A Simple Guide to Rocket Aerodynamics!

[BagelRabbit]

Hello! My name is UpsilonAerospace.

I've had loads of experience with model rocketry, so I have a fairly good general idea of How Aerodynamics Work. However, it seems as if many new players of KSP are a bit weirded out by the aerodynamics that were introduced in 1.0 and subsequent patches.

Thus, I'll take this opportunity to give an overview to aerodynamics as a whole. I'm going to start out rather simple and then get more technical as I go, so just start reading wherever something catches your eye that you may not have known before, okay?

Good. Let's get started.

A Brief and Simple Guide to Rocket Aerodynamics!

(I thought I could make this guide shorter. Perhaps I was overly optimistic with my title... oh well.)

~~~

Let's first get some simple terminology out of the way:

The center of mass (or center of gravity) is a rocket's horizontal balance point. There is an equal amount of mass above the C.G. (as it's abbreviated in model rocketry) as there is below it.

The center of pressure (or C.P., as it's abbreviated in model rocketry) is a little more difficult to find. All bits of your rocket have a certain amount of drag, and if you took the location and amount of drag of every part and averaged it all out, the average point would be the center of pressure. Fortunately, there's an easier way to do this in Kerbal Space Program, as we'll soon see.

~~~

So, where should the CG and the CP be on your rocket?

I'll use the example of a badminton birdie (or shuttlecock). Here's a friendly picture:

Now, as anyone who has played badminton knows, the birdie flies with the bit on the left pointed in the direction it's traveling. If you tried to throw it while it was pointed backwards, it would correct itself so that it was pointed forwards again. In other words, it's 'stable.'

Of course, it would be nice to have your rocket pointed in the direction it's travelling. If it wasn't, it would lead to instability: flipping and possibly crashing. You don't want that to happen.

So, where's the center of gravity and the center of pressure on a badminton birdie?

The front of a badminton birdie is most definitely heavier than the back: the center of gravity is almost all the way at the front of a birdie. Because the back of a birdie has a lot of drag, its center of pressure is located a bit further back. In other words, the center of gravity is in front of the center of pressure to make the birdie stable.

A stable rocket must have its center of gravity in front of its center of pressure. It must have enough drag on the bottom to move its center of pressure behind the center of gravity. Some people think that this is counterintuitive: they think that the rocket should be bottom-heavy, so that the bit with the most weight would be pointed towards the ground. But that's simply not true.

NASA agrees with me here, so I think I'm doing something right:

...oh, and by the way, badminton birdies make great model rockets, though they're a bit too draggy to win any competition awards.

~~~

How can you get the CG and the CP to be in the right place?

There are several important things to note here:

• A tube-shaped rocket with a nose cone on top will be unstable. The CG will be at about the middle of the rocket, but the CP will be near the top: even the pointiest nose cone will still be more draggy than any other part of the rocket.
  
• The faster you go, the worse the problem gets if your rocket is unstable. Or, to be more precise, the more airflow over the rocket, the worse the problem gets. In the lower atmosphere, rockets have an especially hard time with this.
  
• It's much easier to move the CP than it is to move the CG. While you have to completely rearrange your rocket's layout to change the CG, you only have to add some special bits called "fins" onto your rocket to change the CP. (More on this later.)
  
• However, the CG moves around quite a bit during flight! As the tanks drain, the rocket's center of gravity gets lower and lower, causing it to be less and less stable. This is worrisome, and you should keep it in mind.
  
• Placing very draggy bits at the bottom will actually help! If you put a big ol' set of structural panels or wing bits at the base of your rocket, arranged to be as high-drag as possible, they will make your rocket more stable (by lowering the location of the CP). They will also be as draggy as all get-out, so you probably shouldn't do this if you want your rocket to go to space. There are better ways of doing this, folks!
  

~~~

What are fins, and what do they do?

Fins are protrusions outside the rocket's airframe. Their goal in life is to lower the center of pressure on the rocket. Therefore, the most ideal fins are rather small and as far towards the bottom of the rocket as possible.

Fins keep the rocket stable without much drag in a very clever way. When the rocket is travelling straight up, fins have very low drag. However, if the rocket begins to tip slightly to one side, the fins' drag increases, making the rocket briefly more stable. They also exert a force on the base of the rocket ("Lift") that causes the rocket to remain pointed up. Lift plays a far larger role in keeping your rocket pointed in the right direction than drag does, but both are quite capable of helping.

Here's another diagram from NASA. (This one is an expanded version of the precious one.)

