What makes this plane spin out?

[GoSlash27]

Red Iron,

I'll skip the obvious stuff, as I know you've probably already covered that.

The one thing that stands out to me is the wing sweep caused by attaching them to a surface that's not aligned with the centerline. In my experience, the stock aerodynamics engine does care about that. If I understand the game physics, the lift vector is rotated outboard on each wing, causing a dynamic instability in yaw.

Hope that's it!

-Slashy

[Red Iron Crown]

The one thing that stands out to me is the wing sweep caused by attaching them to a surface that's not aligned with the centerline. In my experience, the stock aerodynamics engine does care about that. If I understand the game physics, the lift vector is rotated outboard on each wing, causing a dynamic instability in yaw.

That's new info to me, thanks. That could very well be the problem as both the wings and vertical stabilizers are attached in a swept fashion to match the taper of the tanks.

[Overfloater]

I'm not sure what you mean by this exactly, can you clarify?

I meant two things, 1st is make sure your plane has the ability to stop the spinning, 2nd is this:

Don't mean to question your vast experience, but just pointing out that your design is making it easy for itself to enter spin-out.

[Red Iron Crown]

I meant two things, 1st is make sure your plane has the ability to stop the spinning, 2nd is this:

http://i.imgur.com/Cs1SdnI.jpg

Don't mean to question your vast experience, but just pointing out that your design is making it easy for itself to enter spin-out.

Ah now I get you thanks to the diagram, point the engines through the CoM so that there is no resultant torque if one has less thrust than the other, at the expense of slightly less thrust due to cosine losses. Can engines be mounted that way without turning on part clipping? Might not be ideal for this design anyway as I'd have to turn them in 30 degrees or so.

As for my "vast" experience, almost all of it is with conventional rockets and spacecraft, I'm still a newb when it comes to spaceplanes where I defer to your much greater experience.

[r4pt0r]

One thing i read before is that you might get a-symetric thrust/intake problems if your intakes are placed using the symmetry feature. If you place them individually the game reads them properly or something.

[Kasuha]

Can engines be mounted that way without turning on part clipping?

Definitely. Surface-mount a cubic octagonal strut in place where you plan to put the engine. Note it must be surface-mounted. Rotate it the required angle, then rotate it 180 degrees more so it clips inside. Then attach an engine. You don't even need fuel lines.

Maybe it can be done better way (without a crack), this is just the first that came on my mind.

In case of your plane, though, the angle will be relatively large and so will be cosine loses.

[Red Iron Crown]

One thing i read before is that you might get a-symetric thrust/intake problems if your intakes are placed using the symmetry feature. If you place them individually the game reads them properly or something.

It should be fine in this case because I made a complete side pod with intakes and the jet engine, then duplicated that with symmetry.

Definitely. Surface-mount a cubic octagonal strut in place where you plan to put the engine. Note it must be surface-mounted. Rotate it the required angle, then rotate it 180 degrees more so it clips inside. Then attach an engine. You don't even need fuel lines.

Maybe it can be done better way (without a crack), this is just the first that came on my mind.

In case of your plane, though, the angle will be relatively large and so will be cosine loses.

Thanks for that tip, I'm adding it to my bag of tricks. Maybe I can get away with just pointing them closer to the CoM instead of straight through it, that should reduce the torque compared to having them longitudinal while having lower cosine losses.

I definitely need to play with this design more tonight.

[Overfloater]

...at the expense of slightly less thrust due to cosine losses. Can engines be mounted that way without turning on part clipping? Might not be ideal for this design anyway as I'd have to turn them in 30 degrees or so.

I never used part clipping,

and this technique works well if the jets are further behind, so they are almost straight.

Like in my plane:

I made the plane long, so the jets are only on a slight angle and no compromising thrust.

[AlexanderB]

For me, all the plane-whackyness has been solved since I learned about placement order of air intakes. "plenty of intake air" means 1 engine can possibly still hog all the air, while the other is starving off and reducing thrust, causing a flat spin. Its all in the placement order.

I've had flame-outs on 20km altitude with 0.11+ intake air still showing on the bar. One engine hogged all the air, the other starved. Fixed the build order, and presto, I could nearly orbit on the two jets alone. (made 100+ by 35 orbit, just coast to AP, and circularise. Burn a little during the coast if needed to preserve the AP height.)

[r4pt0r]

For me, all the plane-whackyness has been solved since I learned about placement order of air intakes. "plenty of intake air" means 1 engine can possibly still hog all the air, while the other is starving off and reducing thrust, causing a flat spin. Its all in the placement order.

I've had flame-outs on 20km altitude with 0.11+ intake air still showing on the bar. One engine hogged all the air, the other starved. Fixed the build order, and presto, I could nearly orbit on the two jets alone. (made 100+ by 35 orbit, just coast to AP, and circularise. Burn a little during the coast if needed to preserve the AP height.)

that sounds like the stuff i was referencing. keep us updated on the build.

[Red Iron Crown]

For me, all the plane-whackyness has been solved since I learned about placement order of air intakes. "plenty of intake air" means 1 engine can possibly still hog all the air, while the other is starving off and reducing thrust, causing a flat spin. Its all in the placement order.

I've had flame-outs on 20km altitude with 0.11+ intake air still showing on the bar. One engine hogged all the air, the other starved. Fixed the build order, and presto, I could nearly orbit on the two jets alone. (made 100+ by 35 orbit, just coast to AP, and circularise. Burn a little during the coast if needed to preserve the AP height.)

I'm actually following my own advice on this design, by building a side pod complete with intakes and engine, then placing that with symmetry.

Thanks for the advice anyway, and I'm trying to use a similar ascent profile.

[Claw]

Crown,

I flew your plane.

First off, I want to say that I like the aesthetics of your design.

- I think your design is lacking a little yaw strength, but really it flies fine the way it is.

-- Even if you added more tail fin, it might not help your spinouts since they happen kind of higher altitude. You would need more SAS.

- You might want to consider adding a little more lift rating. When you start getting up high, it takes a lot of AoA to keep it climbing. I think the AoA exacerbates intake problems.

-- Typically I aim for 0.5 lift rating per ton with stock aero. You can get away with a little less than this with your setup since you have a bit of excess TurboJet thrust.

+ I know you said you watched very closely for this, but I encountered asymmetric thrust. I'm not sure if it is the AoA or just the way the air system works, but I definitely was experiencing asymmetric thrust problems. Depending on how aggressive the climb out was, the thrust difference was as low as 4kn (not very aggressive) to a maximum of one full engine flaming out. At 4kn, the plane was controllable with some extra inputs. It would not fly straight by itself and wanted to flat spin when it initially happened. I'm not sure how fast it wants to spin out on you, but at 4kn the initial part of the spin has a fairly slow onset rate.

-- If you didn't know, you can encounter low grade thrust "rollback" before full asymmetric flameout. I think Squad did that to allow some room to recover from asymmetric thrust conditions. At high altitude, it isn't always enough.

Something else I noticed during testing is that thrust on one of the engines would start to flicker between full thrust and something lower. As if it's right on the edge of not getting enough air. But this was only happening to one engine. So, like I said, it might be the AoA causing intake problems, or it could just be KSP's asymmetric thrust. Or perhaps you could try removing and replacing the intakes/engines to see if you can reset the intake/engine pairing.

Throttling back did the trick and I was able to make orbit with a combined LV-N and TurboJet reduced throttle. You might also want to put the TurboJets on an action group to shut them down symmetrically.

Fly your profile again. Right Click on one engine, then ALT+Right Click on the other. This will allow you to watch the thrust on both engines while you fly your profile. See if the same thing is happening to you.

Cheers,

~Claw

[Red Iron Crown]

Claw, thanks for testing and the detailed report. Have some rep!

I had a heck of a time getting the CoL where I wanted it to be so I'm not sure how feasible it will be for me to add more lift. The 0.5 lift per ton is a useful rule of thumb that I wasn't aware of, I've just been throwing wings on and testing so thanks for that.

I was watching for asymmetric thrust the way you describe but didn't have any difference until the plane was already spinning. It is possible that it may be PC performance related (the MET timer is almost always yellow for me), maybe the display can't update fast enough.

I thought I did have the jets and intakes set to toggle on action group 1, but I may have messed it up (never got far enough to actually toggle them so I wouldn't have noticed).

