How to stop jet spinouts at very high speeds?

[Kingtj44]

This is a unique question that requires some background.

I am currently developing a hypersonic jet named BlackStar. This is not my first attempt, and I have gotten my other hypersonic crafts up to 2,315 m/s, but I want to go even faster... no mods no exploits

My BlackStar spins out very easily compared to my earlier hypersonic jets, likely due to the huge amount of new intakes. I am guessing that the large amount of drag makes the aircraft unstable.

In fact, the rear of the aircraft tend to wants to point prograde as I am flying, which I have also noticed in other planes. It basically means my aircraft has an urge to flip around the entire flight.

Is there a way to stop this? A way to make spinouts much harder to do? At these speeds control is very difficult and any tips on how to stabilize the aircraft are welcome!

Edited November 25, 2013 by Kingtj44

[Vanamonde]

It's hard to say without seeing your vessel, but is it possible that as you burn fuel, the center of mass is shifting towards the back (where the weight of the engines generally can be found)?

[Kingtj44]

I'll put SPH pics up with center of mass etc. soon. I'll try using more stable wing positions to decrease spinouts, I'll see if it works.

[Bilfr3d]

You'll also notice that if you turn on the weight and lift indicators in the SPH, if the life is IN FRONT of the weight, that will slowly send it upwards (depending on how far forward it is), and if your weight is all the way at the back, that will send the back down. Usually you want to aim for the lift to be either just behind the weight indicator or right on it. That gives the best amount of manueverability.

[Kingtj44]

You'll also notice that if you turn on the weight and lift indicators in the SPH, if the life is IN FRONT of the weight, that will slowly send it upwards (depending on how far forward it is), and if your weight is all the way at the back, that will send the back down. Usually you want to aim for the lift to be either just behind the weight indicator or right on it. That gives the best amount of manueverability.

The issue doesn't seem to be that it slowly creeps one direction or the other, its that even small movements from the prograde position will send the craft into a severe spinout in any speed over 1,000 m/s. My guess is that the drag makes the craft extremely sensitive to lift changes and will have a seizure at even slight off-balances.

Edited November 25, 2013 by Kingtj44

[aksentti]

Intake drag. I think intakes in front of center of mass will do that.

Of course moving them back would require lots of other changes to the craft, which is a shame, as it looks quite nice as it is.

[Cruzan]

Hey there, based on the images you posted it looks like your CoL is way too close to your CoM. Try and shift the CoL back a hair so that the majority of the CoL bubble is at least sticking out of the CoM bubble. The way you have it set up at the moment your plane will be more than happy to fly forwards or backwards which is what causes it's flippiness.

As for the spinout issues it could be

• Flameout which is caused by lack of air at certain altitudes. One jet shuts down causing the others to send you in to a spin. (Seems obvious maybe, but needs to be mentioned )
  
• Asymmetry. I can't tell from your pics, but everything seems to be symmetrical so this probably isn't it
  
• Imbalanced fuel drain. Again I can't see fromthe pics, but If there are fuel tanks on the sides of your plane that aren't draining at the same rate, then your CoM will shift to that side and become misaligned with your thrust. This will cause a more gradual tendency to spin.
  

That's all I can think of without flying the plane for myself Good luck!

Cool design though btw! I like the way you did you wings.

EDIT: Also if it's just a very minor pull to one side or the other, you can try slapping on more SAS modules (reaction wheels) as they will help counteract this a lot!

[Kingtj44]

I raised the wings to increase stability. I also removed the frontal winglets (you can see them in the pics in front of the intakes) so the COL is farther back from the COM. It works much nicer now, but still at speeds 1500 m/s+ it seems to be sensitive to directional change. To help with high-speed control I added another rudder, which also seemed to help a little. Moving the intake around would require huge aircraft changes. I'm not sure how to get most intakes behind the center of mass because the engines are by far the heaviest thing, naturally putting the center of mass pretty far back.

I know the engines are not flaming out, it seems to be purely structural. All of the fuel tanks are central, no side fuel tanks, so that should not be too much of an issue.

Edited November 25, 2013 by Kingtj44

[Cruzan]

I would still move that CoL much further back. As your fuel drains, unless you've wired it up to drain fuel back to front, your CoM will move backwards while your CoL stays in place. It might be tough to add more lift towards the back to keep your design aesthetic, but maybe some swept wings angled down from the top of edge of the body that meet up with the main wings would look okay while also adding more lift towards the back? Idk, just spitballing

If you want to toss out a .craft file we could probably help you even more

[Kingtj44]

idk how to put a .craft file out tell me and I'll give you the plane. Its a beast though, around 350 parts so...use caution. The COL bubble is just barely touching the COM bubble, unless its best to put it farther out?

