How to keep heavy SSTO's under Control?

[Kobymaru]

Hi!

So I built a "heavy" SSTO of about 130 tons. It flies OK, and it can theoretically go to orbit (I did it once). The problem with it is however, that it just LOVES to flip out of control.

It lower atmosphere, going over an AoA of 10° (no matter if up or down) makes it flip out, and I can't get it back under control.

In higher atmosphere, when accelerating to orbital speeds, even 5° AoA are enough. That's especially problematic since this is where you have to initiate a climb to orbital altitude. The plane gets out of control and undergoes rapid unplanned disassembly due to high aerodynamic stress.

And I have to say, I don't understand what is going on here!! My CoL is way behind my CoM, and I should have enough control authority.

Here are some pictures:

Javascript is disabled. View full album

So how do you guys control your heavy SSTO's? Should I just be "careful"? Or use mods? Or Rebuild my plane?

ps.: I hate the Stock SAS. It was always pretty bad, but ever since the pilot capabilities got updated, it is just absolutely useless for anything slightly bigger than a 5-part ship. Oscillations everywhere, and it rolls when it shouldn't and doesn't roll when it should. Can anyone recommend a good Fly-By-Wire Mod?

Edited September 15, 2015 by Kobymaru

[UnusualAttitude]

Hey there,

Nice looking ship. You're right, stock SAS is usually very unforgiving, especially with larger craft. The Pilot Assistant mod has a range of functions to help atmospheric flight (http://forum.kerbalspaceprogram.com/threads/100073-1-0-x-Pilot-Assistant-Atmospheric-piloting-aids-1-11-7-(Aug-20), just tuning the SAS PID parameters should help.

Not sure about the flipping though, as I can see nothing immediately wrong with your design. Try the following however:

- Lock off the forward fuselage tank until you need it to make sure your CoM is not drifting behind the CoL.

- Check the lateral tanks during flight for unbalanced fuel consumption (bicouplers sometimes cause this). If so, you might be able to fix it with fuel ducts.

Good luck, let us know if this helps.

UA.

[jiminator]

is your center of thrust going through the center of mass or above it? That angled tail piece may be your problem

[p1t1o]

How fast/high are you when you generally lose it? On ascent, fly to maximise your speed, rather than height. If you try to fly as fast and as high as possible,you can easily go too high for your current speed - this causes lift to drop away, leading to loss of control authority and the plane usually flips out.

Once you feel like you are approaching the upper portion of your speed range, slow your ascent and pick up horizontal speed. Once you reach around 18-20km your ascent should be quite shallow, as you start to approach orbital speeds you will find that you pitch up naturally due to passing around the curvature of Kerbin anyway.

Also, I don't know if this is just me or what, but I always turn off reaction wheels on (SAS on, reaction wheels off) my aircraft, they always seem to battle the aerodynamic controls and cause excessive "nodding" and contributing to loss of control events.

[Champ]

I hate the Stock SAS. It was always pretty bad, but ever since the pilot capabilities got updated, it is just absolutely useless for anything slightly bigger than a 5-part ship. Oscillations everywhere, and it rolls when it shouldn't and doesn't roll when it should. Can anyone recommend a good Fly-By-Wire Mod?

Did you assign roles to your control surfaces ? This helps

[Brainlord Mesomorph]

Looking at the pics:

you have 4 RAPIERs pretty much in line with CoL and CoM, but those turbojets are well above CoM and CoL. That would push the nose down.

When you try to fly straight and level, is SAS neutral (centered)? or is it pulling up trying to compensate for a CoT that is too high? (if so, as soon as you touch the controls, you override that, at now you're out of control)

[Captain H@dock]

I'm not on top of how lift changes with AoA on different wing parts, but could the delta wing lift increase faster than the airliner ones? IF this is the case (someone would have to confirm), your craft CoL would tend to move forward when the absolute value of the AoA increases.

The mk2 fuselages on the side might also contribute in such a way. F12 might be able to confirm/infirm this theory.

Edited September 8, 2015 by Captain H@dock

[Merandix]

I tried to recreate it... it stands out that at some point you moved the tailplane forward or backwards, and you also angled it. With all wings at 0 degrees angle of incidence, this plane has it's centre of lift in front of it's centre of mass, and it flips out on take-off. If I tweak the tailplane a bit like you do and move it backwards as well, I also have it flipping out at altitude. I can tweak further, and it will likely just move the point where it will flip out. I don't really know what is happening here, but I do suspect the tail is the cause of your problems, especially in combination with the two jets being in a raised position (forcing the nose downwards, even if it wants to go upwards, hiding your problem).

You need someone more experienced with wings under an angle of incidence to fully explain what the heck is going on here, but I highly suspect that is the cause. You may be able to tweak this until it remains relatively stable. But I doubt that will help you as the thrust from your rearmost engines will decrease or cease. They also may flame out at altitude, as there's fairly little intake going on on your plane, 4 intakes for 6 engines....

[GoSlash27]

It looks like your rear airliner wing was cocked up at high incidence while the front ones weren't. That would cause some very odd behavior.

Best,

-Slashy

[Kobymaru]

Hi! A huge thank you to all for your input!

I have "solved" the Issue with the following steps:

- Moved the CoM directly into the CoL. Since the old rule "CoL behind CoM for stability" did clearly not hold enough. Therefor I gained some authority and the stability will come from software.

- Flattened out the elevators

- Flattened out all the wing business. Originally, I wanted to tilt the wings like I read somewhere in another thread but in my case it messed with the aerodynamics

- Maybe I will come back to the concept after further development

- Using the Pilot Assistant mod as suggested.

Controllability got a little better and I got it to orbit with Pilot Assistant. But I still wouldn't classify the craft as particularly safe and I'm a little unhappy with the controllability at the moment.

I'll keep working on the craft though, and see what I can come up with.

Looking at the pics:

you have 4 RAPIERs pretty much in line with CoL and CoM, but those turbojets are well above CoM and CoL. That would push the nose down.

The turbojets get turned off at higher altitude, I use only the Rapiers to accelerate.

I'm not on top of how lift changes with AoA on different wing parts, but could the delta wing lift increase faster than the airliner ones? IF this is the case (someone would have to confirm), your craft CoL would tend to move forward when the absolute value of the AoA increases.

The mk2 fuselages on the side might also contribute in such a way. F12 might be able to confirm/infirm this theory.

This is worth looking into. I'll check back later.

[chrisb2e9]

- Moved the CoM directly into the CoL. Since the old rule "CoL behind CoM for stability" did clearly not hold enough. Therefor I gained some authority and the stability will come from software.

I know that people will argue with me on this... but whatever.

CoM in the real world on real planes is the center of gravity. The center of gravity is a point about which the airplane rotates on all axis. More detail here if you like:

https://www.grc.nasa.gov/www/K-12/airplane/cg.html Or here:

https://en.wikipedia.org/wiki/Center_of_gravity_of_an_aircraft

CoL, or center of lift. Is not something that we generally think about when flying a plane. Because the wings are fixed, with the exception of elevators and ailerons.

Having said that, imagine you are sitting in a plane and you pull back on the controls as far as you can. You just moved the "CoL" forwards. In front of the CoG (center of gravity). So the nose of the aircraft is going to lift, and the tail will be pushed down.

So, picture where the CoL ends up. That's your forwards limit. If the center of gravity, is close to that limit. You will need constant back pressure on the controls to keep the nose up.

Hard to picture I know.

Basically what I am saying(and trying to convey 2.5 years of studying theory of flight), is that you should have the CoL at the CoG when you make your plane. Putting the CoL behind the CoG does not add stability as some people say. It just means that the plane is going to be nose heavy, and you will need constant pressure(also known as trim) to move the CoL to the CoG to keep the plane level when in straight and level flight. If you picture the CoL as a point in which all lift acts, it's behind the CoM so it's pushing up on the tail. That's not stable... The two points need to be at the same point so that the forces are in balance.

This doesn't help with your issue that you first posted about(I just hate hearing that CoL needs to be behind CoM for stability)... I need to know more about that aircraft and your game. Modded aerodynamics for example? How much do your wings flex? Or even the fuselage of the aircraft? Could be a lot of factors really...

I will say one thing, I would remove the three wings and get the B9 procedural wings. I personally find that they work a lot better.

[Vegetal]

I know that people will argue with me on this... but whatever.

Ah I will

First, do you have any example? I mean a craft file?

Second, when flying an actual plane, you do use trim to keep the nose level, and if you reduce speed, you have to pull in order to keep it level. Why? As the CoL is behind the CoG, when the main wings lose lift, it wants to pitch down, and the reduced speed makes you gradually lose control authority.

But another important factor we don't have in stock ksp is center of pressure, which is also very important for stability. I've had mk3 planes with cargo bays in front which were very flip happy, but only when a certain AoA was exceeded. I think FAR has tools to analyze that, but I never used it.

What area from aviation are you from? I'm curious, because I come from aviation too.

[ajburges]

I know that people will argue with me on this... but whatever.

A challenge!

Your description of torque behavior would work if the lift vector was constant in scale. However, KSP lift vector scale is proportional to AoA. Abstracting the CoL to be the CoP, lift will always work with a strength proportional to the AoA. If the CoL is behind CoG, that force will generate torque which will work to reduce the angle (and increase if positions are reversed). When AoA is negative, so is there scalar component.

The aspects of CoP != CoL is true. For the scales we use CoM == CoG. Rounding errors mean we can't construct scenarios where CoG and CoM are appreciably different.

[Kobymaru]

So, picture where the CoL ends up. That's your forwards limit. If the center of gravity, is close to that limit. You will need constant back pressure on the controls to keep the nose up.

Hard to picture I know.

Basically what I am saying(and trying to convey 2.5 years of studying theory of flight), is that you should have the CoL at the CoG when you make your plane. Putting the CoL behind the CoG does not add stability as some people say. It just means that the plane is going to be nose heavy, and you will need constant pressure(also known as trim) to move the CoL to the CoG to keep the plane level when in straight and level flight. If you picture the CoL as a point in which all lift acts, it's behind the CoM so it's pushing up on the tail. That's not stable... The two points need to be at the same point so that the forces are in balance.

No, it's perfectly clear actually. I know this constant back pressure, and that's exactly why I decided to move the CoM into the CoL. Especially since Trim doesn't really exist in KSP. I know you can do Alt-WASD, but unfortunately it only works for a neutral position, and does not apply when you apply control input, which makes it jumpy.

Or something. I don't know. Trim in KSP is not how trim should be.

This doesn't help with your issue that you first posted about(I just hate hearing that CoL needs to be behind CoM for stability)... I need to know more about that aircraft and your game.

I have a new version which flies a little better. Also, I can upload the craft file if you want.

Modded aerodynamics for example?

Stock aero. Although it has been a while since I repaired my install (and the Physics.cfg). Dunno.

How much do your wings flex?

Visually, not at all. Especially compared to other planes.

Or even the fuselage of the aircraft?

It does on the initial physics-jerk, but not during flight.

I will say one thing, I would remove the three wings and get the B9 procedural wings. I personally find that they work a lot better.

Well, I would be inclined to do this, but I kinda wanted to keep my install as close to stock as possible, at least parts-wise.

By the way, some other person had a similar problem and he seems to have nailed it down to the Mk-3 Cargo bay. Could this be the issue?

I recently noticed that the flipping becomes more of a problem as soon as I go supersonic.