Sikorsky X2 coaxial heli developments.

[Graviman]

Riff_raff,

(if that is your real name )

I imagine helicopter gearboxes use straight cut gears for improved power to weight. Helical gears were invented to reduce geartrain NVH, and might help with the noise concerns (it was mentioned somewhere). To avoid the side thrust the norm would be to use two mirror image helical teeth (herringbone). If hobbing machines required an excessive central gap then two seperate gears could be bolted or friction welded together.

Load sharing between multiple planets is an interesting problem. The first solution which springs to mind is the idea of a compliant mount for each bearing housing, restrained by a guide plate to limit radial movement - the problem would likely be how to avoid transmission resonances excited by the rotors. If a hydraulic system is allowed on the planet carrier then each planet could be postioned by a small oil hydraulic ram, so that all planets see equal load. The next problem is then how to stop the rams all bottoming out over time if the oil in the hydraulic circuit is not replenished. This implies the need for a complex mechanical load sharing arrangement. Either way the clever stuff needs to happen in the planet carrier(s).

Of course all this assumes that tail rotor drive requires the one-way clutch to act on the high rpm side of the transmission (otherwise worm gears would get a mention). Are electric motor driven tail rotors outside the scope of the project?

Mart

[IFMU]

I have often wondered why helicopter gearboxes use straight cut gears, when helical gears were invented to reduce geartrain NVH.

Is this really true? I thought that a bunch of the helo gears I've seen were helical cut.

-- IFMU

[Dave_Jackson]

Face gears were being considered for rotorcraft about 10 to 15 years ago. Are they still being considered?

Dave

[riff_raff]

IFMU,

I don't mean to drag the thread off-topic, but to answer your question about the type of gears used in recent helo drivetrains, non-parallel shaft gearsets are predominantly spiral bevel, parallel shaft gearsets are preferably double helical, and planetary gearsets tend to be spur. Each type is used for practical reasons.

While a spur gear might be best for efficiency, spiral bevels or double helicals give better performance with regards to the dynamic tooth loads produced in the typical rotorcraft geartrain with high pitch line velocities. This is due to their better face contact ratios.

MRGB planetaries tend to use spurs gears simply due to the fact that it is very difficult to assemble a planetary with double helicals. Single helicals are not used much since they always have an unbalanced thrust load which makes the bearing arrangement more difficult.

As for Dave Jackson's comment about face gears, NASA spent a huge amount of time and money trying to develop the analysis and manufacturing technology for high performance face gears. Initially they had a lot of promise, but in the end what became apparent is that they had many practical issues that limit their theoretical performance advantages. The main one being sensitivity to mesh displacements under load. Regardless, I believe the Apache block III upgrade includes a face gear in its MRGB.

Regards,
riff_raff

[Graviman]

Riff_raff,

Your comments about gearbox design are most welcome.
Anything which gives technical background will add to the discussion about X2.

[riff_raff]

Graviman,

I'm glad that you found my gearbox comments helpful.

While I know a bit about helo gearbox design, something I'd really like to learn more about are details of the X2's rotor hub and feather bearing design. To me, that seems like a very challenging piece of design work!

riff_raff

[Graviman]

Yes, more info on X2 gearbox and rotorhead would be interesting.

Originally Posted by riff_raff

As for Dave Jackson's comment about face gears, NASA spent a huge amount of time and money trying to develop the analysis and manufacturing technology for high performance face gears. Initially they had a lot of promise, but in the end what became apparent is that they had many practical issues that limit their theoretical performance advantages. The main one being sensitivity to mesh displacements under load. Regardless, I believe the Apache block III upgrade includes a face gear in its MRGB.

riff,

Regarding sensitivity to bearing/housing movement: The norm for ground vehicle axle differentials is to use spiral bevel gears for crownwheel and pinions. The spiral allows the line of contact (generally over 3 teeth) to rotate about a point above or below the axle line. This means the differential housing can flex without causing the teeth to spall at the edges. Another side benefit of this is that some worm gear action can be designed in to reduce the ratios a little more, albeit at the cost of efficiency.

Mart

[Dave_Jackson]

riff-raff,

Thanks for the information.


