylindrical shapes (like rockets) like to fly/fall on their sides because the aerodynamic forces are even. use f11/f12 to see what's happening.

We counteract this with fins, thrust vectoring, and weight distribution.

PS I'm trying to find a mythbusters episode where they put bullets in a wind tunnel

Guns compensate for this with rifling (imparting a significant spin rate on the bullet)

[cicatrix]

LOL I explained this same thing just today.

Nearly all problems with lifters start at the moment of gravity turn. You want to steer this thing very gently and few degrees at a time. Basically, if you don't have fins the only way to steer your rocket is the engine gimbal which is usually not enough. You can enhance your steering authority by adding RCSs, reaction wheels or aerodynamic control surfaces (fins). What happens is as you start steering your rocket it has too much momentum to immediately change its heading so it continues flying upwards but this time the aerodynamic pressure on the top of your rocket forces it downward which results in increasing the 'angle of attack' which in turn exposes more area to the airflow. It snowballs after that.

To avoid that you should have something that will oppose the increasing aerodynamic pressure on the side of your rocket. As I said it can be RCS, reaction wheels or fins. The heavier your rocket the more steering authority you should have in order not to flip.

Also, keep in mind that TWR greater than 1.3 is bad. If it's ~2 or more then the chances of your rocket to flip increase up to nearly 100%.

Many who used stock aero before 1.0 still do the same thing - up to 10 km then 45 degree east and up till the desired height. This doesn't work this way now. I used FAR before and it it taught me the proper way of doing that.

1. Have fins + reaction wheels. Always.

2. Have TWR less than 1.2 (max 1.5)

3. Start turning earlier (6-7 km).

4. Turn slowly, wait your prograde marker to respond to your steering before continuing.

[0x7be]

Sooo.... if I get it right, real rockets also struggle with this problem?

Realistic solution would be to use smoother ascenstion profile.

- - - Updated - - -

I don't like fins. Not that they are bad from engineering point of view. But real world rockets somehow go without them and I like my KSP rockets to look realistic.

[sal_vager]

Yeah, as smooth as possible helps.

Early rockets had fins, and they are great for the first stage, even the Saturn V and N1 had fins.

[cicatrix]

Sooo.... if I get it right, real rockets also struggle with this problem?

Realistic solution would be to use smoother ascenstion profile.

- - - Updated - - -

I don't like fins. Not that they are bad from engineering point of view. But real world rockets somehow go without them and I like my KSP rockets to look realistic.

You can get by without fins. They just help in KSP. In real world the rockets are turning very smoothly and very slowly. Also SOME of the rockets actually have 'fins' of sorts:

[Harry Rhodan]

You can get by without fins. They just help in KSP. In real world the rockets are turning very smoothly and very slowly. Also SOME of the rockets actually have 'fins' of sorts:

But the Falcon 9 only uses them to land and the Proton for the LES.

[Agate]

cicatrix wins the thread.

But real world rockets somehow go without them and I like my KSP rockets to look realistic.

Real world rockets have a computer that's a lot better at keeping the rocket pointed in the direction of the airstream than you can manage piloting by hand.

[sal_vager]

Use other things then, as well as reaction wheels you have vernors, gimballing main engines, and the Twitch and radial Ant have much bigger gimbal now, letting them act like real world vernier engines.

[cicatrix]

But the Falcon 9 only uses them to land and the Proton for the LES.

Yes, modern rockets can fly without fins, but they have very sophisticated flight control computers that balance thrust, gimbal angles and more. They also have to fight wind, non-uniform fuel burn, etc (you can check out my blog about difficulties that engineers have to fight with in real world http://forum.kerbalspaceprogram.com/entries/3200-Hidden-complexity-of-rockets-Part-1).

As I said if you're smooth you can get by without fins. They simply make your life easier.

[waterlubber]

Play 1.0, which has aerodynamics that make sense. 1.0.2's aero SUCKS.

[grawl]

Hello,

I have not played 1.0, only 1.02, and only around 1 hour.

