Thread: Borg Warner viscous, Care & Feeding

OK, after looking at dozens if not hundreds of sites her eis what I feel is the most authoritative description as it is by the makers.
http://www.gkndriveline.com/drivelin...scous-engl.pdf

It is interesting in that the "hump" mode is described as being from the increase in volume of the fluid contributing to "hump" mode.
I am not looking any more.
Regards Philip A

This is the best explanation I can find:

PRINCIPLES OF OPERATION

Torque transmission in a viscous coupling is based on transmission of shearing forces in fluids. If two surfaces, such as the opposing surfaces of an inner disk and an outer disk, move relative to one another in a fluid, a shear stress is produced in the fluid filling the gap. In a viscous coupling, this relative motion of the surfaces is achieved through a difference in the speed of rotation of the inner and outer disks.

The inner and outer plates are arranged alternately in a fluid-filled housing so that the outer plates are connected to the housing and the inner plates to a hub. The gap between the disks is filled with a high-visosity fluid that transmits the torque without the disks being in contact with one another. The working space of the coupling is sealed off with special seals, ensuring that one filling of silicone fluid will suffice for the lifetime of the coupling.

The silicone fluids used for viscous couplings are clear and nontoxic and usually have a nominal viscosity between 5000 and 300,000 centistokes. The length of the molecule chains determines the flow properties. The longer the molecule, the greater the viscosity of the fluid. According to Newton's law, shear stress in ideal fluids is proportional to the relative speed difference. Silicone fluid is non-Newtonian and has, depending on viscosity, a more or less degressive characteristic. The higher the nominal viscosity of the fluid, the greater the torque that can be transmitted. It is relatively easy to produce silicone fluids with many different and very high nominal viscosities. Thus, the performance characteristics of the vicous coupling can be easily tuned to the vehicle in which it is to be used. The width of the gap between the inner and outer plates also affects the torque transmission curve. The smaller the gap, the greater the velocity gradient and, consequently, the greater the transmitted torque at a given speed difference.

The viscous coupling normally operates in the viscous mode, where torque is generated by viscous shear as described earlier. However, prolonged slipping under severe starting conditions causes the inside of the coupling to heat up. The relatively large coefficient of thermal expansion of the silicone fluid causes the fluid inside the viscous coupling to expand considerably as the temperature increases. After some seconds it is expanded to such an extent that it fills all the available space inside the coupling, and the pressure increases rapidly. This forces the disks together, and metal-to-metal friction occurs. The result is a substantial increase in torque transmission, known as the "hump" effect or self-induced torque amplification. For a given speed difference, the point at which the hump begins can be precisely determined by the design and setup of the coupling. The purpose of the hump is to protect the coupling against overheating and for higher torque transmission under extreme conditions, providing up to 100 percent lock-up, even when one wheel is on a very low-friction surface, such as ice, when attempting to free a stuck vehicle. Another characteristic feature of the hump is that it disappears after a few seconds, whereupon normal operating conditions are restored.

Viscous couplings enter mainstream vehicles: external mounting techniques and reductions in cost and weight make it feasible to design these couplings into existing vehicles with minimal modifications. | Business solutions from AllBusiness.com

Also:

A Comprehensive Study of Self-Induced Torque Amplification in Rotary Viscous Couplings
J. Tribol. -- January 2003 -- Volume 125, Issue 1, 110 (11 pages)
doi:10.1115/1.1504087

You are not logged into the ASME Digital Library.
Log in

ABSTRACT
REFERENCES (18)
Author(s):
Sankar K. Mohan
Advanced Engineering, New Venture Gear Inc., 6600 NVG Drive, East Syracuse, NY 13057

Bandaru V. Ramarao
State University of New York, Environmental Science and Forestry, 1 Forestry Drive, Syracuse, NY 13210
Rotary viscous couplings with interleaved, perforated plates and viscous fluids are used in automotive systems to transmit torque. During operation, viscous dissipation raises fluid temperature, lowers fluid viscosity and causes the torque transmitted to drop monotonically to unusable levels. Couplings designed with certain plate geometry exhibit a reversal of the torque trend with temperature, and transmit increasingly high torque even under continuous operation. Such couplings achieve torque amplification factors in excess of twenty, compared to earlier couplings. This torque amplification phenomenon has been utilized by industry without fully understanding the mechanisms involved. A comprehensive theory is proposed to explain the complex sequence of events that results in this "anomalous," but useful phenomenon. Mathematical models are developed for each interdependent process. A visual simulation tool is used to model the intricate dynamics inside the coupling. Results from the simulation model are compared with experimental findings. The various thermodynamic, hydrodynamic, structural and mechanical processes are delineated and tested with a combination of theoretical analysis, computational simulation and experimental observations. The proposed theory identifies, defines and explains the conditions necessary for initiating and sustaining the self-induced torque amplification. The hypotheses are validated by the reasonable agreement of the model with the test results.

©2003 ASME

So - it seems that those who say that the fluid increases in viscosity with temperature are incorrect. The fluid viscosity decreases with temperature in the same way as other fluids (it seems).

However viscosity increases with shear. But the above quotes seem to suggest that VCs work best when cold, and deteriorate with temperature, until they get hot enough that the pressure inside the VC is sufficient for them to hit the "hump" and lock.

To you guys that responded with info, thank you all, I now have been able to see and read about something that I have pondered for a long time, (I am not very computer savy and probably would not have thought to look up stuff about the VC, this thread just got me thinking again), even though throughout my working life I had dealt with these Viscous Couplings, I had only ever changed them out and sent them back to the OEM for recon or scrapping, industrially they are generally considered a low maintenance item.
The detailed descriptions of function and the physics were really great and a wealth of knowledge now I only hope that the unit we have fitted into the old LR County holds together ok, it is certainly more silent in operation than the original gearbox and transfer originally fitted behind the V8, might see about one for the Disco.