Mart,

The face gears were intended for ratios higher then the Crown & Pinion's maximums of; Spiral Bevel @ 4:1, and the Straight Tooth @ 6:1


Dave

[Graviman]

Apologies, Dave. I could have been a little clearer in my post.

What i was driving at is that by putting a spiral in the gears (as opposed to a simple helix or even no helix) then the the gear contact position can rotate about a point offset from the face gear axis. This means that the bearings can flex a little without causing fatigue hot spots near the tooth edges.

For this application i would imagine that it would work if the spiral centre was in plane with the driving gear axis (at contact centre) and the bearings were allowed to float axially. Any flexure in the casing would be taken up by the driving gear moving about the spiral centre to keep an even pressure.

Gear hobbing of crownwheels is done by a tool rotating at the spiral centre. Since i am not sure how face gear profile changes with radius are generated then i cannot be certain whether what i suggest is practical.

[riff_raff]

face gears

Graviman,

I believe face gear teeth can be made with a helix angle if desired, or possibly with a shaft offset like a hypoid. Here's a neat website on weird bevel gear geometries.

With shaft angles at/near 90deg, face gears begin to have serious limits on face width, due to thinning of the outboard tooth tips on the gear. Of course spiral bevel gears also have face width limits. I believe Gleason recommends no more than about 30% of the outer cone distance as a rule of thumb.

What was a big development issue with the Apache face gear design was the tendency for the mesh contact to move from heel to toe under load. Some of this can be compensated for using lead correction or face profile mods, but these changes may also reduce ultimate load capacity. The Apache xmsn engineers also tried things like cutting the pinion teeth using a slightly tapered blank to improve their face gear performance.

riff_raff

[Graviman]

Riff, that website (and the consultancy) looks pretty good:



So the geometry of face gears is similar to bevel gears, but with the addendum/deddendum radii changing across the contact zone to keep the gear flat. It makes more sense to me now.

I had wondered about contact shifting from heel to toe due to tooth separation loads: One thought that occured was to fit the pinion (if that is the correct terminology for a face gear set) to a spline cut with some spherical curvature. As long as the spline at the pinion centre has more tooth contact area than the gear contact then the stresses would be sensible. The idea is for a simple face gear to mesh with a pinion free to misalign its axis, with the spline acting as a CV joint. The balance of contact forces would make the pinion align itself to the face gear teeth, compensating for any bearing and xmsn distorsion. This could save a bit of weight in the casing etc.

I'll have to get myself onto the next Cranfield (UK) course with some of these ideas in mind. It would be nice to have a better understanding of X2 xmsn first...

[riff_raff]

Face gears

Graviman,

For the most basic form of face gearing, the pinion is simply a conventional spur gear. And that is why face gears seem appealing at first glance, the spur pinion does not require an accurate axial location like the pinion of a spiral bevel gear set would.

Where face gears become tricky is due to the change in pressure angle of the face gear tooth as you move radially along the face width. It is a lower pressure angle at the toe (inner edge) and a higher pressure angle towards the heel (outer edge). This makes them very difficult to grind accurately.

There's lots of good face gear tech data on NASA's server. If you want a good text book, get Faydor Litvin's.

Good luck.
Terry

[heli1]

Let me preface this question with the fact that I am not an engineer but I am confused regarding the X2 drive system.
OK...so theMGB has to drive the coax/contrarotating main rotor ....but also presumably the auxiliary propulsion airscrew too or is that driven separately from the engine in some way? Thre's obviously a cluch in there somewhere too to help transfer the power.
I've tried looking at patent diagrams but still can't figure it out so obviously need a primary grade explanation !

[Dave_Jackson]

Mart (graviman)

Simple envisionment: Consider the face gear as having an infinite diameter. In this scenario, its teeth would be cut to gear rack specifications and a perfect mesh would exist. This implies that very high ratios will allow for better fitting teeth.


Heli1

'..... I am confused regarding the X2 drive system."

You are not alone. Everyone, outside of the X2 project, must be confused.
Perhaps those inside the project are also confused; which is not totally unusual for R&D.



I sometimes wonder if the X2 project is nothing more than a unique marketing campaign that keeps the brand name 'Sikorsky' in public view while little else is going on.