I noticed the rocket flipping stuff, and as a casual stock player it surprised me ! After a few adjustments on thrust limiter... Who would have thought boosters could need LESS thrust ??? ... It is now going smooth as cruise

waterlubber, what were the differences in 1.0 ?

[Nutter]

1.0 has less drag and reentry heating is more intense, but really, I think the former applies more to planes than rockets, and those did look a bit over the top in 1.0.

And as for the latter...well, you can probably still use that heat shield to make sure you won't die horribly in your capsule, but I heard planes are a mite easier to land by comparison.

[EtherDragon]

Flipping rockets and one of little annoyances of 1.0. Why exactly this is happening? I'am talking about physics of that process.

Is this considered a bug or not?

No, it is not a bug and all due to real-world phenonema: Here's a handy Tutorial in the Tutorials sub-forum, that explains it all, with pictures!

Also -

on how to safely fly your finless rocket into Orbit on "National Don't Use Fins Day." =D Edited May 15, 2015 by EtherDragon

[Laie]

You don't necessarily need to have fins... but your rocket will be a lot easier to control if the center of drag is well behind the center of gravity. Fins are a lightweight method of creating drag when and where you need it the most. Other solutions require precise steering and/or a lot of control authority through gimbals or RCS. I think it's safe to say that fins are far and away the easiest approach, especially when using WASD controls.

I much prefer Kershan's picture to illustrate the problem:

[kzauner]

LOL I explained this same thing just today.

1. Have fins + reaction wheels. Always.

2. Have TWR less than 1.2 (max 1.5)

3. Start turning earlier (6-7 km).

4. Turn slowly, wait your prograde marker to respond to your steering before continuing.

And, if you are limited regarding tanksizes (as you still need to unlock some of them):

5. Try locking the topmost tank, so your COG shifts not that much down.

But don't forget to open the valves before you run out of fuel (just bind it to an action-group not used for rockets, for instances 'brakes').

An option 'invert fuel flow logic for this stage' would come in handy with the new aero.

[Leszek]

I fly with no fins on my KSP rockets all the time.

It isn't hard to keep near the center of your velocity vector. It is just a matter of getting your gravity turn right. I think people are still launching with 1.5 or 1.7 TWR. You would have to start your gravity turn right away almost to get that to work.

I use 1.25 starting TWR and do my 5 degree kick at 70 meters per second. After that I just follow the center of the velocity vector. You can lag your VV to slow down your gravity turn, or lead to increase but I stay well within the circle the whole way. I use a 20km AP by 45 degrees as a milestone. (I want to have a 20km before I reach 45 degrees.) It isn't until I reach 30 kilometers in altitude that I leave the VV circle and when I do I leave it smoothly.

When I first switched to NEAR (and later FAR but there was not a big difference) I had issues with my rockets flipping all the time. That is until I had experience doing it. With the new aero both V1 and V1.0.2 I have had almost no issues. It is very similar to how NEAR was and my launch profile is not much changed.

Really I think the root cause was that the stock aero wasn't changed earlier and now everyone has to deal with the learning curve at release instead of before the bad habits got firmly intrenched. I made the switch to NEAR after only a month or so after I got the game so the vertical to 10k with a 45 degree turn never had a chance to become normalized.

P.S. Just to be clear, you don't need fins, you don't need extra reaction wheels, you don't need to worry about your CG. It is just a matter of practice to get any reasonable rocket to orbit without flipping. I never worry about any of these things, and the fins, even the small ones makes it harder. Sure you don't flip, but you can't control your gravity turn easily.

Edited May 15, 2015 by Leszek

[eataTREE]

I am having an enormous problem with this. While my smaller rockets fly just fine using the advice presented, I cannot seem to launch any large payload ( > 90t) in an aeroshell to orbit successfully. They all fall victim to uncontrollable spin in the high transonic range.

Fins? Got 'em. Reaction wheels? Got 3. TWR? I have acceleration clamped down to 12.5m/s. Gravity turn? Starts at 125 m/s, 3 km, just like it's supposed to.

