[Stock Helicopters & Turboprops] Non DLC Will Always Be More Fun!

[Mareczex333]

Simply, awesome!!

[ThorBeorn]

This is so cool Azimech. But how do you get the props to rotate? I mean you must attach them either using surface attach or node attach right? Or are you using the rotating docking part from Infernal Robotics or something else, rotating joints?

I saw you used landing gear to keep the propeller in place in one engine, but how did you get it in there in the first place without using a "stiff" attachment?

[Azimech]

I use a separator :-)

[ThorBeorn]

I thought about that for a brief moment. "Hmm could be that, but no thats too ridiculous". Well there you go. I think I'll use the docking wafer or whatever it's called from IR instead. That's not too cheaty right?

Edit: So I built some engines using the docking washer from IR. The most overengineered one had eight blowers (20 kn static thrust each), 32 fan blades and four propeller blades. Two of these babies got my aircraft up to a whopping 15 m/s on the runway. I guess there's room for improvement here...

Edited August 4, 2015 by ThorBeorn

[Azimech]

How's development going? KSP 1.1 will bring new jet engines, so I might get back into building them.

[BlueCanary]

How's development going? KSP 1.1 will bring new jet engines, so I might get back into building them.

I'd imagine 0.625m jets would be perfect for this, especially since they seem to have really good TWR

[Azimech]

That is important, but static thrust as well.

[BlueCanary]

That is important, but static thrust as well.

Only 41kn, but still at a quarter of the weight and like an eighth of the volume they might have a whole lot less work to do.

EDIT: Here it is:

EDIT 2: Edited wrong post, sorry

Edited September 5, 2015 by BlueCanary

[Azimech]

I'm curious ... where did you find these numbers?

[BlueCanary]

I'm curious ... where did you find these numbers?

In the imgur album for 1.1, it's all still very much subject to balancing I guess but you can see most of the stats (no isp unfortunately) in one of the screenshots.

[Azimech]

41kN static thrust would be most enjoyable ... this is showing it at Mach 1.7, same as the wheesley. I think I'll start a petition on Reddit.

[FlipNascar]

It was time to join in on the helicopter fun...

So I started tinkering, and can get it to rotate quite happily. Until of course you add rotor blades, then it slows it down massively. And with this being my first foray into stock turbines... I have a working "helicopter" that relies on Twitches to spin the rotors and it works. I can fly it from the Runway around the KSC happily enough before pilot error kicks in. Ultimately I wanted to try and find a way to create a working swashplate so that you can say goodbye to the reaction wheel spam. But not found the right way to do that yet. I'm wondering if redone wheel physics may help in 1.1. A promising prospect was using inverted rovermax 2 wheels - but I wanted to lock their suspension, can't. Don't know how to mod them to do that. But the ultimate problem was keeping them away from the plate and gently adding them through control inputs using control surfaces. But you can't attach stuff to control surfaces...

Bearings seem solid enough, although what I haven't tried yet is to hybrid it - add some Twitches to the rotors, nor what makes the best rotor blades, nor how many, nor their static AoA.

I suspect that it's got something to do with the relative lack of rigidity in the support trusses, but when I strut them, then it just causes more issues. In principle, once the rotor disk loads up the weight and is dragging the rest of the heli this relative lack of rigidity doesn't seem to affect handling - I presume, based on tests from the powered rotor heli. Although while writing this, I went and added a bunch o' struts and it worked... It didn't fly, but it stayed together at full power.

More info ramblings in the spoilers and links to craft files. If anyone feels like firing them up and passing on your critiques, I am all ears. I am way into unexplored territory here... The concept is simple enough. But making it all come together... And yes, the designs are fugly because they're unrefined.

My tech demo/proof of concept. Boring and basic as you like. But it'll spin. And it'll go like the clappers - show me the autobhan fast.

Craft file here (most likely doesn't feature that rotor disc. Went through all sorts of variations, but this is the one with way too much mass...)

I suspect that in reality I am asking too much of the bearing in terms of speed, and that I am trying to compensate for the "small" rotor disc by spinning it faster.

Can't remember if I actually made this work. My gut feeling is no. It probably failed spectacularly because of having the "blowers" that high up the mast. So I started from scratch - this is the "rotor system" from my only flyable (hah!) stock helicopter.

Like any great helicopter this thing looks so ugly that my lexicon fails me. Craft file here

I spawn my Kerbals with Take Command...

To operate:

Infinite Fuel Cheat, set about 75% power, RCS on.

Stage.

SAS On, Set power about 50%, stage.

It should wobble around a bit. Okay a lot, and you should have maybe enough lift to play in what would be ground effect. Downside is, you have to switch between the rotor disc and the main craft to adjust power. A notch or two about 75% will give you flight. It doesn't do high-speed or anything like that. You can use main craft's throttle as a pseudo tail rotor, but of course the lateral drift is just huge, so this is why I wanted a swashplate. Trim in a bunch of left cyclic to counteract that lateral thrust - if it's good enough for a real helicopter, it's good enough for my KSP aberrations.

The fuel hose is a left over from trying... something. That clearly didn't work. Although this is quite the fun little death trap...

- - - Updated - - -

Edit: I should probably go crunch numbers from the working "rotors" to figure out what I need in terms of "lift units" to lift mass... hmmm.

[BlueCanary]

FlipNascar, I heard somewhere that wing lift is applied from the root, so it might be that your first rotor disc was taking the very slow speed at the roots of those wings and applying lift based on that, whereas if you flipped them around it could be much more effective due to higher speed at the tips.

[Redshift OTF]

I've gone back and had a brief fiddle in the hope I can get helicopters to work again but haven't had a lot of luck. I'm not sure if they toned down the Jet thrust or if it's a case of the Jet engines are too weak when static. Also I'm not sure if those new small gear wheels have too much friction. Any thoughts?

The smaller jets should help in the next update.

[FlipNascar]

The smaller jets should help in the next update.

This is what I'm wondering. And if with the reworked wheel physics it may be easier to get this sort of thing to work.

@BlueCanary, no idea if that will work. Will try it out.

[Azimech]

FlipNascar, download my propeller thrust tester, Part Angle Display and Infernal Robotics, with those three you can find the ideal propeller/rotor blade pitch. Also fiddling with blower angle helps a lot.

http://kerbalx.com/Azimech/77I-Propeller%20Thrust%20Tester%2011

[BlueCanary]

Is there any way to make variable pitch props in stock? I made a tip-jet powered helicopter but the lag between tipjet throttling and rotor speed makes it very hard to control. I was thinking varying rotor pitch a la real life would work, but if not what about running the rotor at a speed slightly too low to support lift and using auxillary lift jets to provide control?

Edited September 6, 2015 by BlueCanary

[FlipNascar]

BlueCanary: Try it and see...

But, and here is where it's been far too long since I read about all this so I may well be wrong... The issue with trying to control the pitch on the rotor blades is that you need to then be able to control each blade independently. Otherwise you're not actually going to get a proper control input. Furthermore... You also then need some wizardry that will adjust the blades as they [the blades] change from advancing to retreating. Secondly, if you then take the load off the bearing you are probably going to get RUD as you use the VTOL assistance, unless you also design the bearing to be solid and not move when it's at less than 1G (relative to the rest of the craft). That would be my guess... But I'm no rocket surgeon.

However, if you could create some such way, then you could get much greater efficiency out of turboprop aeroplanes. You could set a fine pitch for take off and the coarsen that puppy out for the cruise....

The way I see to do this in KSP is that you need to mimic a helicopter's swashplate, so that you can create the effects of changing the pitch. But the swashplate... Well that's the kicker, how to create that in stock? You need a little bit of freeplay in your bearing, then you need to have something that pushes on the rotor disc's assembly. So when you want to go left/roll left you need the right side of the rotor disc to be pushed up so that the rest of the helicopter gets dragged along.

Fundamentally a real helicopter is just being dragged along by the rotor disc. When you apply control inputs you directly affect the rotor disc, and then a few moments after you make your input the rest of the helicopter catches up.

What I've tried in terms of swashplates... create a rotor that will spin and spins quickly. Then using oxstats I created a plate around the outside. Now there's something to push against. I tried using control surfaces attached to wheels and offset, but a) they are backwards for roll inputs and they just catch on the plate and get dragged around. Can you strut an elevon?

Of course the real kicker in all this - assuming that it ends up being similar to real helicopters - you need a very delicate control input. Oh wait reaction wheels + SAS! Woo!

Among fixed wing pilots there's the expression that a helicopter doesn't fly, it's just so ugly the Earth rejects it. There may well be some truth in that...

Pic below shows one idea I had for the swashplate. It didn't work. And that's not because my rotor disc flew away... You can't directly affect the suspension of the rover wheel.

Azimech: Will do. Although I'm wary of 1.1 and wondering just how close that is, and what wheel physics update will do to bearings that exist.

PS: I apologise for the wall of text.

[BlueCanary]

I assume gyroscope effects make it too hard to just brute force tilt the entire rotor disc with something?

Also maybe there's a way you could just use control surfaces as rotors and somehow find a way to tilt them at the right times? Maybe KOS would have a way to do it? Or by making it a chinook style or even quadrotor system you could just adjust pitch for the whole of one rotor to get control?

[FlipNascar]

TLDR: It has to be brute forced, but, how do you brute force it when it's actually kinda delicate, and it's moving PDQ.