I'll do some more testing tonight and see if I can make it work more reliably. Thanks again.

[Claw]

Claw, thanks for testing and the detailed report. Have some rep!

Thanks!

I had a heck of a time getting the CoL where I wanted it to be so I'm not sure how feasible it will be for me to add more lift. The 0.5 lift per ton is a useful rule of thumb that I wasn't aware of, I've just been throwing wings on and testing so thanks for that.

Yep, CoL/CoM is always a challenge. Since it sounds like you haven't heard them already, I've pasted my space plane rules of thumb below.

I was watching for asymmetric thrust the way you describe but didn't have any difference until the plane was already spinning. It is possible that it may be PC performance related (the MET timer is almost always yellow for me), maybe the display can't update fast enough.

That's possible. It happens pretty fast for me. If you have performance problems, consider adding more torque (maybe via the SP+ probe body). This could give you a bit more time to react. I will say that even at 4kn differential, the plane wanted to spin pretty quickly. I think that's because you don't have a lot of yaw inertia with your design, but there isn't much you can do about that.

I thought I did have the jets and intakes set to toggle on action group 1, but I may have messed it up (never got far enough to actually toggle them so I wouldn't have noticed).

Then that's my fault. I simply didn't notice the engines in that action group.

~Claw

These provide a starting point:

- Generally 1 TurboJet per 12 tons (Up to a maximum of 15 tons.)

- ~150 units of fuel per TurboJet (no less than 100 units)

- 0.5 lift rating per ton (no more than 1.0 lift rating per ton)

- 3 ram-intakes per TurboJet

Generally speaking, if you increase performance in one area, you can decrease it in another. So if you have excess thrust (say 1 TurboJet but only 9 tons), then you can ease up a little on the lift requirements, fuel, or intakes.

I use these numbers while flying a launch profile that roughly equates to flying stock "terminal velocity" all the way up (maybe slightly faster, depending on the ship). If you level off too much in the 25km range, it can sometimes be difficult to get your vector going back up again.

If you get curious as to why I use the above rules of thumb, I can explain how I came up with all of that if you care.

[Wanderfound]

Claw, thanks for testing and the detailed report. Have some rep!

I had a heck of a time getting the CoL where I wanted it to be so I'm not sure how feasible it will be for me to add more lift. The 0.5 lift per ton is a useful rule of thumb that I wasn't aware of, I've just been throwing wings on and testing so thanks for that.

This doesn't work in FAR, but it does work in stock: winglets. Winglets everywhere. Give yourself some X-wing style and stick a pair of angled winglets in as many places as you can find. To avoid disturbing CoL, balance every forward winglet with a rear winglet and every dihedral with anhedral.

You can double your lift without increasing your wingspan if you do it right. Doesn't look bad, either. Don't try it in FAR, though.

[Red Iron Crown]

I've done some further testing, thrust asymmetry was definitely the problem. I've reworked the design a bit heeding the advice in this thread, particularly Claw's. I added moar SAS, moar wing, moar tail area and moar intakes and have had good results. Thanks everyone, switching this to "Answered".

Happy flying!

Edited August 26, 2014 by Red Iron Crown

[Wanderfound]

I've done some further testing, thrust asymmetry was definitely the problem. I've reworked the design a bit heeding the advice in this thread, particularly Claw's. I added moar SAS, moar wing, moar tail area and moar intakes and have had good results. Thanks everyone, switching this to "Answered".

https://dl.dropboxusercontent.com/u/61004449/ksp/sp%2B/screenshot194.png

Happy flying!

Looks sweet; nice.

Stick a docking port in there and bring it to the races on Minmus.

BTW, folks: the easy way to avoid assymetric thrust issues on a pure RAPIER plane is just to set an action group to change modes and disable automatic switching. As soon as the thrust gets wobbly, flick to closed cycle for a moment and you're back to symmetry. Then either boost for space or stabilise, drop the nose, throttle down and change back to air-breathing mode. Same trick works for central-turbo builds.

I'm surprised I don't see more central RAPIER / lateral aerospike setups. Even turbo/spike works. Try it, you'll be happy.