Nevertheless, my rocket tumbles out of control around 500 m/s no matter what I do at that point. Even if I'm > 30,0000 km, suddenly aerodynamic forces become uncontrollable.

Any ideas? This is becoming truly frustrating. I'm already chopping up my payloads into tiny pieces, now I can't even get one piece into orbit!

[moronwrocket]

I think it is because your tanks are way too long on a single rocket. As they drain, they drain from the top, and so the heavy back end can resist the effects of drag better than the empty, light front end. So the rocket flips. Try adding a middle stage to your boosters?

[Empiro]

Bigger rockets should be more stable than small ones (aerodynamic forces have relatively less of an affect compared to small rockets).

However, there are a couple of possibilities:

-As you burn fuel, your CoM is shifting, and is moving too low

-You're pointing too far away from prograde

-Your rocket is flexing, and bending too much in the airstream

Adding more struts to the top of your rocket can help. As can turning off SAS. I've found that SAS offten causes wobble as it tends to over compensate. You don't need it at all. A well designed rocket automatically will point itself into the velocity vector.

[Motokid600]

I am having an enormous problem with this. While my smaller rockets fly just fine using the advice presented, I cannot seem to launch any large payload ( > 90t) in an aeroshell to orbit successfully. They all fall victim to uncontrollable spin in the high transonic range.

Fins? Got 'em. Reaction wheels? Got 3. TWR? I have acceleration clamped down to 12.5m/s. Gravity turn? Starts at 125 m/s, 3 km, just like it's supposed to.

Nevertheless, my rocket tumbles out of control around 500 m/s no matter what I do at that point. Even if I'm > 30,0000 km, suddenly aerodynamic forces become uncontrollable.

Any ideas? This is becoming truly frustrating. I'm already chopping up my payloads into tiny pieces, now I can't even get one piece into orbit!

No that's incorrect. That's too fast and too high. I personally don't go by speed when it comes to my pitch maneuver. I go by sea level thrust to weight ratio which I usually set around 1.5.

At that TWR I start the turn between 3 and 400 meters. So almost instantly after leaving the pad. The lower the sea level thrust the later and softer the turn. The higher the thrust the earlier and harder the turn. ( not much harder. Like an extra degree )

No you don't need fins, no you don't need reaction wheels and no you don't even need verniers. What you NEED is to stay inside that little yellow circle. The trick is getting that yellow prograde market to fall at the right pace. Turn too early and it'll fall to fast and the rocket will flip trying to fight it back up. Turn too late and the market won't fall fast enough and again the rocket will flip trying to get it where it wants to be.

Only once your above 30km or so then you can play around a little more.

-Your rocket is flexing, and bending too much in the airstream

This too. Make sure when your done building a rocket, before you hit launch for gods sake change the root part to the one closest to the center of mass. Means the world for large builds.

Edited May 15, 2015 by Motokid600

[almagnus1]

The trick with getting to orbit is learning how to do the gravity turn, and here's what I'm doing:

1) Take off vertically as you normally would, and get to over 100 m/s surface velocity. Have SAS turned on.

2) Go two tick marks from top of the navball on the 90 degree heading, and make sure you're stable in that direction.

3) Once stable, turn OFF SAS, and let the physics engine finish pulling the rocket over as you continue to burn upwards. This starts the gravity turn.

4) Watch in the ascent as the prograde vector and heading drop towards the horizon, and turn ON SAS to lock in the heading if it's going down to fast, or use SAS to force the heading towards the horizon if it's not dropping fast enough.

5) Finish the circularization burn once you are apoapsis as you normally would.

I've found this works out rather well, as you basically let the physics engine deal with getting the rocket to orbit, and once you get a feel for doing gravity turns, it actually turns out to be decently fuel efficient.

It's a completely different way of flying from prior to 1.0, but it's actually easier on the ascent but actually a bit more engaging IMO as you are constantly monitoring the craft on the way up.

Edited May 16, 2015 by almagnus1