Some configurations of rocket need larger fins than others. For example, rockets with large fairings have a lot of drag up top, and thus they need larger fins to bring the center of pressure down to the right spot. In addition, the sharper the nose cone (and the less drag) up top, the lower the need for large fins.

To properly stabilize the rocket, at least three fins are needed. Most rockets use three or four fins; there's really no reason to use any more than six.

~~~

Okay, but how can you tell where the CG and the CP are?

There's currently a very good way to find your rocket's CG: click the button at the bottom right of the VAB that says "Center of Mass." Finding your rocket's CP is a bit harder. The indicator that says "Center of Lift" does not give your CP. As a matter of fact, it doesn't give any data at all if you don't use the wing bits found in the 'Aerodynamics' tab! Thus, I would personally recommend using a part from the aerodynamics tab as your fins.

Once you have placed your fins, you can pull up the 'Center of Lift' tab and the 'Center of Mass' tab. If your 'Center of Lift' is one caliber or more behind your 'Center of Mass,' you should be good. A 'caliber' is the width of your rocket's fuel tanks: if you have a 1.25m tank, for example, you need to place the fins so that the 'center of lift' is about 1.25m further down the rocket. (You can eyeball this if you want.)

If your rocket is too stubby to have the center of lift a full caliber behind the center of mass, that's okay. Your rocket still should fly safely, though it may not have as good of a margin of stability.

~~~

How do fins affect your rocket's flight path?

Fins' ability to make your rocket more stable lends to some interesting characteristics. Most notably, rockets with fins consistently point prograde (in the direction the rocket is travelling). If your prograde marker is slightly off-vertical and you turn your SAS off, the rocket will begin to very gradually tip in that direction. Because your thrust vector will no longer be vertical, your prograde marker will move further towards the horizon. Your rocket will, theoretically, tip until it is ninety degrees from vertical and then start descending... if you don't design it properly.

In real life, this happens quite frequently if a model rocket launches into the wind or without sufficient thrust. It's called 'weathercocking.' Here's a picture of this effect, originally taken by JCRocket.com:

To combat this, you can do several things. One is to place fins that can swivel on your rocket. These fins will do a better job of keeping the rocket pointed at exactly the angle you want it to. Another thing you can do is add some SAS. Finally, you can throttle up all the way or otherwise increase the amount of thrust your rocket produces relative to its weight: its 'thrust-to-weight ratio.' This will ensure that the rocket does not tip as quickly, because the amount of thrust is basically pushing you further upward before you tip too much more.

Note, however, that if you can get your rocket tipped very slightly early in the flight so that it is nearly horizontal when you get to the upper atmosphere, you've done a great job creating an 'orbit-friendly' flight profile! Because your rocket was pointing prograde the whole way up, your burn to orbit will be quite efficient and it may not even require the SAS unit! In other words, this rocket behavior can be annoying, but it can also help.

~~~

Do you need fins on your radially-mounted boosters?

Likely not, at least with shorter boosters.

Radially-mounted boosters are usually mounted near the base of the rocket. This actually moves both the center of gravity and the center of lift towards the base. If your rocket is stable without boosters, it should be okay with boosters. You probably don't need larger-than-normal fins on the main stage or separate fins on the boosters. However, if your rocket wasn't that stable to begin with, you might want to consider adding fins onto the boosters and/or the main core to ensure stability.

~~~

Do you need fins on each stage of your rocket?

It depends.

If your rocket has multiple stages that fire within the thick of the atmosphere, each one should have its own set of fins. The fins on the lowest stage should almost certainly be the largest and most prominent, with every subsequent stages' fins getting smaller and smaller. Don't forget that you're raising the center of pressure significantly with each set of fins.

If you're firing your upper stages at 30km or higher, especially with some SAS on them, you don't need fins. Your rocket will fly fine without them.

~~~

Is there any reason to put fins at the top of the rocket?

Not really. The only good reason you would do something like this is either if you had an upper stage or if you wanted to build a scale model of a real rocket.

~~~

Are there other ways to make my rocket fly properly?

Sure!

Option 1: Engine Gimbal

Pros: Probably applies to most rockets you build, meaning you don't have to add extra parts. Somewhat effective, especially at lower velocities.

Cons: Less effective (to completely ineffective) at higher velocities. Requires an engine with gimbal capability. The larger the gimbal, the greater the percentage of fuel wasted while burning in a direction other than straight up. Gimbal may overcorrect, resulting in wobbly rockets.

