• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

Need help with my overly ambitious Tilt Rotor project

#1
So I've taken on quite the project. I'm building an AW609 Tilt rotor as a platform to develop Tilt rotor specific control software. The In order to fit all of my equipment I made this thing pretty large. It runs 2 Cobra 4130/16 390kv motors with Master Airscrew 16x10 3-Blade props. Now if you're not already familiar with how Tilt Rotor aircraft work then take a minute to google the v-22 osprey or the version I'm building, the AW609. Note the nacelles and how they pivot. During takeoff the motors in my aircraft are directly above the pivot point providing downward thrust. They can(should) rotate 20 degrees aft, and 90 degrees forward of this upright position. Now here's my big question. How should I drive the nacelles? Imagine I have a standard servo as a direct drive inside the nacelle (they fit) on the pivot point. During transition we rotate these motors to 45 degrees. I now have a constant torque load on the servo. With these motors and props weighing about 1.2 lbs each, what sort of equipment will I need to pull this off. I know there are servos powerful enough to do it, but I'm uneasy with the amount of current draw that comes from a sustained load like this that is a major part of the overall function of the aircraft. There's no need for high speed transitions in servo angle either. What would be perfect is a servo built for low speed, somewhat high torque, and good efficient holding power. Looking at specs just isn't helping me make a decision so hopefully some input will help guide me along here. Thanks! I'll post pictures of what I have already.
 

PsyBorg

Wake up! Time to fly!
Mentor
#3
You could run a carbon shaft to hard mount the motors to. Split that in the center in the fuselage and use a ring and pinion set up on a constant run servo. Then just set min / max revolutions to achieve desired angles at the motors. The torque and speed of the rotations can be determined by gear ratios.
 
#4
I've taken a look at gearing within the nacelles and also using a drive shaft coming from the fuselage. The first thing I don't like about the fuselage approach is it adds a great deal of complexity to the design. The AW609 wings are swept up at 3 degrees and forward 5 so I have to add in a U joint. Did some playing around with this with some RC car U joints and just wanted to avoid that route. Also the amount of space I have to work with is pretty limited since the drive shaft comes so close to the top of the airplane. In order to gear down I need the larger gear on the shaft... it gets messy pretty quick. I would still consider gearing options for the nacelle if someone has a good sit / resource for tiny gear sets.

Thinking through quite a bit of those scenarios and designs is what lead me to want to go direct drive with a hefty servo. The 1lb load sits about 2.3 inches from the pivot point so there's hardware that can do it and keep the design super simple... I just don't know about the power consumption that is going to come from that kind of constant load.
 

quorneng

Active member
#5
Just out of interest do you have a method to control roll, pitch and yaw when in hovering mode?
This control gets even more involved during the transition to conventional flight.
 
#6
During a hover yaw is controlled by rotating the nacelles against each other. So say the neutral straight up position is 0 degrees. I need to yaw right. The right nacelle rotates to 2 degrees (rotates forward) and the left nacelle rotates aft to -2 (points at the tail). Roll is controlled through differential thrust. There's really no way to control pitch unless I move the CG so while entering the hover I need to detect which way its leaning from the get go. If the CG is nose heavy, the neutral position is now starting at -2 degrees and the aircraft will hover with its nose pointing at the ground until there's enough air over the tail to bring the elevator into play. Now imagine I'm flying with the nacelles at 45 degrees forward. That means that differential thrust now effects yaw and roll as does rotating the nacelles against each other. I'm hoping at this point that all these effects transition linearly and that the control surfaces will be able to come into effect in time.
 
#7
I got those yaw numbers backwards. Ironically something similar happened in a test. Forgot that the servo values needed to be mirrored and sent the thing into a spin. Started simulating the code before physical tests to avoid any more duck and cover situations.
 

quorneng

Active member
#8
A bit concerned at your lack of pitch control.
Rotating wings are not intrinsically stable so you will need either some form of cyclic pitch control (like a helicopter) or an additional horizontal prop at the rear (like a multi copter). Either system is almost certain to need full gyro input to get the control responses down to a 'human reaction' level.
Then you have the changes in the props centre of thrust as you translate into forward flight.
You could be in for a long development!

Please keep posting on how you get on. I will be following with interest.
 

PsyBorg

Wake up! Time to fly!
Mentor
#9
The angles you mention are small enough that a single shaft / split shaft could still easily be done using a bearing at the nacelle at the lowest and thickest part of the outer wing and another closer to the fuse set close to the rear most edge of the wing. The pivot could be standard issue fulcrum / lever used in normal elevator set ups. I don't think torque would be an issue with the bearings taking up the major resistance and weight. I also think the rotation of the motors will give a gyroscopic effect while spinning lessening any stress of the rotation.

Similar to this effect


From the design of the plane you are modeling your build from all pitch control is balanced neutral by wing placement over the planes fuselage at cog on the x axis. Placement of necessary components ( load management) would further aid this. The real plane does not have any external controls for this which is why I think this way unless there is a sliding internal counter weight that is shifting to control pitch while in a hover which again would be easy enough to do with a continuous turn servo on a screw drive that shifts the load (batteries) fore and aft.
 
Last edited:
#10
Yes the lack of pitch control before a decent IAS is a concern. That said there's a bit of software magic planned to account for the lift off differences that might lead to a stable nacelle up (and negative) / nose down hover. As for what PsyBorg hinted at .. a dynamic center of gravity is possible but not preferred. I've considered it and don't want to rely on shifting the battery for level flight in order to take off. As far as all these points go the weight isn't my issue as much as the current draw from the components like a single servo holding a 1lb motor at a 45 degree tilt for extended periods of time. This thing is seriously overpowered at the props so I may just need to divide the nacelle and thrust batteries, it can lift both. Unfortunately I think the big limitation on this airplane will be run time. It may only be able to run the servos and motors for a few minutes at a time considering all the stress events involved.


BTW I saw that video a while back and keep wondering if the gyro effect plus thrust will negate my fears of the load on the servo.
 

PsyBorg

Wake up! Time to fly!
Mentor
#11
I would seriously think the stress on the nacelles against the servo would be less then what happens at a rudder or elevator with it having to fight against wind trying to pin it flat again. The props are pulling the weight of the plane and are what directs the forces thus not producing any lateral resistances. The only thing it should be exerting any force on is the pivot point where it rotates and that should only be on the lead edge of the direction the motor is pulling. The only time it should see any resistance to the servo is when transitioning from one angle to the next and I think most of that will be negated by that gyroscopic effect.

The one you should really get in contact with is Peter here at Flite Test. He is the guru on all things tilt wing. That's how he got involved with Flite Test to begin with was with a home design tilt wing / rotor aircraft.
 
#12
I would seriously think the stress on the nacelles against the servo would be less then what happens at a rudder or elevator with it having to fight against wind trying to pin it flat again. The props are pulling the weight of the plane and are what directs the forces thus not producing any lateral resistances. The only thing it should be exerting any force on is the pivot point where it rotates and that should only be on the lead edge of the direction the motor is pulling. The only time it should see any resistance to the servo is when transitioning from one angle to the next and I think most of that will be negated by that gyroscopic effect.

The one you should really get in contact with is Peter here at Flite Test. He is the guru on all things tilt wing. That's how he got involved with Flite Test to begin with was with a home design tilt wing / rotor aircraft.
1. "Less than what happens at the rudders"
Somewhat right. Rudder input is a limited high amp problem just like the elevator. But the Tilt rotor is continuous.. in other words it "seems" extremely expensive to hold a 1lb motor at 45 degrees for a long period of time on a 5 cell battery. My original post was to inquire on specialty servos that my be better suited for holding that kind of load.

2. Prop pull
I hope this elevates a bit of stress. they won't be lifting thattt much compared to the average airplane. Two arduino's and an iPhone seem .. reasonable.

3. Gyro "lift"
Maybe that's my third savior on energy draw. I can tell you these 16-10-3 Airscrew props generate a TON of torque. As long as they're not grinding the tilt axis that may help the servos with the inherent resistance to changing direction.

4. Peter.
I grantee he'll love this project but would be smart enough to stay away until I sort out this rats nest of problems :p
 
#13
ACorlett48
If you can arrange that the prop thrust line goes through the nacelle pivot point and that the CofG of the nacelle is on the pivot point there would then be no significant thrust on the servo regardless of the angle of the nacelle.

"Better to eliminate a problem rather than to overcome it".
 
#14
Okay one problem I've had is how to run the wires into the nacelle. I've tried all sorts of cutout designs through the nacelle and the end of the wing so that they remain hidden. Getting the wires to move the full 120 degrees always has been an issue and I don't like them rubbing/twisting that much. SO, I finally decided to go off the 609 design and run exterior wires out the bottom of the wing to the nacelle and use some good lookin' wire wrap to make it look intentional. So there we go, one problem down. I'm taking your advice and making sure the thrust line goes right though the pivot point. That's the last of my big design issues so we are getting closer to hover tests! Pictures before then :)
 

PsyBorg

Wake up! Time to fly!
Mentor
#16
They did one over the past winter with him for that indoor event in Ohio where he had the prototype out. It was pretty interesting how that works with that in line over under configuration.
 
#17
Did you find the answer for nacelle turning? I'm working on an osprey and the current Nacelle angles are not very accurate and smooth. I've been looking into redesigning the nacelle to bring the pivot point more forward towards the motor placement where the weight is at. I've been looking at using a stepper motor for slower-smoother-more staying power as well.
What has worked for you?
 
#18
If you can arrange that the prop thrust line goes through the nacelle pivot point...
Pretty hard to do given that the prop thrust line goes through the rapidly spinny things that make that thrust. They are pretty large and that pivot point you refer to must be connected to the wing. Props and wings don't really coexist in the same plane in space very well.
 

quorneng

Active member
#19
Tabasco2931
Surely the Osprey's thrust line goes through (or pretty close to) the nacelle pivot axis.
You should do the same so the pivot servo does not 'feel' the thrust forces.

Further if you arrange the thrust line to be a bit below the nacelle pivot axis the thrust forces could actually relieve the total load on the servo. It would then only have to resist any nacelle 'weight' when the props were stationary.
 
#20
Surely the Osprey's thrust line goes through (or pretty close to) the nacelle pivot axis.
V-22-166480-FAR01.jpg

actually it's not that close at all. The force of thrust is going to be created on the tip path plane of the props which sit pretty far in front of the nacelle pivot point. The greater the distance, the longer the "pendulum arm" that the servo has to move becomes and the more weight it must rotate. They help counter that arm by having the motor sit farther back in the nacelle which offsets the weight the props are picking up. This is only a real concern as the Nacelles transfer closer to the 45* line as in vertical mode the thrust line is in line with gravity and in forward flight the "weight" of the airplane is transferred to the wings.

(They also have variable pitch props which can help offset the weight during movement as well)
 
Last edited: