View Full Version : 'Thompson Coupling', a pretty interesting design.


rotorque
26th Feb 2008, 08:54
Interesting design, the ‘Thompson Coupling’ as apposed to the ‘Thomas Coupling’. It’s a true constant velocity coupling designed and built in Australia.

There is a Youtube version of the movie that the company has made that shows it going through a variety of tests etc and seems pretty convincing. Welcome to Thompson Couplings Limited - Home of the Revolutionary Thompson Coupling (http://www.cvcoupling.com) and <object width="480" height="295"><param name="movie" value="http://www.youtube.com/v/xgQgm3GwaFs&hl=en&fs=1"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/xgQgm3GwaFs&hl=en&fs=1" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="480" height="295"></embed></object>

At one particular point the commentator makes mention of its use as a fully articulated rotor head (with a one ton(ne) weight suspended from the rig) and I have to admit that I could see the potential.
Are there any heli designs out there that could apply this technology, specifically to the rotor system?

(I’m not affiliated with the company, but have to admit to being a believer in the technology…. Can’t buy any shares at the moment though :( )



Graviman
26th Feb 2008, 12:42
I'm pleased this design is still being developed. I first learned about it 10 years ago. The problem is generally a catch 22, in that major OEMs won't invest until the company demonstrates production capacity, but that requires investment...

Universal joints in drivelines are generally limited to a maximum recommended angle of 3 degrees (shaft vibration limitation), and a minimum angle of 1 degree (to keep grease flowing). Constant velocity joints can handle higher angles, since there is no shaft accel/deccel, but are limited for torque by bearing B10 life.

This design captures the best of both world by allowing higher angles at higher torque capacities. For applications like tail rotor drives i imagine this design would be very robust. Dave Jackson also had a clever application for an articulated head, although the trend now is towards higher effective hinge offset with designs to handle lead/lag forces.

NickLappos
26th Feb 2008, 13:48
The drive shafts want much lighter simpler couplings, those described are great for vast angles, but overkill for helo purposes. The Goodrich Ti couplings on the 92 have no moving parts at all, the misalignment is taken up in flexure of a Titanium welded coupling, shown in pairs at the end of the shaft sections below. They are hollow welded flanges that flex as needed:

http://www.fus.goodrich.com/images/S92Coupling.jpg

Dave_Jackson
26th Feb 2008, 18:49
rotorque.

Doman had a similar idea.
Web page (http://www.aviastar.org/helicopters_eng/doman_lz-1a.php)
Sketch of rotorhead (http://www.unicopter.com/Doman_LZ-1A.html)


As mentioned by Mart, a few years ago while developing this Hub - 3-blade - Constant Velocity Joint w/ Hub Spring (http://www.unicopter.com/S_Hub_CVJ.html) the gentleman at Thompson and I discussed CVJ's.

If interested, the following is the fundamental difference between our two ideas.

The Thompson coupling converts the double universal joint into a Constant Velocity Joint by the use of linkages to maintain the correct position of the simple housing. The Thompson coupling may be an ideal solution for conventional CVJ applications, since it offers no resistance to angular change and can except axial loading. However, for a helicopter rotor, a spring resistance to changes in the axial angle between the driver shaft (mast) and the driven shaft (rotorhub) is desired, since it acts as a hub spring (http://www.synchrolite.com/B329.html#Hub_Spring), in addition to maintaining position.


Dave

ShyTorque
26th Feb 2008, 22:27
Looks good to me!

I learned about the limitations of the basic "Hardey-Spicer" type of UJ when I owned a Triumph Spitfire Mk3 in the 1970s. This type of car has a UJ each side of the fixed differential housing to solid wheel driveshafts on a single transverse spring "swing axle" rear suspension. Each driveshaft runs to a fixed "right angled" wheel bearing.

They should have designed in a second UJ at the wheel end (GT6's did have a second one). The vibrations caused by the driveshafts accelerating and decelerating as they rotated were noticeable right through the car as soon as slight wear took place in the needle roller bearings in the UJs. I kept the car three years and changed so many UJs I could do it in 45 minutes even though it involved stripping the entire rear suspension.

Graviman
28th Feb 2008, 08:43
Nick, do you have a handle on the torque, rpm and mass of those S-92 drivelines? As the shafts are tubes, i assume mass is ~linearly proportional to length and torque. I can't really contact Goodrich for tech specs on the pretext of being a potential customer!

http://www.fus.goodrich.com/aircraftapplications/Helicopters.html
http://www.fus.goodrich.com/images/Helicopter.jpg

NickLappos
28th Feb 2008, 13:09
Grav, tell them it's for a 100 ton truck!

The engine input shafts to the transmission, and all the TR shafts use the Lucas (now Goodrich) couplings, so you can get up to about 3000 HP at 18,966 rpm. They are a marvel of new material science. Each is made of basically two Ti disks with flanged lips that are welded together. The drumhead of the disk flexes to take up the angular deflection needs, and the deformation is within the infinite life regime for the material. To prove that, each one is spun for 10e8 cycles at max deflection prior to delivery, proving them fit within that boundary. Where more than about 1.5 degrees are needed, two are used in series.

They are very light weight, basically just .050" Ti disks about 7 inches in diameter. The Apache also uses them. No parts, unlike Thomas couplings, which have many parts and many layers, and that slide against each other (fretting and cracking are the issue there.) the Thompson couplings are very nice design, but too many parts and complexity for simple shafts, IMHO. Could they replace the intermediate and tail GB?

Graviman
28th Feb 2008, 18:22
Hehehe, the problem here is the cost-mass-performance triangle, Nick. If I had my way our customers would be hovering that payload direct from mine to mill!

I had wondered about the use in the intermediate gearbox, if only as a means of saving weight. The link below suggests the Thompson Coupling could handle 20 degrees, but unlikely at high torque (Dana would certainly not approve a UJ for 15 degrees). Since I imagine the tail boom / fin corner is positioned to allow main rotor clearance then that angle is not enough. However, the standard joint is designed for grease lubrication inside rubber bellows – if the joint was oil immersed it would handle higher angles without heat build up. Since the joints are constantly in mesh then you would not be carrying around unmeshed teeth on gears, and the fixed angle removes the need for the complicated mechanism.

http://cvcoupling.com/index.php?option=com_content&task=view&id=12&Itemid=44

For helicopter applications Thompson would need to do a dedicated design study for an oil housed assembly with several couplings in series for say 45 degrees. This might result in half the mass and cost, over a gearbox, since there is less machining and heat treatment. Certainly there would be increased damage tolerance, from secondary particulate contamination such as tooth spalling & failure from say a bearing breakup. I would be happy to make some enquiries on Sikorsky’s behalf…


Dave, for the purposes of driveshaft mass guestimation:

3000HP (2237kW) / 18966rpm (1986.1 rad/sec) = 1126 Nm

4.506 g/cm^3 (Ti) / 1000 x 2 PI x ~7”/ 4 (50mm) x 0.050” (1.27 mm) = ~1.8 kg/m

Mass per torque per length = 1.6 kg/m per 1000 Nm input torque.

I’ve taken some liberties here for a first estimate, but maybe i'm wrong to criticise some of your designs...

Dave_Jackson
28th Feb 2008, 18:50
Mart,

Which driveshaft or design are you refering to?

Dave

Graviman
29th Feb 2008, 03:39
Dave, i have always criticised your interleaver designs as being potentially heavy because of the drivelines. The above guestimate, albeit making some wild assumptions about the construction of the drivelines (based on Goodrich Ti coupling photograph), indicates that the drivelines could be a reasonable mass. However, the concerns about operating costs of such a complex drivetrain are still valid. Equally, so are the concerns about machine rotor system width limiting general purpose landing site access.

160thfan
29th Feb 2008, 04:04
Hey Nick,

What is the life of the joint you are describing on the S-92?
Thanks

Dave_Jackson
29th Feb 2008, 07:23
Mart,

The cross-shaft on the V-22 was changed to carbon composite a year or two ago. This should make your calculations look even better.


What Interleaving? This week's project is the 2x4 PropRotor (http://www.unicopter.com/AeroVantage.html). ;)


Dave

rotorque
29th Feb 2008, 11:59
Didn't Bell's K-Flex design (OH58/206) work on a similar concept to the S92 coupling, no fretting or rubbing...

I could be wrong, but I was under the impression that each K-Flex flexure has an infinate working life as well. They are/were a great improvement on the older grease filled flex couplings.

Are the above couplings a true constant velocity or do they also suffer a slight change in speed per revolution as do UJ's?

:)

Graviman
29th Feb 2008, 22:36
160thfan ;)

The drumhead of the disk flexes to take up the angular deflection needs, and the deformation is within the infinite life regime for the material. To prove that, each one is spun for 10e8 cycles at max deflection prior to delivery, proving them fit within that boundary.


Rotorque, the Goodrich joints will be CV, but the 1.5 degrees is more to take up build tolerance (unavoidable on any structure), and airframe flexure from inflight loads.

Dave_Jackson
1st Mar 2008, 02:17
rotorque,
You are correct about the KAflex coupling. It has been in use since 1878. Ten years ago they were referring to it as 'Extended Service Life'

The subject of what constitutes a 'Constant Velocity Joint' is an interesting one. IMHO, it appears that all CVJs have one or more components that experience acceleration/deceleration. This component(s) may slide or roll and need lubrication, or it may flex and have a life limit.

