The Aeroplane Doctor

[kiwinanday]

Planes not flying right? I can help turn your Christmas Bullet into a Spitfire. Post a screenshot with all three SPH indicators on (like below) as well as a short description of what it's doing or not doing, and what you want it to do. I'll read it over and do my best to offer some tips and tricks to get it flying purrrrrfect.

 

Some other examples of my planes:

Spoiler

 

All of these are excellent flyers, and perform exactly as the pilot commands. Prop engines from Kerbal Standard (nee Airplane Plus)

 

Edited March 26, 2017 by kiwinanday

[Guest]

Looks nice, I'll check it out when I get back to KSP

[TheRagingIrishman]

@kiwinanday in general, what should the com/col placements look like?

[qzgy]

18 minutes ago, TheRagingIrishman said:

@kiwinanday in general, what should the com/col placements look like?

Like, for most planes? CoL behind CoM, or on it, depending on want of maneuverability. Farther back, harder it is to fly. I saw also this really nice guide on how to make planes. Link to forum thread:

 

[kiwinanday]

@TheRagingIrishman CoM should always* be in front of the CoL. Think of a paper airplane—you fold the paper over in the nose so that the CoM is "pulling" the CoL forward, generating lift. Generally, the closer they are together, the more maneuverable an aircraft will be. This is because the mass of the airplane wants to pivot around the CoM, but the lifting forces pivot around the CoL. The closer they are together, the closer they are to pivoting around the same point. The aircraft in the OP is very agile, because of how close the Cs are to each other.

If that makes sense.

*Inherently unstable aircraft, like the X-29, have the CoL in front of the CoM for maneuverability, but they require constant computer corrections to keep the airplane from tumbling. If you use Atmosphere Autopilot or similar, you can build inherently unstable aircraft, but the fly-by-wire will always have to be on for it to be flyable.

6 minutes ago, qzgy said:

Like, for most planes? CoL behind CoM, or on it, depending on want of maneuverability. Farther back, harder it is to fly. I saw also this really nice guide on how to make planes. Link to forum thread:

 

Thx. Figured there were others, but thought I'd lend a hand.

Edited March 26, 2017 by kiwinanday

[StahnAileron]

4 hours ago, TheRagingIrishman said:

@kiwinanday in general, what should the com/col placements look like?

@qzgy Covers it and the guide linked is like 80% relevant. (You can ignore the parts taking about NEAR/FAR vs Stock Aero since those are from the Beta and Alpha days of KSP.)

In my experience recently, if you want high stability, CoL should be a fair distance behind CoM and maybe even a bit above it as well. You won't have much pitch control, though. (Which is pretty important, unless you can place your pitch control surface even farther for leverage... Won't help much if your primary CoL is really far behind the CoM.)

Agility would have you placing the CoM/CoL closer together (even over lapping) and perhaps with the CoL below the CoM. However, unless you have enough lift (high Lift to Mass ratio) to deflect your flight path about as quickly as you can turn (that is, pitch up/down), you'll just flip your craft if you do any hard maneuvers. (And be wary of body lift if your CoM happens to be really far to the rear of the aircraft itself. This can and will wreck your stability past a certain AoA, dependent on the design. My Shuttle rip-offs using the MK3 parts as the basis have this problem.)

I did design a fairly light weight craft with lots of wing area for absurd lift for its size and weight.

At this point, if I'm making a relative traditional style plane and am somewhat methodical about it, I build the main body first, place the main wings (and any control surfaces it warrants) and adjust the CoL to be near the CoM, then place my stabilizers last. This helps ensure my primary lift component from the main wings actually works to LIFT the plane without acting too much like stabilizers as well. As such, I can actually turn well instead of being a glorified lawn dart.

@kiwinanday Looking at your designs, have you considered using the FuelTanksPlus mod? Its assortment of fuel tanks have longer fuel tanks with fuelswitching. I use it since the parts look nicer and I can get away with less parts. (It also has a bunch of smaller/shorter tanks.) This is mainly in reference to the booms on the drone.

Edited March 26, 2017 by StahnAileron

[AHeroReborn]

can i just say i like the idea of this thread!