Very interesting thread

Back a few posts to how the VC and the transfer box work. I'll quote from the P38 w/shop manual:

"Drive from the intermediate shaft is transferred by a morse chain to the differential unit. The differential unit comprises sun and planet gears. The rear output passes through the differential unit sun gear shaft and engages with the planet carrier. The splined forward end of the rear output shaft provides location for the viscous coupling unit inner spline. The outer diameter of the sun gear shaft engages with the outer splines of the viscous coupling unit.
Viscous coupling unit
The viscous coupling operates in conjunction with the differential unit to control the proportion of drive torque transferred to the front and rear drive shafts. The viscous coupling is a sealed unit filled with a silicon jelly which surrounds discs within the unit. The silicon jelly has properties which increase its viscosity and resistance to flow when agitated and heated.
During normal driving conditions, slight variations in the relative speed of each drive shaft is insufficient to increase the viscosity of the silicon jelly. Therefore the resistance within the viscous coupling is low.
In off-road conditions, when the wheels lose grip on loose or muddy surfaces, a greater difference in the rotational speeds of the front and rear drive shafts exists. The slippage, due to the difference in rotational speeds of the drive shafts, within the viscous coupling agitates the silicon jelly causing heat which increases the viscosity. The increased viscosity increases the drag between the discs forcing both sets of discs to rotate at similar speeds, reducing axle slippage and increasing traction. The viscous coupling removes the need for a manually controlled differential lock.

What is happening is the mechanical differential which is "open" is joined to the VC which is acting as a limited slip unit with lock up. In normal operation the viscosity in the unit is sufficient to distribute torque 50/50 to the front and rear drive shafts (this is why Ian says it's pre-locked when he bench tests it)

This why we have "constant 4WD"

The clever bit is when the front or rear axle spin at different speeds it produces a rapid change in rotational speed between the plates, the VC will behave as follows (and we're talking milliseconds of reaction time):
1) Shear
2) Heat rapidly
3) Viscosity increases/Silicon expands
4) VC begins to lock....
5) "Hump effect" locks VC

However there is a limit: and if the loads are too great for too long a period the viscous fluid will indeed "cook" and the VC will fail - think of it as a fuse, and why you shouldn't tow behind a lift truck!

That vid of the VW transporter demonstates this very well, and also demonstrates why bench testing will only show up a cooked VC, not demonstrate how it works

So, you're all correct - sort of!

Chris

... Finally learning something !

Which now brings us to the most important bit, how to look after the coupling so that it does'nt get Viscous and bite our wallets again...

The obvious is tyre size. - No more buying new tyres and sticking them on the front. (as mine was, with 50% worn on the rears)

Would this explain the rather odd tyre pressures as found on the Tyre Plaque, - 28 psi Front and 38 Rears, loaded or unloaded. ?
- When I bought mine the pressures were...you guessed it, same all round.

Is there a definitive reason for it, - and which approach annoys the VC ?

I suppose the only way to monitor the aggravation in the VC is to have a read-out of the relative RPM's of front/rear prop shafts. Or better still, something found in twin - engined boats, a simple centre-zero meter which displays the faster/slower engine.... But thats overkill !

Originally Posted by Hobbes

The silicon jelly has properties which increase its viscosity ... when ... heated.


1) Shear
2) Heat rapidly
3) Viscosity increases/Silicon expands
4) VC begins to lock....
5) "Hump effect" locks VC

From the rheology paper it would seem that this statement is completely incorrect.
Which holds with physics - liquids decrease in viscosity when heated, gases increase.

So your list would more correctly be:
Shear increases viscosity of the silicone, however this is countered by a decrease in viscosity with heat, however with enough heat then you get enough of a pressure increase inside the VC to induce the hump.

I don't think the VC is going to much troubled by minor differences in tyre diameter. It's significant differences in rotational speed that create heat in the unit. In my North American Spec service bulletin for the BW (from I think 96) it suggests that "due to the intense heat generated by the sealed VC and absorbed by the ATF, regular fluid changes are important particularly if the vehicle is used off-road on a regular basis"
It suggest 48,000km in regular service; 24,000km for frequent off road use.

(Similarly as a comparison on our Audi Quattro you're not supposed to exceed 50km on the space saver as it's (significantly) different diameter can overheat the Haldex Viscous)

However, back on tyres my RRC90 manual gives 28F/35R & 41R for sustained speedsin excess of 160km/h or with heavy rear axle loads.

Just to clarify in my previous post:

In normal operation the viscosity in the unit is sufficient to distribute torque 50/50 to the front and rear drive shafts

that's not accurate as it's the open diff that gives the 50/50 split, until there's a difference in shaft speed at which point the the VC is will very quickly react to redistribute torque as required - in motorsport terms it's tuned "stiff" ie LSD > Lock quickly

Rangies prior to the BW transfer had the same tyre pressure spec; these are presumably related to handling characteristics rather than VC longevity. Interesting point about the spacesaver tyres and speed limitation- I had assumed that this also was about handling rather than drive-train care.

Had another fiddle with the BW transfer I have stripped in the shed and I am pretty well convinced that my theory of acheiving 4wd with open centre-differential by cutting down the output shaft in the setting of a seized VC, would work. Apologies Ian.
James, glad to hear you have temporarily circumvented your problem. I don't think your fuel economy will improve, as you are still rotating all the driveline components. Perhaps if you fitted free-wheeling hubs- or removed the CVs in the front hubs? Don't consider this if you have ABS.

A little over 30 years ago when I was training as a mechanic, I was taught that the (then new-fangled) viscous fan hubs contained a liquid that increased in viscosity with temperature. Having dismantled a failed one, I'm not so sure.