Edited August 26, 2014 by Wanderfound

[AlexanderB]

I don't see the point in aerospikes for most SSTOs, my (smaller) planes get to orbital speeds on turbojet alone, no point in carrying a 1.5 ton aerospike (or multiple) if two, 90 kilo 24-77's will do, even if they have 25% less ISP, their weigth advantage more than makes up for that.

The only real advantage a higher ISP engine would have, is if you'd fly the whole airplane to another planet, but i'd pick the nuke for that.

Most of my planes are just for reaching LKO and back, and delivering standalone upper stages, cargo (fuel), or kerbals there.

I'm actually following my own advice on this design, by building a side pod complete with intakes and engine, then placing that with symmetry.

Thanks for the advice anyway, and I'm trying to use a similar ascent profile.

Ah, you're right. still, others might not have seen the advice?

It was very profound for me, I was unable to build a stable multi-engine aircraft, and a single jet or rapier only lets me take a few kerbals into orbit, now I can build a lot bigger and better.

[StrandedonEarth]

Thanks and more rep for Claw and Red Iron Crown for a useful thread that helped me turn my flying kludge into an orbiting kludge.

Edit: Oh look, I made ground crew! Woohoo!

Edited September 5, 2014 by StrandedonEarth

[Wanderfound]

I don't see the point in aerospikes for most SSTOs, my (smaller) planes get to orbital speeds on turbojet alone, no point in carrying a 1.5 ton aerospike (or multiple) if two, 90 kilo 24-77's will do, even if they have 25% less ISP, their weigth advantage more than makes up for that.

The only real advantage a higher ISP engine would have, is if you'd fly the whole airplane to another planet, but i'd pick the nuke for that.

Most of my planes are just for reaching LKO and back, and delivering standalone upper stages, cargo (fuel), or kerbals there.

A few reasons for the Aerospikes:

* See sig below; FAR in effect. Ferram's alteration of the speed/thrust curve makes it difficult to get a pure air breather much above Mach 5 these days, especially if you don't want to waste ten minutes on a run-up.

* I don't have any trouble getting my designs to orbit (or interplanetary), so there's no reason to sacrifice coolness or sporty performance for ultimate efficiency. My stuff tends to be either "flying sportscar" or "orbital Concorde"; even the utilitarian things are intended to be fun to fly and very, very fast.

* Aerospikes look way nifty, and having a big chunk of extra thrust available that is relatively fuel efficient in both space and atmosphere is a bonus. I use them as if they were afterburners for the jets; going to space is not the only use for rocketry.

* I don't like LV-N's, for self-imposed-handicap reasons, so I don't use 'em much. On things like the Migration Aerotrain, the Aerospikes are intended for use as the interplanetary transfer engines (while still making a substantial contribution to atmospheric performance, and having enough grunt that it doesn't take forever just to boost for Minmus).

[FleshJeb]

I'm getting better at spaceplanes and I think I've got the basics down, but I still have some flat spin problems with my current design between 25km and 30km as it slows its climb and starts to build speed. The CoL is just behind the CoM, and the plane has a twin tail arrangement with control surfaces for yaw authority, gimbals on the jets are enabled, plus there is a fair amount of torque available from the pod and probe core. It's stable at lower altitudes and speeds, but once the jets get into the meaty part of the power curve it starts to yaw in one direction or another and eventually spins. If I throttle back a bit it can limp up to 30 km and then fly fine. What am I doing wrong?

I know this is already listed as Answered, but I like aesthetics of the original craft too much to let it go.

The evidence suggests that you actually have TOO MUCH yaw control. At max power, those gimbaled engines are providing an overwhelming amount of yaw force, and they're overcorrecting, causing the oscillation. Set yourself an action group to toggle the gimbaling and see how that goes.

[Engineer of Stuff]

The parts that you described putting in the cargo bay are massless and physicless and it does not matter if they are on the outside of the craft. They will change nothing.

[Wanderfound]

The parts that you described putting in the cargo bay are massless and physicless and it does not matter if they are on the outside of the craft. They will change nothing.

I think that was aimed at me (quoting relevant text = good), but again, see sig: FAR assumed.

Batteries etc do affect drag in realistic aero. And, even in the soupmosphere, the cargo bay trick still has aesthetic value, and a short cargo bay is a fairly low-mass way of enhancing stability with a lengthened fuselage.