[qzgy]

7 minutes ago, StahnAileron said:

 

At this point, if I'm making a relative traditional style plane and am somewhat methodical about it, I build the main body first, place the main wings (and any control surfaces it warrants) and adjust the CoL to be near the CoM, then place my stabilizers last. This helps ensure my primary lift component from the main wings actually works to LIFT the plane without acting too much like stabilizers as well. As such, I can actually turn well instead of being a glorified lawn dart.

For me, the best way is to make all the aerodynamic surfaces on the body, and adjust with the offset tool or add more lift surfaces if needed so that the CoL is near CoM

[StahnAileron]

1 minute ago, qzgy said:

For me, the best way is to make all the aerodynamic surfaces on the body, and adjust with the offset tool or add more lift surfaces if needed so that the CoL is near CoM

I used to do that, but for some designs, a single CoL indicator is actually a curse. You THINK your plane is agile because the indicator (which is an average of all lifting surfaces) is near the CoL. In reality, it isn't. I made a few planes that had this problem. The solution didn't dawn on me until recently. (I've been out of KSP until recently, really.) I'm trying to get into a habit of seeing were my primary lift is sitting relative to the CoM first, THEN placing my stabilizers and pitch/yaw control surfaces. I find my designs behave more like I intended them to with this process. (It also helps to place your primary control surfaces a fair distance from the CoM for leverage. Give you more authority.)

I also try to add angle of incidence to my main wing so it's ALWAYS generating some amount lift at speed, even at zero AoA. Stabilizers should be parallel to the main body. Ideally, they shouldn't be generating any real lift, i.e. supporting the mass of the craft in the air. That's the main wing's job. Of course, this is at ideal cruising speed where, for the Angle of incidence of the main wing, you can fly straight and level with zero AoA. Still, the brunt of the heavy, uh, lifting should be done by your main wings (i.e. wing load.)

Well, there's body lift as well, but KSP's application of it is simplified. (I know of no body parts, stock of modded, that generates lift at zero AoA. I imagine in real-life lifting body designs, some lift would be generated at zero AoA for practical designs.) As such, body lift would ideally be zero in KSP. Meaning your main body is parallel with your flight path, a.k.a. Zero AoA, and therefore producing the lowest drag possible for the design. (Thereby increasing efficiency of the design.)

KSP's simplification is great for getting started, but you'll curse it if you start to understand what the hell you're doing because it hides the little details. (I mean, you need a mod to see DCoM vs. CoM unless you like draining and filling tanks when building to check long term stability.)

Now that I think about it, it'd be kinda nice if RCSBuildAid could show independent indicators for Roll, Yaw, and Pitch, and Static lift. (Though really, I only care about Pitch vs. Static...) Hmm... Perhaps a visit to the RCSBA thread is warranted...

[kiwinanday]

I'm much more intuitive with my airplane design. I start with a visual picture of the entire aircraft and what it's meant for (how maneuverable it will be), and my mind can somehow just see how the air will flow over the surfaces, and how the plane will react, so I just intuit everything and go from there. The main aircraft in the OP required a bit of vertical engine adjustment to balance the lift of the angled wing, but it was otherwise ready to go on first build. And with FAR, I'm even better, because the air behaves more like an actual fluid. The only one of the shown aircraft that was real trouble was the delta-winged SSTO spaceplane (last pic), because it weighed 25t fully loaded. And the one above is shamelessly just a EF Typhoon knockoff. I definitely prefer canard aircraft to conventional, though they're a whole different animal, and I have a small handful of them, but I'm keeping them in my pocket for now ;-)

Edited March 26, 2017 by kiwinanday

[Azimech]

9 hours ago, kiwinanday said:

@TheRagingIrishman CoM should always* be in front of the CoL.

 

Unless you have gigantic pitch/rudder control authority but usually you're right.

https://kerbalx.com/Azimech/77I--TCCS-Pallas-12

Edited March 26, 2017 by Azimech

[kiwinanday]

If your CoL is ahead of your CoM, gliding isn't really possible without fly-by-wire.

Edited March 27, 2017 by kiwinanday

[Matuchkin]

On 25/03/2017 at 11:20 PM, kiwinanday said:

CoM should always* be in front of the CoL.

Wouldn't it be better to put the CoL in front of the CoM, on aircraft such as bush planes, light and small utility planes (such as the Cessna 172), and other such civilian aircraft?

[kiwinanday]

No. Placement of CoL and CoM has no effect on the Lift:Mass ratio (which I think is what's in your mind).