The fuel drains very oddly. It drains from the two small ends into the center tanks, which is better than nothing I suppose. (You can see the 4 central tanks that make up the body of the craft in the pics)

[Cruzan]

Here's a wiki article about Sharing .craft files

Can be a little difficult to find the folder the first time, but after the first time it's a breeze. I've got a shortcut leading to my saves folder since I do so much craft sharing

As for the bubbles, with how long your plane is, I'd say if your blue CoL bubble isn't at least completely outside of the yellow one then you'll have issues later on. There isn't anything wrong with having it a little further back than that either. Looking at the side view of your plane, do you see that 2nd vertical line from the back on the body (the wingconnector thing you're using)? I'd center the CoL in-line with that for starters

[Kingtj44]

Here's a wiki article about Sharing .craft files

Can be a little difficult to find the folder the first time, but after the first time it's a breeze. I've got a shortcut leading to my saves folder since I do so much craft sharing

As for the bubbles, with how long your plane is, I'd say if your blue CoL bubble isn't at least completely outside of the yellow one then you'll have issues later on. There isn't anything wrong with having it a little further back than that either. Looking at the side view of your plane, do you see that 2nd vertical line from the back on the body (the wingconnector thing you're using)? I'd center the CoL in-line with that for starters

I think the craft file is here:

https://drive.google.com/file/d/0B3xj2-JHf0wEMVdIcnNxZ2lnb1E/edit?usp=sharing

If something isn't working right with the download tell me

Have fun, and also I moved the entire wing section back, you may notice.

[1greywind]

SPH\VAB CoM indicator gives you position with fully fueled craft. In flight your engines sucks fuel from forward fuel tanks first and your CoM shifts back - past your CoL and your craft becomes statically unstable. To solve this issue you need to eighter rearrange all your fuel tanks in such way that CoM will not shift, or you need to shift your CoL to back so it will be past your CoM even with empty fuel tanks.

[Cruzan]

I think the craft file is here:

https://drive.google.com/file/d/0B3xj2-JHf0wEMVdIcnNxZ2lnb1E/edit?usp=sharing

If something isn't working right with the download tell me

Have fun, and also I moved the entire wing section back, you may notice.

Nice!

Did you try flying it with this setup yet?

I just took it for a ride and didn't notice any issues. At low speeds and low altitudes it does weird things because the wings flex/wobble, but at a steady heading, going 2100m/s+ it does just fine as far as I can tell. I flamed out a couple of times and those were the only times it seemed to want to yaw/spin.

[Kingtj44]

Nice!

Did you try flying it with this setup yet?

I just took it for a ride and didn't notice any issues. At low speeds and low altitudes it does weird things because the wings flex/wobble, but at a steady heading, going 2100m/s+ it does just fine as far as I can tell. I flamed out a couple of times and those were the only times it seemed to want to yaw/spin.

Yes and I just hit 2,323 m/s at 43,000m I believe its max speed is around 2,324 m/s. Fastest jet I have made yet, and yes it understandably sucks at low altitudes. Ceiling seems to be around 43,000m as well in steady level flight at full throttle. At these intense speeds, it just needed much more specialized design that focused on stability much more than my other craft. With some tweaking, true max speed is likely 2,325 m/s

It actually seems to be one of the fastest jet-only aircraft I have ever seen. There have only been a few that have been faster, but they were pretty much blobs of intakes with a jet engine, whereas mine could be crowned the fastest "real" jet aircraft to fly.

It might even be an SSTO, since it is extremely easy to "slip" right into orbit at those high speeds.

Edited November 25, 2013 by Kingtj44

[Cruzan]

It'll probably go faster if you raise its ceiling and push it higher by throttling back You'd definitely need a little boost from rockets/RCS to get orbital since, as far as I know, you can raise your periapsis above 70km without an extra boost. The Blackstar could probably make it with a single toroidal tank and a 48-7S engine!

[Kingtj44]

It'll probably go faster if you raise its ceiling and push it higher by throttling back You'd definitely need a little boost from rockets/RCS to get orbital since, as far as I know, you can raise your periapsis above 70km without an extra boost. The Blackstar could probably make it with a single toroidal tank and a 48-7S engine!

I actually slipped into 200,000m orbit on accident without rockets, I skipped through the atmosphere again though so it was an incomplete orbit. A small rocket could boost it easily into orbit.