I should note that many people have kept their rockets quite stable with a combination of engine gimbal and SAS (seen below). If you're a more advanced player, you shouldn't have much of a problem with a finless rocket.

However, doing this requires you to manually point your rocket within five or ten degrees of prograde throughout the ascent. If you're using this, you should always keep your SAS on as well: turning it off will quite likely cause your rocket to flip over.

Finally, fins are all-but-required if you have a solid rocket booster (SRB) or an engine without a gimbal for your first stage. Engines without gimbal are unable to correct your rocket's position when it deviates even a tiny amount from prograde, thus making your rocket far more likely to lose stability. You can check your engine's gimbal range by right-clicking on the engine's icon on the left side of the VAB.

Option 2: SAS

Pros: Allows for complete control at all levels of atmosphere, whether the engines are burning or not. Rather nice and simple.

Cons: Rather heavy, reducing amount of payload that can be carried into orbit. Can be twitchy.

Option 3: RCS

Pros: Allows for acceleration in different directions, both in the atmosphere and in space.

Cons: Way too heavy. Depletable. Mostly pointless. Why did I even add this option?!

Option 4: Spin Stabilization

Pros: Rather simple. An interesting real-life concept. Can make unstable rockets stable without adding much mass.

Cons: Impossible to orient oneself. Impossible to do a gravity turn. May waste fuel by spinning. Causes Kerbal nausea. (No one likes Kerbal nausea.)

