KSP (Stock) Aerodynamics

[Kobymaru]

Hello friends.

I have been more or less successful in building Spaceplane-SSTO's, I have two in my hangar which get me up to LKO with one/three Kerbenauts as payload.

While I'm pretty proud of my achievement, I wanted to step up my game and bring some knowledge into the development process (like we all did when we learned about the rocket equation and Delta-V).

So I tried to dig up some ressources on how aerodynamics behave in KSP, but I failed. Here's what I want to know specifically:

• The drag is dependent on the square of the airspeed. But it is not dependent on the Angle of Attack, is it? So the drag is the same from every direction?
  
• Drag is not (really) dependent on which parts are used, because almost all parts have a drag coefficient of 0.2. It does not add up if I am correct?
  
• Is drag different in subsonic/supersonic speeds?
  
• How does lift behave?
  
  • Is it dependent on the speed? If yes, how?
    
  • How does it depend on the air pressure?
    
  • Does the center of lift move backwards in supersonic speeds as in real aerodynamics?
    

In stock KSP, I lack the diagnostic tools to build a proper craft, and I am not quite ready to install Ferram Aerospace just yet.

Note that I could probably look it all up in some Real-World textbooks/ressources, but if I understand correctly, KSP aerodynamics are vastly different from Real-World aerodynamics.

Edited July 3, 2014 by Kobymaru

[capi3101]

http://wiki.kerbalspaceprogram.com/wiki/Drag#Drag - that should answer your first three questions succinctly.

To be thorough:

1) Drag is independent of AoA

2) You do use the average drag coefficient, which in the stock game is 0.2 for almost every part - so much so that you can in general use 0.2 in all cases and get away with it.

3) Drag is calculated the same at both subsonic and supersonic speeds

As for lift, that's another story; to be honest I'm not sure how it's calculated.

Edited April 16, 2014 by capi3101

[Claw]

- Parasitic drag doesn't depend on AoA, but induced drag does.

- There is a lift bug with control surfaces. Their lift increases all the way to 90 AoA.

- Lift production is dependent on arspeed, AoA, and air density. The equation is consistent throught the flight envelope. So no, the center of lift does not move (unless you envoke the control surface bug). I have equations if you really want them.

- Drag (as the wiki shows) is a weighted function. Basically part mass times the drag value. So things like nose cone drag are vastly overshadowed by fuel tank drag.

- Part drag is independent of placement. So big flat pancake rockets have the same drag as tall sleek rockets.

In my opinion, KSP does a reasonable job of approximating aerodynamics enough to have reasonable aircraft behavior. It does have some quirks and a limited flight envelope (super and hypersonic aren't really modeled), but it does okay. I run all stock aero, but most people who want "hardcore" aero use FAR.

[Kobymaru]

Thanks for your information! That's helpful.

- Parasitic drag doesn't depend on AoA, but induced drag does.

Excuse my ignorance, but what is the difference?

- Lift production is dependent on arspeed, AoA, and air density. The equation is consistent throught the flight envelope. So no, the center of lift does not move (unless you envoke the control surface bug). I have equations if you really want them.

Equations would be really nice, thank you. How did you figure them out?

- Drag (as the wiki shows) is a weighted function. Basically part mass times the drag value. So things like nose cone drag are vastly overshadowed by fuel tank drag.

- Part drag is independent of placement. So big flat pancake rockets have the same drag as tall sleek rockets.

In my opinion, KSP does a reasonable job of approximating aerodynamics enough to have reasonable aircraft behavior. It does have some quirks and a limited flight envelope (super and hypersonic aren't really modeled), but it does okay. I run all stock aero, but most people who want "hardcore" aero use FAR.

So far, I really dislike stock aerodynamics. Weighting with mass and drag independent of placement sounds really stupid, IMO. I sincerely hope Squad fixes aerodynamics some time along the way.

I'm hesitant about FAR, because I read that it makes the launch easier and I want to stay away from rescaling mods. But since I'm growingly disappointed with stock aero, I might reconsider FAR.

[Jimbimbibble]

FAR does knock some delta V off the launch but it makes the rocket more difficult to fly and it forces you to fit stuff into fairings. In FAR, if you pitch your rocket over 45 degrees it will spin out and the center of lift actually matters. I think you should try it for a week and see how it goes.

[Claw]

You are very welcome.

Parasitic drag is drag due to air flowing over the surface of an object. For example, a flat plate going through the air edge first has less drag than when it goes through face first. And a ball has less drag than a cube.

Induced drag is a byproduct of creating lift. Some component of the lift is always acting opposite of the direction of airflow. As AoA increases, more of the lift vector points backward which effectively acts as drag.

I don't have the equations with me at the moment, but I can post them later if someone doesn't beat me to it. I actually didn't figure out the equations, but several other forum members did.

I am very confident squad will update the aerodynamics. The game is only v0.23.5, so there is certainly a lot of work to do.

I'm happy enough with the way stock aero behaves now that I accept it's faults. FAR fixes your two big complaints. Use it if you want, ditch it if you don't. Some people get so angry with stock aero that they wouldn't even play without FAR. I suppose I'm a little more forgiving.

I avoid mods too because I grew tired of crashes. Also, I didn't want all of my craft to be at the mercy of waiting for mod updates, especially when the game updates. All I use right now is MechJeb and Kerbal Alarm. Ironically, MJ is causing me grief at the moment in v0.23.5.

Don't get me wrong, there are a lot of awesome looking mods out there and I don't want to belittle anyone. I just seem to crash a lot more often with the ones that I did try.

[Supernovy]

From what I recall, lift is linearly proportional to velocity while drag is proportional to velocity squared. This may contribute to infiniglide, with lift dominating drag and low speeds and vice versa at high speeds.

Craft will attempt to orient so that the lowest-drag parts are facing toward the airflow and the highest-drag parts away. This can be seen with re-entry and parachutes. Undeployed chutes have 0.22 drag, while capsules have 0.2 drag, so the end with the parachute points away from the airflow.

Drag is indeed a weighted sum, so adding a large mass of low drag parts will increase your terminal velocity. Note that some aircraft cockpits and all wings have low drag, so aircraft tend to have lower drag than rockets. Solid rocket boosters have high drag, too. RCS thrusters have low drag.

[Claw]

Thanks for your information! That's helpful.

Equations would be really nice, thank you. How did you figure them out?

Ask and ye shall receive...

Lift equation for Control Surfaces: sin(AoA)*pressure*speed*lift

Lift equation for Wings: sin(AoA)*(1-|sin(AoA)|)*cos(AoA)*pressure*speed*lift

From what I recall, lift is linearly proportional to velocity while drag is proportional to velocity squared. This may contribute to infiniglide, with lift dominating drag and low speeds and vice versa at high speeds.

I believe infiniglide is a bug induced by an error with the control surface lift equation (as can be seen above). Control surface lift continues to increase until 90 deg AoA (sin 90 = 1). So if you flutter them, part of the lift component is thrown forward and generates ridiculous amounts of thrust.

If you put a small control surface at 90 degrees, you can make an airplane that uses an ION engine for thrust. That's one of the wacky KSP aero effects.