 Look at the 172:

The big, heavy engine is up front, ahead of the wing, and the big, heavy passengers sit below the wing, so the net mass is ahead of the wing. As mentioned above, an aircraft without fly-by-wire can't glide unless the center of mass is ahead of the center of lift. To prove this, fold a basic dart paper airplane, and try to throw it backwards.

Edited March 27, 2017 by kiwinanday

[Hodo]

The whole debate over COL vs COM discussion really depends on several factors, most of which dont apply here to KSP.  

When you fly with FAR installed it does make a HUGE difference over the stock "aerodynamics".

And a supersonic aircraft is going to behave differently than a subsonic aircraft designed for just that.

I know some of my most successful designs were designed strictly for one thing and did that ONE thing VERY well.  

 

Like this old design..

It was a switchblade wing design... worked extremely well for speed runs to orbit.  

And I have designs that are "combat" designs that were VERY good at their mission.

So again... if you design it for one job it will be VERY good at that one job.

Dont worry about the CoL vs CoM, worry if it can at least do the job it is meant to do.  

[Azimech]

Indeed, I have a few gigantic cargo planes which have to operate with a wide range of CoM shift. I just give them a very long tail and a lot of control surfaces. With FAR I would change the design somewhat. With my helicopters it's a different story, I always put CoM in front of CoL because stock rotors can't compensate forward flight.

[kiwinanday]

@Azimech wouldn't it be better to design the cargo bay to be centered under the CoM, so that your payloads don't have an effect on flight characteristics?

Edited March 27, 2017 by kiwinanday

[Azimech]

1 hour ago, kiwinanday said:

@Azimech wouldn't it be better to design the cargo bay to be centered under the CoM, so that your payloads don't have an effect on flight characteristics?

It really depends on the payload ;-)

[Hodo]

1 hour ago, kiwinanday said:

@Azimech wouldn't it be better to design the cargo bay to be centered under the CoM, so that your payloads don't have an effect on flight characteristics?

 

When dealing with large cargo aircraft, you have to account for the cargo's CoM being different than the CoM of the aircraft.  

 

Like this.. 

It hauled two of these into orbit.

The CoM was pretty easy on those.. they were centered on the cargo.  And had very little effect on the crafts CoM.

BUT... some of my earlier cargo monsters.

That one was tricky.. The cargo CM was WAY far forward of the CM of the craft.  While lighter than anything that craft normally hauled into orbit, it was not the way I liked it.  Still ended up using way more fuel to use brute force to over come the shift in mass.

[kiwinanday]

True that (nice lifters, btw), but it helps if you can design a payload to be as balanced as possible.

[Rath]

On 3/26/2017 at 4:30 PM, StahnAileron said:

I used to do that, but for some designs, a single CoL indicator is actually a curse. You THINK your plane is agile because the indicator (which is an average of all lifting surfaces) is near the CoL. In reality, it isn't. I made a few planes that had this problem. The solution didn't dawn on me until recently. (I've been out of KSP until recently, really.) I'm trying to get into a habit of seeing were my primary lift is sitting relative to the CoM first, THEN placing my stabilizers and pitch/yaw control surfaces. I find my designs behave more like I intended them to with this process. (It also helps to place your primary control surfaces a fair distance from the CoM for leverage. Give you more authority.)

I also try to add angle of incidence to my main wing so it's ALWAYS generating some amount lift at speed, even at zero AoA. Stabilizers should be parallel to the main body. Ideally, they shouldn't be generating any real lift, i.e. supporting the mass of the craft in the air. That's the main wing's job. Of course, this is at ideal cruising speed where, for the Angle of incidence of the main wing, you can fly straight and level with zero AoA. Still, the brunt of the heavy, uh, lifting should be done by your main wings (i.e. wing load.)

Well, there's body lift as well, but KSP's application of it is simplified. (I know of no body parts, stock of modded, that generates lift at zero AoA. I imagine in real-life lifting body designs, some lift would be generated at zero AoA for practical designs.) As such, body lift would ideally be zero in KSP. Meaning your main body is parallel with your flight path, a.k.a. Zero AoA, and therefore producing the lowest drag possible for the design. (Thereby increasing efficiency of the design.)

KSP's simplification is great for getting started, but you'll curse it if you start to understand what the hell you're doing because it hides the little details. (I mean, you need a mod to see DCoM vs. CoM unless you like draining and filling tanks when building to check long term stability.)

