Introduction

The Viscous Coupling (VC) is the drive train component that transmits power from the center drive shaft to the front differential, and on to the front wheels.  Inside the coupling there is a silicon liguid/goo that turns nearly into a solid when exposed to shear force caused when the VC input and output shafts rotate at different speeds.  The input shaft is connected to the rear wheels.  The output shaft is connected to the front wheels.  When the front and rear wheels turn at different RPMs (at a rate greater than 6%), the shear force raises the temperature and viscosity of the silicon liquid inside the VC.  The silicon becomes solid like and engages plates inside the VC with the result that power is transmitted from the center drive shaft (the engine) to the front differential (the front wheels).

When the front and rear wheels begin to rotate at relatively the same RPMs again, the liquid "deactivates," becomes less viscous and more liquid like, and as a result the front wheels/diff. disengage, and the van is powered again only or primarily by rear wheel drive.

VC Fail - Causes & Effects

A VC can fail in two basic ways: it can stop engaging altogether or it can remain constantly engaged (sometimes only at higher operating temps).  VC fluid loss caused by leaking seals seems to be the most likely cause of a total failure of the VC to engage at all.  When this happens, you never have 4 WD.   The more insidious and costly failure occurs when the VC remains engaged when it should not be engaged.  This can lead to destruction of the entire drive train, including the expensive transaxle.

There appear to be two primary causes for the VC to engage when it should not: (1) having tires that are not all the same size and wear (all 4 must be the same), and (2) age...VCs appear to have a nautural life span, at least where subjected to routine high operating temps.  (One shop also claims that having a drive train that is not properly aligned also causes premature engagement, but this theory seems questionable.)

When tires of different wear or size are used, it causes the VC to engage prematurely or even constantly.  It makes the VC "think" your wheels are slipping and that you need 4 WD.  Premature or constant engagement of the VC, particularly at highway speeds, overheats the VC, thereby "cooking" the viscous fluid.  Over time, this causes the properties of the fluid to change so that it engages the VC prematurely or even permanently,  thereby stressing the other components of the drivetrain.

A typical scenario leading to a cooked VC would involve a syncro with tire sizes that vary slightly in treaddepth.  The syncro is regularly driven at highway speeds for hours at a time over a period of time.  Eventually, the driver notices  binding in the drive train whenever she pulls off the highway into a gas station for gas.  It may be subtle at first.  Eventually, though, as the fluid gets cooked and ruined,  the wheels seem to stiffen or bind much easier than before, and ultimately at the slightest turn of the wheel.  When it gets bad, the drive train may lock up completely in the parking lot at very slow speed upon a relatively slight turn of the steering wheel.  This total engagement of the drive train puts tremendous strain on the components of the drive train when the van is moving at speed under power.  With continued use, the transaxle soon fails, the drive shaft and CV joints are also strained.

There is some dispute as to whether a properly functioning VC will cause binding in very tight turns, like when turning while backing out of a driveway, or doing a sharp turn in a parking lot.  The VC engages when the front and rear wheels turn at different speeds, greater than 6% in relative RPMs.  When the front and rear wheels turn at a greater difference in RPMs during sharp turns (above 6%), it would seem that the VC would engage and cause some binding.  However, a brand new VC will not do this, even when warm.

Chirping/binding in tight turns at least provides good cause to check your tire tread depth and to keep an eye on whether the symptoms become progressively more pronounced.  However, symptoms indicating that the VC is engaging sooner than it used to or should be, such as obvious binding at low (parking lot) speeds, easily induced binding (i.e., upon less turn of the steering wheel) and heat related binding (after long high speed summer trips), should not be ignored and should be investigated immediately in order to avoid huge repair bills.  Brand new OEM VCs can be had for about $1100.  From the dealer, they cost over $2,000.  If you cannot afford or find one right away, remove your center drive shaft until you do to avoid causing damage to the drive train.  Have a qualified mechanic do this, or follow the procedures in Bentley.  There are a few not apparent procedures that should be followed.

As the chart below depicts, at 110 degrees centigrade, the VC silicon fluid viscosity increases dramatically, thus engaging the VC.  Since temperature is a cause and not just an effect of VC engagement, it  seems probable that the "engagement temperature" of a "cooked" VC (a VC where the fluid has been ruined by prolonged overheating) is lower than for a non-cooked VC, and that with cooked VC fluid, the VC engages at the higher end of normal VC operating temperatures, with the result being that the VC is always engaged at the higher normal operating temperatures even if the wheels are turning within the 6% threshold for engagement.  This puts incredible strain on the entire drive train.

VW's original (German) repair manual doesn't say much about how to test the viscous coupling.  They only recommend placing the rear wheels in a break testing stand.  If you then switch to  the G-gear (creeping gear), the front  wheels should move the van  out of the  test stand as soon as the engine is revving slightly above idle.   If the front wheels fail to do so the viscous coupling is to be replaced, VW says.  VW adds another tiny sentence  to this, saying that only  when the engine  is revving at idle and with the G-gear  switched in, the viscous coupling is able  to absorb all the torque to the front wheels and keep them from moving.

Using a heavy floor jack with wheels, put a block of wood on the jack and raise the rear of the van using the skid plate.  Be careful, you can bend it.  The wood block, if long enough, distributes the weight across enough of the skid plate to minimize that risk.   Lift both back wheels off the ground 6" or so on a smooth level parking lot, with the jack's wheels parallel to the Syncro's.  The rear of the van will be moving on the jack wheels, so you need to make sure there are no obstructions that could catch the jack wheels and cause the van to fall off the jack.

Now, put a 1x1 piece of wood in front of each front tire, 2x4 may work too.  You need to block the front wheels like this to be able to test whether the VC is capable of absorbing the spinning of the rear wheels without locking up and causing the front wheels to engage and climb over the wood.  If you can get the rear wheels to turn/spin in the air with the clutch fully disengaged, and without the van climbing over the wood blocking the front wheels (can't be too high...1-2"), then the VC is definitely good, or the fluid is not cooked.  The van should climb over the blocks as soon as you increase the engine RPMs.

It may take several tries to get the van to do this.  The VC is very sensitive, and it will want to engage as you let the clutch out.  Some advise using the hand brake to help slow the spin of the rear wheels ... or to start them spinning slowly at first.  It is really neat when you get it to work.  Suddenly, the VC is working before your eyes in a very graphic way!

________________________________________________________________

Warning! You could conceivably get killed or kill someone performing this test.  If you die doing this, don't get mad at me!
_________________________________________________________________

A test to see if the VC is engaging is to take your syncro out into some fresh snow or mud and induce wheel spin.  If you get none up front, but the rears are spinning, you have a problem...which could be blown VC seals or perhaps a bad front differential.  This test, however, does not help you diagnose a VC that is in the process of self-destructing, or is putting undue strain on the rest of the drive train, because it is engaging when it should not.

For more information on the VC by Derek Drew and Jim Davis, see   vanagon.com/syncros/technica/