Brute forcing it is what I've tried unsuccessfully (but only because I cannot find a way to actually input the brute force onto the disc without breaking the input thingy or the disc). Realistically within the limitations of stock and KSP I think that a brute force approach will be the only way to do that.

The issue with using the control surfaces is using the control surfaces. Don't forget that in KSP it's not a single craft. It's two craft that are interacting together. With the rotor assembly spinning how can you affect a proper control input - I tried it with my twitch rocket powered rotor disc. If you could find someway to have the probe core controlling the rotor disc not be spinning like the rotor disc, then perhaps you could get it to work. But then surely you're back to square one because you've then got a separate craft that you're trying to control the rotor disc.

I swear I don't mean to be the negative nancy...

If you download the craft file in my earlier post of the fugly jellymacopter with the rocket powered rotor, that rotor is made of control surfaces. And you can enable them and try it. There's a probe core on top... Hmmm maybe changing the way the probe core is aligned it could work?

But then you're back to the original problem... I have to switch between craft to adjust power or direction...

However, I don't know what would work. I've tried some things that didn't work... Maybe if you use SAS and set the rotor's probe to point a direction - I can't visualize which way would work, would perhaps be dependent on rotor direction too? And then can you use that in some way to stabilize it?

Please feel free to prove me wrong! No really, then you'll solve one of my problems for me...

[BlueCanary]

So if there was a bearing that didn't require craft to be seperate it'd have a lot more chance? Maybe using structural pylons, I saw something somewhere about them being used for rotation.

[FlipNascar]

Bonne chance mon ami. Let me know how your kludgery goes. However, without being the Debbie downer... Unlikely. But try it. One thing of I've learnt with KSP is to never say never. Even if it, to me, seems preposterous.

[selfish_meme]

So if there was a bearing that didn't require craft to be seperate it'd have a lot more chance? Maybe using structural pylons, I saw something somewhere about them being used for rotation.

I tried a tank out that used structural pylons, the turret sank into it, i dont think they are rigid enough.

- - - Updated - - -

TLDR: It has to be brute forced, but, how do you brute force it when it's actually kinda delicate, and it's moving PDQ.

Brute forcing it is what I've tried unsuccessfully (but only because I cannot find a way to actually input the brute force onto the disc without breaking the input thingy or the disc). Realistically within the limitations of stock and KSP I think that a brute force approach will be the only way to do that.

The issue with using the control surfaces is using the control surfaces. Don't forget that in KSP it's not a single craft. It's two craft that are interacting together. With the rotor assembly spinning how can you affect a proper control input - I tried it with my twitch rocket powered rotor disc. If you could find someway to have the probe core controlling the rotor disc not be spinning like the rotor disc, then perhaps you could get it to work. But then surely you're back to square one because you've then got a separate craft that you're trying to control the rotor disc.

I swear I don't mean to be the negative nancy...

If you download the craft file in my earlier post of the fugly jellymacopter with the rocket powered rotor, that rotor is made of control surfaces. And you can enable them and try it. There's a probe core on top... Hmmm maybe changing the way the probe core is aligned it could work?

But then you're back to the original problem... I have to switch between craft to adjust power or direction...

However, I don't know what would work. I've tried some things that didn't work... Maybe if you use SAS and set the rotor's probe to point a direction - I can't visualize which way would work, would perhaps be dependent on rotor direction too? And then can you use that in some way to stabilize it?

Please feel free to prove me wrong! No really, then you'll solve one of my problems for me...

I found that using sas can impart more velocity, so by adding or detracting from the force using sas you can somewhat control thrust without having to control a different part, but it puts more stress on the bearing.

[Azimech]

Here's something I did in February. Using stock parts you have a number of problems.

1: The joints system is too flexible and it's flexibility is only a factor of part mass. This is ridiculous.

2: Moving parts are too slow for helicopter operation, they have a fixed animation frame rate (i.e. fixed pulse width modulation)

3: In my example I used landing gears, but there's no way to control the animation to stop half way (i.e. no amplitude control).

4: Control surface operation is always a slave to root part orientation. I once had the idea to put flaps on my propeller blades for pitch control.

5: Colliders aren't being respected enough. This might improve by changing the physics/fps ratio but already in 0.90 part count meant a slow down.

Things aren't much better in 1.0.4.

Javascript is disabled. View full album

[FlipNascar]

Kudos for going as far as you did down that road. Would love to see a craft file so I can pull it apart... If that's not too much to ask for.

In my mind, there's got to be someway in which you can manipulate control surface inputs (because they're modulated) and translate that motion onto a rigid object that will push against the disc. Obviously it'll be backwards, but if the theory can be made to work, how hard can it be to make it all work the otherway... Constantly fly backwards from a probe core? Down will be up... Left will be Right. Which is what you want.