I also have some more stats and info for it:

Glide landing possible: Yes, very easily

Approximate fuel time: 22 minutes of flight

Maneuverability: Poor

Top speed: 2,325 m/s

Stability: Moderate/good

Structural integrity: Moderate

Max full-throttle altitude: 43,500m - 44,000m

Cruising altitude: 43,000m

Responsiveness: Poor/moderate

Edited November 25, 2013 by Kingtj44

[numerobis]

First to the OP: note that the landing gear mess up the center of mass calculation. In the VAB/SPH they are accounted as having 0.5t mass. In flight, they have no mass at all. I tend to take off the gear, fix the CoL / CoM / CoP issues (it sure would be nice to have a CoP marker... I mostly just put high-drag items, like intakes, at or behind CoM) and then put the gear back on.

It'll probably go faster if you raise its ceiling and push it higher by throttling back

You reach top speed when you are at the highest altitude you can still be at full throttle. Higher, and you have to throttle back exactly as fast as drag falls, so there's no advantage. There's a minor disadvantage, which is that Isp falls, but that's more relevant when you have relatively few intakes per engine: going up 5km from 21km altitude (1 intake) will change your Isp from about 1265 to 1225, whereas going up 5km from 43km will only change your Isp from 1200.8 to 1200.3

To reach your highest horizontal speed you want to be going in a circular orbit, with no vertical component to your speed and no thrust dedicated to keeping you at your altitude. By coincidence, a polar orbit gets you just that: orbit speed at that altitude is very close to the speed you'll be going on jets, and a polar orbit means orbit speed is roughly equal to surface speed. Lift off and go due North, then just keep following the prograde marker. Near the poles, your heading changes; just follow the prograde marker (you don't need to do anything if you have intakes behind the center of mass, you'll just fly straight).

[Cruzan]

You reach top speed when you are at the highest altitude you can still be at full throttle. Higher, and you have to throttle back exactly as fast as drag falls, so there's no advantage. There's a minor disadvantage, which is that Isp falls, but that's more relevant when you have relatively few intakes per engine: going up 5km from 21km altitude (1 intake) will change your Isp from about 1265 to 1225, whereas going up 5km from 43km will only change your Isp from 1200.8 to 1200.3

To reach your highest horizontal speed you want to be going in a circular orbit, with no vertical component to your speed and no thrust dedicated to keeping you at your altitude. By coincidence, a polar orbit gets you just that: orbit speed at that altitude is very close to the speed you'll be going on jets, and a polar orbit means orbit speed is roughly equal to surface speed. Lift off and go due North, then just keep following the prograde marker. Near the poles, your heading changes; just follow the prograde marker (you don't need to do anything if you have intakes behind the center of mass, you'll just fly straight).

I'm not sure about this one. I don't know what KSP's drag model is actually like, but from practice, my planes go their fastest when I've parked them up as high as I possibly can without burning out. This usually involves throttling back down to the last or second to last tick. Often increasing from something like 1800m/s to 2150+m/s (surface v, I believe my fastest is 2380m/s orbital v) with the latter velocity being when I'm under 1/3rd throttle. I'm sure it varies with each plane's design, but I wouldn't say that throttling back and loss of drag from gaining altitude are directly offsetting. Often the extra altitude offsets the slight loss of thrust. I've heard that KSP's drag model is very harsh which might be accentuating this. Maybe using mods like FAR things would be more like you're describing?

This has just been my experience with designing quite a few SSTO's in KSP. You can test it out for yourself and find the butter zone for your own designs, but I would venture a guess to say that you'll go faster if you park at a higher altitude and throttle back accordingly

Edited November 25, 2013 by Cruzan

[Zephram Kerman]

Taking a look at those images, I'm inclined to blame directional instability on those vertically mounted structural wings on the sides of the fuselage, especially the ones up front. While they look very cool, they also create lateral "lift" as soon as they get any angle of attack. As that slip angle increases, they get more and more lift, so the problem is self-reinforcing and inherently unstable.

In the SPH, turn the entire plane on one side, so a wingtip is at the floor. I suspect the CoL will suddenly jump way forward of the CoM.

If you want to keep the look, replace those structural wings with the square structural plates instead. They have no lift rating, so they would be totally passive. Alternatively, add one tail fin in back for each wing up front.

Edited November 25, 2013 by Zephram Kerman

[Kingtj44]

I am not sure if I want to make an even faster aircraft. The absolute maximum speed of A jet aircraft is 2,400 m/s because turbojets simply stop creating any thrust at that speed. Obviously there is no such thing as the perfect aircraft so 2,400 m/s is unreachable. Not sure, even major redesigns would only result is a few more m/s

[numerobis]

...