...yep, I'll go with fins, personally.

~~~

Is there anything else to say?

Probably. Let me know if you have any unanswered questions, or if you find inaccuracies within this guide! I'll try to address your concerns as well as I can.

I would like to note that this thread is basically an extension of the "Why does my FLIPping rocket Always Flip Over!" thread. I think I covered some ground that EtherDragon didn't. I actually started this tutorial only hours after 1.0 was released, but unfortunately I haven't been able to post it until now, due to computer problems. Sorry.

Oh, and I'll likely make a short video that covers much of this information!

Thanks for reading this rather long post. I hope it's helped you with building better rockets. If you learned something new, I suppose some Reputation would be nice, though I don't really need it.

Best of luck!

-Upsilon

Edited May 17, 2015 by UpsilonAerospace

[NathanKell]

Awesome writeup!

One nitpick: torque isn't a way to make your rocket statically stable, it's a way to make it do what you want. There's a difference.

(And indeed, too much stability can be bad too--makes it harder to turn)

[Legendary Emu]

Thank you so much for writing this! Now I understand why every one of my attempts to put a spaceplane atop a rocket flips over mid flight.

[Eric S]

I think I'll be using a link to this page for people that don't get why their rocket is flipping or even the ones that argue that a high CoM leads to tipping. As is, I'm linking the simpler explanations several times a day (mostly on reddit).

[Watoh]

Thanks for the write up, definitely cleared a few things up for me!

[Rokmonkey]

A good write up, this should help all those naysayers to learn the new aerodynamics model.

I will comment however, fins are not the only GOOD answer. There are no fins on the Falcon or Atlas rocket families. Another strategy to recommend making stable rockets, is to setup the staging of a rocket such that the center of mass always move upwards.

Very long, minimally staged rockets with payloads lighter than the engine (approximate of course), have the center of mass move towards the bottom of the rocket. By placing an upper stage in there, as the fuel drains from the first stage, the CG moves towards the front of the rocket. And this is the same for each subsequent stage.

A rocket design considers two things now. Boosters to get off the ground, and balancing each stage of a rocket through the ascent.

Edited May 4, 2015 by Rokmonkey
I no spell good

[Jouni]

Rocket fins are kind of like training wheels. They're useful when you're still learning to fly, but once you can reach orbit reliably, they're usually just extra mass with no real function.

[Wesreidau]

Rocket fins are kind of like training wheels. They're useful when you're still learning to fly, but once you can reach orbit reliably, they're usually just extra mass with no real function.

Unless your fins double as landing legs. Gotta love those 50's comic-book Nerva ships.

[BagelRabbit]

Awesome writeup!

One nitpick: torque isn't a way to make your rocket statically stable, it's a way to make it do what you want. There's a difference.

(And indeed, too much stability can be bad too--makes it harder to turn)

I tried to make your first point relatively clear, though I do suppose I said "Are there any other ways to make the rocket stable?" instead of "Are there any other ways to make my rocket fly properly?"

Thanks for the positive feedback! I'm sure you know more about this than I do...

Thank you so much for writing this! Now I understand why every one of my attempts to put a spaceplane atop a rocket flips over mid flight.

Yep. You can place a spaceplane on top of a rocket, but you need some really big fins on the bottom to lower the center of pressure. Otherwise, that thing will be absurdly unstable.

I think I'll be using a link to this page for people that don't get why their rocket is flipping or even the ones that argue that a high CoM leads to tipping. As is, I'm linking the simpler explanations several times a day (mostly on reddit).

Thanks for the write up, definitely cleared a few things up for me!

Thanks! I hope this guide is a more-or-less complete look at what makes rockets stable, and what you can do to improve stability.

A good write up, this should help all those naysayers to learn the new aerodynamics model.

I will comment however, fins are not the only GOOD answer. There are no fins on the Falcon or Atlas rocket families. Another strategy to recommend making stable rockets, is to setup the staging of a rocket such that the center of mass always move upwards.

Very long, minimally staged rockets with payloads lighter than the engine (approximate of course), have the center of mass move towards the bottom of the rocket. By placing an upper stage in there, as the fuel drains from the first stage, the CG moves towards the front of the rocket. And this is the same for each subsequent stage.

A rocket design considers two things now. Boosters to get off the ground, and balancing each stage of a rocket through the ascent.

See, the problem with this is, no matter how high the center of gravity is, the center of pressure is always higher than the center of gravity. This is because the nose is generally the main source of a rocket's drag, so it raises the CP almost all the way to the top. Sure, having the CG rise is nice, and it is possible with a reasonably large upper stage. But it still results in an unstable rocket.

The best stability test you can do is to take your rocket up to about 5km, turn off the SAS, and then press W,A,S, or D for a second. If the rocket corrects back to near-vertical, it's stable. If it continues tipping, it's either 'neutrally stable' or 'unstable.' I'll bet you my own $0.02USD that you can't create a rocket that passes this stability test without placing something at the base of your rocket that adds drag (or turning on the SAS or otherwise cheating).

Oh, and I'll also comment that if the Falcon or Atlas's guidance system failed, they would be unstable. It's engine gimbal that keeps them pointed in the right direction, not aerodynamic stability.

Rocket fins are kind of like training wheels. They're useful when you're still learning to fly, but once you can reach orbit reliably, they're usually just extra mass with no real function.

Eh, not necessarily true.

Fins can actually increase a rocket's efficiency by pointing it in the right direction as it goes to orbit. If you know exactly what you're doing, it's possible to angle the rocket slightly after launch and then just take your hands off the controls as the rocket gradually tips towards horizontal and gains altitude. I have yet to achieve the perfect burn, but I've regularly ended up at about ten degrees from horizontal at 35km, travelling at 1500 or 1600 m/s, with my nose right at the prograde marker the whole way up. No control input, no SAS, just plain ol' aerodynamics. The weight cost for this very nice ascent was only about 0.06 tons.

I would strongly recommend giving this sort of thing a try and seeing whether you can do it too! It saves Delta-V on the way to orbit and it's immensely satisfying if you can do it properly.