Perhaps the claim to being a CVJ has to do with how large the oscillating component(s) is. Even a pair of universal joints provide a constant velocity, if these two universal joints are installed in phase with each other and always operated with equal angles.

Dave

Graviman
1st Mar 2008, 11:46
Dave,

For gound vehicle applications the driveshaft very much defines the "performance envelope" for a given terrain. I normally recommend that for good off-highway performance you should be aiming to isolate the machine from inputs above 1Hz. This frequency is a comfortable walking pace, so the spine nicely cushions the brain from reasonable accelerations. A driver will then try to control his speed so that he remains in control of his vehicle, which is defined as being in good ground contact, so 1g +/-1g. We design suspension travel for 0g to 3g to give a little margin for misjudgment.

Large vehicles have the benefit of compliant tyres, so we can achieve the desired 1Hz for less supension movement. The nominal spring compliance is chosen to be above tyre compliance, so that we can use asymetric dampers. The asymetric damper has a much higher rate in rebound over bound, so that in bound the sping compresses along with the tyre and in rebound the damper can thus absorb enough potential energy to avoid overshoot. For my models i assuming a spring law of Force=Const*Dist^Exponent, for either air or rubber installation. The automotive guys use a package called ADAMS which pushes the ride&handling performance to a different level.

The practical upshot of all this is that for ground vehicles every effort is made to limit the travel that the drivelines have to endure. I can tell you that for anything i look at, i aim to keep the nominal angles between 1 - 3 degrees from unladen to laden, and i aim to keep relative UJ angles within 1 degree. Anything beyond that and the universal points do not last our powertrain design intent of 15'000 hours, since the accelerations cause huge dynamic loads to affect bearing B10 life. I doubt very much that many drivers realise how much thought has gone into such a seemingly simple component.

When i worked at Land Rover (before BMW sold it to Ford) i recognised that the Thompson coupling had the potential to overturn many of the old rules. For helicopters the suspension from the air turbulence is in the blades themselves. I think Nick is quite right in that the applications are more limited, but there are potentials for advancement - like intermediate gearboxes.

If i get any feedback from Thompson i will feed it back to this thread for all to benefit...

Dave_Jackson
1st Mar 2008, 21:04
This Web page is patent searches on Elastomeric CVJs for use in rotor hubs. (http://www.unicopter.com/1334.html)

Nazalas
10th Feb 2009, 15:24
1) the “Thompson Coupling” is essentially an assembly of two universal joints with a linkage that ensures that the angles between the universal joints and the input shaft and output shaft, respectively, are the same

2) the knowledge that this arrangement offers a constant velocity performance is at least 100 years old and is nothing new or novel, a vast variety of linkages have since been developed or patented with exactly the same objective in mind, none of them has ever been considered as a commercially viable product in any industry

3) in order to be competitive with existing technologies in a mass production environment such as the automotive industry, a new product will be assessed in accordance with two selection criteria, “Power Density” and “Cost”; the “Thompson Coupling” or any other “two universal joint linkage assembly” inherently cannot compete in these two selection criteria, by a large margin

4) whilst “inventions” like the “Thompson Coupling” provide good entertainment material for the ABC, it is inexcusable that so called “Professional Institutions” like “Engineers Australia” and others, do not, or more likely, are not capable of researching the background and technical potential of a new technology, and consequently are culpable of misleading the public; the “Thompson Coupling” ranks in the same category as the “Sarich Engine”, neither of them were either novel, at the time of their patent lodgment, or have the potential to ever succeed as a mass product; “Engineers Australia” has promoted both of them

5) there is a fine line between “ignorant inventor enthusiasm” and plain fraud

anyone still interested in investing in the “Thompson Coupling” here is an opportunity

thompson coupling ltd shares (http://www.invested.com.au/79/thompson-coupling-ltd-shares-34713/)

thompson coupling ltd shares (http://www.invested.com.au/79/thompson-coupling-ltd-shares-34713/)

Graviman
12th Feb 2009, 19:17
For off highway equipment, like mining trucks, the ability of a joint to handle large angles without undue wear is a worthy cause. Car CV technology applied to large mining trucks just doesn't go the minimum 15'000 hours between overhauls - that's 900'000 miles or 1'000'000 miles if your round it up...

But this is a helicopter forum, and i think Dave Jackson's variation on the technology to reduce main rotorshaft vibration in teetering machines has real potential. Horses for courses. ;)

What is your connection with Thompson BTW? :confused: