Airplane Guidelines

[PPR]

I learned a lot from my recent forum post of Guidelines for Space Stations and thought I might transition into another area.  Since airplanes have long been one of the more difficult designs in KSP, it seemed a natural choice.  So let's develop some guidelines for building an airplane.

But let me make clear the purpose of this discussion.  This is not meant to be a tutorial on plane design.  There are plenty of YouTube videos out there that do a wonderful job explaining how build a plane (especially those by Scott Manley).  The purpose is to develop guidelines to optimize an airplane design.

     op·ti·mize
     verb
     make the best or most effective use of (a situation, opportunity, or resource).

So what do you do to make your plane the best it can be?

Because airplanes may be designed to do different things, I would ask posters to state what they intend their guideline to help a plane do.  For example, a plane designed for speed would look very different from an airplane designed for high altitude.

If you share an example of your design, please include the usual info: mods, part count,  speed, altitude, endurance, etc...

While I am not exactly a noob when it comes to plane design in KSP, I must admit I spend most of my time with rockets and probes, so it is not a personal strength.  But I'll try to get things started with a few guidelines of my own.

 

For high speed: 

1)  Keep your thrust-to-weight ratio high.

2)  Your aircraft will generate a lot of heat from friction, so you will need to include parts that radiate large amounts of heat (engines, wings, pre-coolers, and radiator panels).   If you want to know how well a part cools, look in the .cfg file for " emissiveConstant = 0.95"

3) Small wings = less drag.

4)  Make sure your parts have a high temperature tolerance (internal & external)

[bewing]

Well, I think we need a bit more info. Are we talking SSTO planes? Or are we talking just basic atmospheric stuff?

Because if it's just atmospheric, you don't need any heat dissipation, even at mach 3. And the basic MK1 parts have plenty of heat tolerance. The constraints of atmospheric design are more: is it easy to fly (stable at all altitudes)? Easy to land? Does it look nice? Is it for long range? How much time are you willing to take to get to your destination? Can you drive it over terrain safely? Do you want parachutes with that? How many seats? Does it come in black?

 

[Rocket In My Pocket]

For atmospheric stuff, I'd say one of the most important things to double check is how stable it is at 4x warp speed.

Some planes will just shred when trying to time warp, and getting anywhere on Kerbin in a plane takes forever even with x4 speed.

Struts can help keep the plane stiff but will increase drag, often times weak or wobbly joints are the cause of a RUD but just as often it's poor aerodynamics or a plane that doesn't fly straight and true without control input.

[Wanderfound]

Hmmn, guidelines...

FAR assumed, but most of these will apply to stock as well:

 

* Roughly 400 units of LF (i.e. one Mk1 or short Mk2 LF tank) per jet to get up to light-the-rockets speed. Don't forget to pack extra if you're carrying nukes.

* If it has enough thrust to take off, it has enough thrust to go to space (assuming Whiplash/RAPIER).

* Low-tech Juno/Wheesley SSTOs are possible, but rarely practical.

* You need a bigger vertical stabiliser than you think, especially if you're headed for thin air.

* On the first takeoff, pull up hard shortly after you get into the air. You want the plane to almost-but-not-quite stall. Tweak your control authority up or down as required.

* Forward control surfaces stall much more readily than rear ones. Keep the canard deflection low unless you've already hit max deflection on the rear surfaces.

* AoA-responsive deflection can reduce stalling tendency: negative for canards, positive for elevators/elevons. Too much of it may aggravate SAS wobble.

* Speaking of SAS wobble: stiffen your airframe. Even a perfectly-tuned SAS will freak out when presented with a flexing airframe.

* Speaking of SAS tuning: PID tuners. Use 'em, they work. Set it up right and your SAS will be rock-steady. Kerbal Pilot Assistant has one.

* Speaking of mods, Dynamic Deflection can be used in a sneaky way to tune high-g aerobatic planes. Set the control surface deflections to maximum, then make use of the correlation between airspeed and dynamic pressure. Set up Dynamic Deflection to reduce deflection at low pressure but increase it at high. With enough fine tuning, you can get it so that full stick always holds your plane right at the edge of a stall, regardless of the airspeed. Great for supersonic dogfighters.

* The time acceleration trick mentioned above is worthwhile. Get into the air, carefully crank up the time acceleration and observe the plane. Reinforce anything that flexes. Big Mk3 monsters are especially susceptible to time acceleration induced detonation,

* If carrying nukes, light 'em at 20km and shut down the oxidising rocketry as soon as your apoapsis approaches 70km. Circularise on the nuke alone.

* Airbrakes make reentry much safer, much faster, and much more likely to reach KSC. With well-set airbrakes, you can (if you choose) come ripping in over the KSC mountains at high Mach, slam on the brakes and drop Stuka-style into a landing approach.

* However, airbrakes are not necessary. Cobra reentries and S-turns; they work.

 

Edited March 25, 2016 by Wanderfound

[qromodynmc]

As for the basic planes, Just check out real life planes, learn how they work, and build similar things to them.

As for space planes, learn making basic plane first, then add moar boosters till they go to space.

[Sma]

14 hours ago, PPR said:

The purpose is to develop guidelines to optimize an airplane design.

     op·ti·mize
     verb
     make the best or most effective use of (a situation, opportunity, or resource).

I like this, it's always nice to have something defined, just to make sure everyone is on the same page. That being said I don't think I really have much to contribute as I don't make planes very often. One time I managed to make an SSTO that could get into "space", and I think it JUST barely got into orbit, and unless I had a rather low space station it wouldn't have been able to dock and refuel.

3 hours ago, Rocket In My Pocket said:

Struts can help keep the plane stiff but will increase drag, often times weak or wobbly joints are the cause of a RUD but just as often it's poor aerodynamics or a plane that doesn't fly straight and true without control input.

I was under the impression that struts were both massless and physicless, meaning that they wouldn't effect drag at all, but I could be wrong.



At anyrate, a few things I could contribute (some may be obvious), for those new to planes and/or the game, when I've tested planes in the past, I usually try to put parachutes on them, so if something goes wrong, I can abort the flight and (if safe) deploy the chutes and recover (most) of the plane, and hopefully the passengers/pilot as well. Another thing (this is one of the obvious things), is to try and make sure your landing gear are near to the CoM/CoL, doing so will make it easier to lift off the landing strip before you reach the end. Of course, at the same time you also have to make sure when taking off not to smash your engines on the runway, I do that often lol.

Some of my best planes (which again, isn't something I do often as I haven't really worked out a good purpose for them, but may try to work with them more in 1.1), I didn't even need to touch the controls to make it lift off the runway. Just tilting the front of the wings up just a little bit tends to help it lift off the ground once it reaches the right speed. 

At anyrate, as frustrating as it can be to make a nice plane in the game, it is pretty rewarding once you get the hang of it.

[Rocket In My Pocket]

3 hours ago, Sma said:

I was under the impression that struts were both massless and physicless, meaning that they wouldn't effect drag at all, but I could be wrong.

Lol, I wish!

You can literally watch the Dv drop as you add struts.

I'm not sure if it technically adds drag to it's parent part or just has drag itself but it most certainly affects drag. As far as mass that's a no brainer, the game straight up tells you they weigh 0.5 a piece.

Part file relevant bits:

    mass = 0.05
    dragModelType = default
    maximum_drag = 0.02
    minimum_drag = 0.02
    angularDrag = 1

[bewing]

So, I don't build SSTOs. For guidelines for atmospheric only planes, I'd say:

Canard-based planes are significantly more efficient than standard tail-dragger types, which is why Burt Rutan always uses canards. So if you care about efficiency at all, then don't build planes that look like the ones you see out your window.

The canards in the game are too small, so tailfins make much better canards.

At high altitudes, even two tailfins are not enough canard to maintain good stability, so use four in a shallow X.

Planes that are meant to land on bad terrain should have at least 5 wheels -- the standard tricycle setup plus two more outriggers near the center of the plane (unless you pick good landing sites).

Learn how to use a Panther. Once you get to 8km altitude then level out and light the afterburner.

If you want fuel efficiency then always fly as high as you possibly can, as fast as you possibly can.

Don't bother with MK2 parts unless you are specifically going for looks, or you are building a seaplane. MK1 parts are better in almost every way.

Type B swept wings are very nice.

800 LF can get you a quarter of the way around Kerbin and back. 1200 can get you halfway around and back.

Retractable landing gear definitely reduces drag, has much better brakes, has landing lights (quite important at night), and allows your plane to "kneel" -- which makes embarcation/debarcation much easier.

Non-retractable nose gear steers much better!

All steerable landing gear uses an immense amount of electricity, so add a few more batteries!

An MK1 command pod makes a very nice airplane nose -- plenty of reaction wheel, very low drag, good electric storage, and a free ladder for doing "EVA report while flying over biome X" experiments.

Edited March 23, 2016 by bewing

[YargJay9991]

As a IRL pilot in training, reading all the flight theory books and learning the physics has actually helped me a lot in KSP.

Happy Flying

YargJay9991

[Rocket In My Pocket]

7 hours ago, YargJay9991 said:

As a IRL pilot in training, reading all the flight theory books and learning the physics has actually helped me a lot in KSP.

Happy Flying

YargJay9991

I fly IRL as well but honestly I never found the book side very useful except for passing tests. (I'm an avid reader so it's not a distaste for books in general.)

No amount of reading can replace actual seat time. I'd rather fly with someone who has 1000 hours of real experience and never read a single book, than someone who has read 1000 books and never flown.

"The difference between theory and practice is greater in practice than it is in theory"

[PPR]

On 3/22/2016 at 9:10 PM, bewing said:

Well, I think we need a bit more info. Are we talking SSTO planes? Or are we talking just basic atmospheric stuff?

Because if it's just atmospheric, you don't need any heat dissipation, even at mach 3. And the basic MK1 parts have plenty of heat tolerance. The constraints of atmospheric design are more: is it easy to fly (stable at all altitudes)? Easy to land? Does it look nice? Is it for long range? How much time are you willing to take to get to your destination? Can you drive it over terrain safely? Do you want parachutes with that? How many seats? Does it come in black?

 

I had intended to just cover air-breathing aircraft, but if you have some guidelines for SSTO's, go for it!

As for the heat dissipation, I'm talking about speeds in the neighborhood of Mach 4.5

Personally, I've always liked the aircraft design principle advocated by real-life pilots: "Does it look sexy standing still?"

[Triop]

I went on the internet and I found this....

 

[YargJay9991]

18 hours ago, Rocket In My Pocket said:

I fly IRL as well but honestly I never found the book side very useful except for passing tests. (I'm an avid reader so it's not a distaste for books in general.)

No amount of reading can replace actual seat time. I'd rather fly with someone who has 1000 hours of real experience and never read a single book, than someone who has read 1000 books and never flown.

"The difference between theory and practice is greater in practice than it is in theory"

Haha Cool! Pilot buddies!  I 100% agree. I'd rather fly with experienced pilots than experienced readers. Just for some reason, having that extra knowledge from both the books and actual flying has helped me build planes in KSP the right way the first time.

Anyway, better keep the thread on it's track.

Happy Flying

YargJay9991