Dave

[riff_raff]

X2 propulsor

heli1,

As far as I can tell, the aft propeller runs at a fixed speed relative to the main rotor speed, and is shaft driven from the MRGB (without a clutch) similar to a conventional tail rotor. Based solely on the recent patent applications I've seen from Sikorsky, I believe the main rotors/aft propeller of a production X2 are designed to be run at two discreet speed ranges. The MRGB can be shifted between the two speed ranges during forward flight, but probably not during hover. The main rotors/aft propeller torque (or load) split is accomplished via blade pitch control of the main rotors/aft propeller.

All of the multi-speed MRGB concepts I've seen really need at least two engines to achieve a safe, fault tolerant drivetrain architecture. The X2 TD only has a single engine. So if the X2 TD does have a gear shift function, it is likely using a friction clutch somewhere in the MRGB, as that is the only practical way to synchronize shaft speeds with one engine. While this might be acceptable for a TD vehicle, in my opinion it would not have adequate fault tolerance capability for any commercial or military application, due to the temporary interruption of drive power from the single engine to the rotor during a shift sequence.

Regards,
riff_raff

[Dave_Jackson]

Off topic ~ Face gear

riff_raff.

Perhaps there's a future for face gears.

Boeing obtained the following patent last Tuesday.
Method of making a pinion meshing with a given face gear in accordance with altered design parameters


Dave

[NickLappos]

No clutches, guys. The whole drive system is tied together, rotors and prop shaft. My guess is that the "permanently" geared prop is a variable pitch assembly that can run at flat pitch, as well as positive (forward push) and negative (aft pull) pitch. If no prop thrust is desired, the pitch is flattened. Prop thrust could be done with a separate thrust lever in the cockpit, near the collective.
The engine governors automatically increase power as prop pitch is increased.
The ABC concept went to ultra speed by slowing down the rotor systems down to about 82% rpm, because the tip mach is the bad actor at ultra high speeds. A guess is that the advancing blade will be at Mach .96 or so at Vmax, so that the tip will be going at .96 Mach - about 1000 fps - as the rotor is turning at 82%Nr at 250 knots. Also, the cyclic pitch is set so that the aft sweeping blade is essentially feathered as the speed increases, since no lift is needed on that portion of either disk (since the other disk has an advancing blade on the same side to provide the lift).
The old ABC pulled about 2.0 G at 25,000 feet, which sent the retreating sides into deep stall, but retained high cyclic control power (the advancing sides retained lots of control.)

[Dave_Jackson]

Rotor Design;

The following is an addition to Nick's comments about the excessive tip-speed on the advancing blades during high-speed cruise.



To reduce this tip-speed, Sikorsky is attempting to minimize the drag at the roots of the retreating blades during high-speed flight. An earlier patent outlined the removal of the airfoil from the root end of the blades, plus some unique blade twisting.

To further minimize this drag Sikorsky was granted US patent 7,695,249, Bearingless rotor blade assembly for a high speed rotary-wing aircraft on April 13, 2010. The new patent is for transmitting the blade pitch directives via a torque-tube directly out to the airfoil portion of the blade.

This results in a fixed pitch at the blade root, which is effectively a spar with no induced drag nor negative lift. Perhaps at a later date they may place an independent controlled airfoil over the root spar, with the objective of increasing lift in all flight realms and reducing the moment arm lengths.


Dave

[Graviman]

Dave,

Do you have any section drawings available for X2 blades? Just email me direct if you do. My heli performance program can now flap a single rotor to equality, so I'd like to see how X2 lift/downwash looks with cyclic trimmed for forward flight. It's early days so i am still developing the algorithms to handle inflow roll and counterotating configs.

---

Nick,

When you comment about S-69 retreating blades going into deep stall at 2.0g at 25'000ft how was this noticed? Was there increased vibration from tip vortex shedding (multiples times per pass) or did the drive torque go up (or autorotating collective down). If rolling to a bank was there a need to monitor the pedals to stop any uncommanded yaw?

[Dave_Jackson]

Mart,

Do you have any section drawings available for X2 blades?

Nope.

You may wish to use the information on patent application US 20070110582 to get an approximation of cross sectional data.


Dave