Anyway, thanks for your comments and suggestions!

-Upsilon

[Jouni]

Fins can actually increase a rocket's efficiency by pointing it in the right direction as it goes to orbit. If you know exactly what you're doing, it's possible to angle the rocket slightly after launch and then just take your hands off the controls as the rocket gradually tips towards horizontal and gains altitude. I have yet to achieve the perfect burn, but I've regularly ended up at about ten degrees from horizontal at 35km, travelling at 1500 or 1600 m/s, with my nose right at the prograde marker the whole way up. No control input, no SAS, just plain ol' aerodynamics. The weight cost for this very nice ascent was only about 0.06 tons.

You can get the same efficiency with active guidance as well.

I experimented with passive guidance when I started using FAR, but it didn't work very well in practice. Most interesting payloads are large and relatively lightweight. In order to get the center of pressure low enough, the fins have to be gigantic.

[Rokmonkey]

Oh, and I'll also comment that if the Falcon or Atlas's guidance system failed, they would be unstable. It's engine gimbal that keeps them pointed in the right direction, not aerodynamic stability.

Oh, absolutely. But we aren't making model rockets here. Early career/science, fins are very helpful, some might say necessary. However, we have gimbaling engine and guidance later on so that is another efficient method. You've dismissed that entirely valid and in many cases preferred method far too easily. By not adding fins, you reduce weight and lower the drag footprint increasing available delta-V.

[WhirlwindMonk]

Thanks for the information! It helps explain why one of my rockets kept tipping over terribly. The giant lander behind a faring up top gave it nice, high CoM, but without anything to correct the CoP, it flipped the heck out every time.

However, we have gimbaling engine and guidance later on so that is another efficient method.

It should be efficient, but is it really efficient as implemented in the game? Maybe there are other issues with my rocket designs, but it seems like when I try to let my SAS follow prograde, it either overcorrects so badly the rocket flips out, or it does so many dozens of tiny corrections every second I have a hard time believing it's efficient (though maybe it is, I'm certainly no expert).

[Jouni]

It should be efficient, but is it really efficient as implemented in the game? Maybe there are other issues with my rocket designs, but it seems like when I try to let my SAS follow prograde, it either overcorrects so badly the rocket flips out, or it does so many dozens of tiny corrections every second I have a hard time believing it's efficient (though maybe it is, I'm certainly no expert).

The gimbal ranges of the 1.0 engines are too high for SAS, if you try using anything other than stability assist. I've got good results with limiting thrust vectoring to 20-50%, depending on the rocket.

[Rokmonkey]

It should be efficient, but is it really efficient as implemented in the game? Maybe there are other issues with my rocket designs, but it seems like when I try to let my SAS follow prograde, it either overcorrects so badly the rocket flips out, or it does so many dozens of tiny corrections every second I have a hard time believing it's efficient (though maybe it is, I'm certainly no expert).

I can't speak to that. I only turn on Stability Assist and manually follow the prograde marker, I don't not have it set to follow prograde. I believe Jouni hit the nail on the head, they have large gimbal and it over corrects. You can try turning on fine controls, or reducing the gimbal range (is that a stock tweakable?)

By just using Stability Assist, you have lots of control over the ascent profile. I have seen many cases where following the prograde would put my rocket in a nearly flat trajectory at 18-22km, and by correcting my rocket to a steeper ascent I had a far more efficient launch (This is with and without fins on the same rocket).

[WhirlwindMonk]

The gimbal ranges of the 1.0 engines are too high for SAS, if you try using anything other than stability assist. I've got good results with limiting thrust vectoring to 20-50%, depending on the rocket.

Interesting, I'll give that a shot. Thanks!

I can't speak to that. I only turn on Stability Assist and manually follow the prograde marker, I don't not have it set to follow prograde. I believe Jouni hit the nail on the head, they have large gimbal and it over corrects. You can try turning on fine controls, or reducing the gimbal range (is that a stock tweakable?)

By just using Stability Assist, you have lots of control over the ascent profile. I have seen many cases where following the prograde would put my rocket in a nearly flat trajectory at 18-22km, and by correcting my rocket to a steeper ascent I had a far more efficient launch (This is with and without fins on the same rocket).

That's what I do, the problem is that I'm only okay at it. If there's another option, regardless of whether it's fins or reducing gimbal, I'd like to at least try it out and see if it can do it better than I can manually. Plus, building a rocket that can very nearly fly itself into space simply by thrusting and letting physics do its thing sounds really dang cool.

[Meithan]

Great write up, very clear and complete. I've sent some rep your way.

As an aside for people willing to discuss the finer points of this, I'd like to add that while it's true that both the lift and drag of the fins contribute to the restoring torque in a stable rocket, I'm under the impression that it's the lift of the fins that generates the dominant contribution.

I can think of several reasons for why this could be so. For small deviations from the direction of flight:

(a) the lift of the fins is perpendicular to the centerline of the rocket, which favors torque production;

( the drag is nearly parallel to the centerline, which is unfavorable for torque production;

© the drag area of the fins is not much larger than when the rocket is pointing in the direction of flight, so the form drag doesn't increase much;

(d) for typical airfoils, lift increases faster than induced drag (for small angles of attack)

So the only way for drag to be the dominant contribution would be for the drag force to be considerably larger than the lift force that can be produced. However, typical wings have lift-to-drag ratios on the order of 10, so the lift-induced torque would tend to be considerably greater for the reasons stated above.

I'm no aerodynamics expert so please do correct me if I'm wrong; I bring this up to hopefully improve my understanding.

[Mulbin]

This is great! Very useful. Also had a chuckle as I hadn't realised Americans call shuttlecocks "birdies"... presumably for the same reason you call male chickens "Roosters"

[Crystal_Mace]

This is great! Very useful. Also had a chuckle as I hadn't realised Americans call shuttlecocks "birdies"... presumably for the same reason you call male chickens "Roosters"

Shuttlecock is a much more fun word! I always call them that, though I am a silly American, and to MY sensibilities, calling the male chicken a rooster always sounded like you were talking about the hen!