Now that I think about it, it'd be kinda nice if RCSBuildAid could show independent indicators for Roll, Yaw, and Pitch, and Static lift. (Though really, I only care about Pitch vs. Static...) Hmm... Perhaps a visit to the RCSBA thread is warranted...

The best way to either remove the lift of the tail or wings while keeping the weight showing is by tilting them vertical with the rotate widget.

[StahnAileron]

44 minutes ago, Rath said:

The best way to either remove the lift of the tail or wings while keeping the weight showing is by tilting them vertical with the rotate widget.

Good point, but I don't mess with rotation unless I have to. I dislike mucking with placement and rotation just to test something in the middle of building (maybe during fine-tuning near the end after the overall building is complete.) I've screwed up countless time setting something in the editor to test conditions or design around it only to forget to reset to normal and screwing up a launch/test flight and wasting time reverting. (In particular, engine thrust levels for VTOL designs that have a mix of vertical and horizontal engines. I later found out that re-ordering engine staging can simplify that, but then I may forget to set-up a proper staging sequence, though most of the time it's of little consequence.)

Something like EEX Redux helps, but it's still something I try to avoid if I can help it.

Weight is less of a concern for me because in most cases, the stabilizers are a relatively miniscule fraction of the craft mass. Tailoring the main wings' CoL to the tailless CoM isn't a problem unless I'm trying to set them very close together or I'm trying to make a very lightweight design with good/even mass distribution.

On ‎3‎/‎26‎/‎2017 at 9:04 AM, Azimech said:

Unless you have gigantic pitch/rudder control authority but usually you're right.

https://kerbalx.com/Azimech/77I--TCCS-Pallas-12

That is fine example of kerbal methodology in practice.

On ‎3‎/‎27‎/‎2017 at 1:33 PM, Hodo said:

The whole debate over COL vs COM discussion really depends on several factors, most of which dont apply here to KSP.  

[...]

Dont worry about the CoL vs CoM, worry if it can at least do the job it is meant to do.  

The main point is stability/controllability versus agility/responsiveness. That's really what the discussion is about. (Well, and general design considerations for those that haven't delved deep into KSP's aero system.)

This can affect how well the design can accomplish what you intend it to do, so you can't just say, "don't worry about it." It's an inherent factor and aspect of craft design. If you screw up the CoL/CoM balance and you keep flipping out on take-off (or can't even take-off on the other end of the spectrum), you won't be accomplishing much. (Unless you're just screwing around. But nothing really matters if you're just screwing around, anyway. ) CoL/CoM balance may not be as important as, say, the CoT/CoM alignment of a rocket, but it still matters enough to take into consideration.

[Hodo]

8 hours ago, StahnAileron said:

 

This can affect how well the design can accomplish what you intend it to do, so you can't just say, "don't worry about it." It's an inherent factor and aspect of craft design. If you screw up the CoL/CoM balance and you keep flipping out on take-off (or can't even take-off on the other end of the spectrum), you won't be accomplishing much. (Unless you're just screwing around. But nothing really matters if you're just screwing around, anyway. ) CoL/CoM balance may not be as important as, say, the CoT/CoM alignment of a rocket, but it still matters enough to take into consideration.

I think you misunderstood me, but I wasnt that clear.  I dont mean to say not to worry about it, but it shouldnt be the primary thing you dwell on.  If you lose the fun trying to make the perfect aircraft, balanced and everything you will find it being more of a job than a game.  Sometimes you can find some interesting designs that actually work quite well but on paper shouldnt work.  

 

[StahnAileron]

24 minutes ago, Hodo said:

I think you misunderstood me, but I wasnt that clear.  I dont mean to say not to worry about it, but it shouldnt be the primary thing you dwell on.  If you lose the fun trying to make the perfect aircraft, balanced and everything you will find it being more of a job than a game.  Sometimes you can find some interesting designs that actually work quite well but on paper shouldnt work.  

 

Ah, okay. Gotcha. Yeah, I sometimes do that. It how this thing got made:

I balanced the CoM/CoL after the fact. This was like the third or fourth version of this craft. The first few relied on RTGs, fixed solar panels, and batteries.

[kiwinanday]

@StahnAileron gorgeous bird! I'm a bit jelly that I didn't have the idea for a ringwing.