Likely the reason you're speeding up while climbing is that you didn't reach your maximum speed at low altitude. You could climb more slowly. Another reason is that as time goes on, you burn fuel, which reduces your drag.

For the relevant math:

drag = 0.5 Cd rho A v^2, where Cd is 0.008, rho is the air density, A is the sum for each part of its mass and its drag, and v is the surface velocity.

airflow = area * rho * cos(AoA) * 40 * (600 + v), where area is the value in the part.cfg for an intake, and AoA is the angle of attack of the intake.

air required = throttle * 225 kN / (Isp . g0) where g0 is 9.82 m/s^2 for KSP, and Isp is between 1200s and 1201s at the relevant altitudes. If you have enough airflow, you can fly at full throttle; otherwise you have to reduce throttle.

thrust depends on throttle and speed, not directly on the air density.

Drag and airflow are both linearly related to air density, so if you reduce air density by flying higher, you reduce drag and airflow by the same relative amount. Below your ceiling, extra airflow does you no good, so you should go higher. Above your ceiling, reducing airflow reduces your thrust, so it all cancels out.

There are some lesser considerations:

Isp falls with altitude, which means you need more airflow to maintain the same throttle, so with one third the air density, you need to throttle back to less than a third, so you shouldn't fly higher than necessary -- but the difference at very high altitude is, as I mentioned earlier, less than a part in a thousand.

The amount of air in your intakes -- which counts as mass for calculating drag -- also falls with altitude. However, there's almost nothing to save here: at 43 km, you have 18 kg/s going through your intakes, so if you're flying at 10 FPS (physics delta-t of 0.1, the maximum allowed in the default settings) you have 1.8kg of air on your plane every frame; compare that with about 800 kg of intakes. Plus, there's a trick: the amount of air depends on your physics delta-t, so you can set that very low in the settings.cfg ; max physics delta-t of 0.01 means you only have 180g of air on your plane.

[Cruzan]

Likely the reason you're speeding up while climbing is that you didn't reach your maximum speed at low altitude. You could climb more slowly. Another reason is that as time goes on, you burn fuel, which reduces your drag.

For the relevant math:

drag = 0.5 Cd rho A v^2, where Cd is 0.008, rho is the air density, A is the sum for each part of its mass and its drag, and v is the surface velocity.

airflow = area * rho * cos(AoA) * 40 * (600 + v), where area is the value in the part.cfg for an intake, and AoA is the angle of attack of the intake.

air required = throttle * 225 kN / (Isp . g0) where g0 is 9.82 m/s^2 for KSP, and Isp is between 1200s and 1201s at the relevant altitudes. If you have enough airflow, you can fly at full throttle; otherwise you have to reduce throttle.

thrust depends on throttle and speed, not directly on the air density.

Drag and airflow are both linearly related to air density, so if you reduce air density by flying higher, you reduce drag and airflow by the same relative amount. Below your ceiling, extra airflow does you no good, so you should go higher. Above your ceiling, reducing airflow reduces your thrust, so it all cancels out.

There are some lesser considerations:

Isp falls with altitude, which means you need more airflow to maintain the same throttle, so with one third the air density, you need to throttle back to less than a third, so you shouldn't fly higher than necessary -- but the difference at very high altitude is, as I mentioned earlier, less than a part in a thousand.

The amount of air in your intakes -- which counts as mass for calculating drag -- also falls with altitude. However, there's almost nothing to save here: at 43 km, you have 18 kg/s going through your intakes, so if you're flying at 10 FPS (physics delta-t of 0.1, the maximum allowed in the default settings) you have 1.8kg of air on your plane every frame; compare that with about 800 kg of intakes. Plus, there's a trick: the amount of air depends on your physics delta-t, so you can set that very low in the settings.cfg ; max physics delta-t of 0.01 means you only have 180g of air on your plane.

Wow that's a lot of good info! Is the throttle value in that equation a percentage value from 0.0-1.0?

I think I may have also been bringing an SSTO mentality to a hypersonic airplane design discussion, as I was neglecting the fact that my total velocity != my horizontal velocity. My bad

[Kingtj44]

After testing the BlackStar with me, is there any more edits I should do? Also how do I change this tread to "answered?"

[numerobis]

Wow that's a lot of good info! Is the throttle value in that equation a percentage value from 0.0-1.0?

I think I may have also been bringing an SSTO mentality to a hypersonic airplane design discussion, as I was neglecting the fact that my total velocity != my horizontal velocity. My bad

Indeed, throttle is between 0 and 1 (not a percentage).

Minor bug: I gave for the "air required" the formula for the total propellent required; air is 15/16ths of that, fuel is 1/16th.