Rear Differential Rebuild/With Pictures

[Wh33lhop]

a Clutch type LSD unit locks because the basket of spider gears in the middle wants to expand when One wheel is significantly different than the other

Always seemed to me that it expanded when torque was applied to the input shaft and the wheels resisted, not when there's a speed differential. Clutch type locking action is proportional to torque applied to the diff.

Don't know much about viscous diffs since I'd much prefer a clutch type in any situation.

[nando]

Chicken egg equation

Realistically the plates inside vlsd will expand with heat.

This is a math bee beteeen tmi school kids

Sent from my SAMSUNG-SGH-I337 using Tapatalk

the plates do not touch (or at least, they shouldn't). they have a micro impregnated surface that engages the silicone fluid. if the plates touch, your VC is toast; aside from metal/metal contact, the special surface will wear away quickly and the plates will warp.

the seal on an ix VC can only take ~300 degrees before it fails. if the locking action worked off heat, you wouldn't have a functional VC for very long.

[Wanganstyle]

the plates do not touch (or at least, they shouldn't). they have a micro impregnated surface that engages the silicone fluid. if the plates touch, your VC is toast; aside from metal/metal contact, the special surface will wear away quickly and the plates will warp.

the seal on an ix VC can only take ~300 degrees before it fails. if the locking action worked off heat, you wouldn't have a functional VC for very long.

Nobody said they touch.

Subaru marketing states that the vc fluid and steels will also expand to close down gaps and lock up.


Sent from my SAMSUNG-SGH-I337 using Tapatalk

[nando]

marketing vs engineering. hrm... ;)

[Wanganstyle]

marketing vs engineering. hrm... ;)

Chicken egg same origin

Nobody can deny the item gets hot inside; and engineering accounted for this heat in operation

The GKN m-variable incorporates a viscous coupler and a stack of clutch packs in the same unit.


Sent from my iPhone using Tapatalk

[The Dark Side of Will]

Chicken egg same origin

Nobody can deny the item gets hot inside; and engineering accounted for this heat in operation

The GKN m-variable incorporates a viscous coupler and a stack of clutch packs in the same unit.

Sent from my iPhone using Tapatalk

If it was not heat expanding the silicon viscous fluid then how else would it lock up? MAGIC???

A *VISCOUS* limited slip operates via *VISCOSITY*.

Put some Aunt Jemima out on your porch when it's 35 degrees (F) outside and then try to pour it. That will give you a feel for a high viscosity fluid. The silicone fluid inside a VC is MUCH more viscous than that.

The plates in the VC are very close together but should *NEVER* touch. As they rotate relative to each other, they squeeze the silicone fluid between them. The thickness/viscosity of the fluid means that squeezing it through the thin gaps in the disks creates resistance... it's just like a hydraulic brake (like a water brake engine dyno) or damper.

As the fluid is squeezed through the small gaps, it heats up. This is the same principle at work when you run more oil pressure in your engine higher than it needs to be and end up with high oil temps. Also the same reason stiffly valved shocks/struts heat up on a rally course.

The point is... it is squeezed FIRST, provides resistance FIRST, and heats up LATER after continued use. A VLSD is instantaneously responsive. That's why a cold VC can pass a jack test.

The fact that the fluid expands when it is heated is why a VC is not filled completely when sealed.

This is precisely why BMW chose the viscous mechanisms for the center and rear of the E30 325iX... they could operate in extremely low traction conditions without any detrimental effects. With the stock preload, a clutch type LSD will act like a spool on the snow, while a VC will still act like an open diff until one wheel starts to spin.

Have you seen this thread? http://www.e30tech.com/forum/showthread.php?t=118249

The GKN Visco-Lock is a different animal.

It uses a pump that moves silicone fluid to squeeze clutch packs.
In the Visco-Lock, the silicone fluid is "pumped" via a friction pump to apply clamping load to conventional clutch packs. Relative motion of the axle and the diff carrier operates the friction pump. It is *NOT* a viscous coupling
(Edit: Visco-Lok is also available for the Viper: http://www.unitrax.biz/?tag=gkn-visco-lok )

GM used a similar, but much weaker application of the same idea in their Versatrak AWD system in the early 2000's.

The GKN uses a gear train to transmit power, while the friction pumps and clutch packs just provide locking torque.

GM used the clutch packs to transmit drive torque but only on an as-needed basis... the vehicle was completely FWD until the front tires slipped. There was no center diff; the T-case output was geared directly to the front axle. This means that when the front tires slipped, the rear diff carrier turned faster than the rear axles. This operated a gerotor pump, with its inner ring attached to the axle and outer ring attached to the carrier, which applied hydraulic pressure to the clutch packs to transmit drive torque.

The GKN unit operates on the same principle, except that it uses a friction pump and silicone fluid to squeeze the clutch packs.

There *ARE* some marketing materials out there claiming that a viscous unit will "lock" when it gets hot. The Chevy Silverado SS T-case uses a viscous coupling and the marketing materials state that hard use will make it lock. They also say that *continued* hard use will make it FAIL. This is ABUSE.

Marketing materials that say such things are trying to dress a bug up to look like a feature.

[Wanganstyle]

Nowhere did I say anything actually touches. Tangent.

Have you ever broken down and examined how a gkn visco lok functions? I have;

Its a viscous unit on the top cap that sends force to activste the left side clutch stack and activates the right side with the spider gears transferring the force (no ramps). Weird item

Sent from my SAMSUNG-SGH-I337 using Tapatalk

[nando]

either way, the VC doesn't work off heat.

as was already stated, the increase in viscosity with shear is a basic material property of the silicone fluid.

[The Dark Side of Will]

Nowhere did I say anything actually touches. Tangent.

Chicken egg equation

Realistically the plates inside vlsd will expand with heat.

This is a math bee beteeen tmi school kids

Sent from my SAMSUNG-SGH-I337 using Tapatalk

Actually, I'm not sure what you said

OBTW, do you know if the 210mm 2.65 ratio gearset fits the same carrier as the 2.93 gearset?

Have you ever broken down and examined how a gkn visco lok functions? I have;

Its a viscous unit on the top cap that sends force to activste the left side clutch stack and activates the right side with the spider gears transferring the force (no ramps). Weird item

Sent from my SAMSUNG-SGH-I337 using Tapatalk

The viscous unit is just a pump that applies axial load to a pretty normal clutch pack, similar to what's in the standard LSD's. It's not a viscous coupling.
As I mentioned, it's similar in principle to something that GM built, but it's just really weird that they went with the viscous pump rather than a gerotor.

[z122643]

Where do you find the side cover bearing races? I am rebuilding the diff that came out of the car. It needs EVERYTHING!

[Wanganstyle]

Where do you find the side cover bearing races? I am rebuilding the diff that came out of the car. It needs EVERYTHING!

Buy them from the same place the bearings are coming from; the part number for the race can be brought to any timken bearing dealer.

Remove all old bearings/races and take to bearing supplier.




In regards to the bmw gkn m.variable diff;

I dont see any pump inside; the top cap has a sealed viscous unit inside that does not come apart with normal means; I've sent 20 tons of press through a scrap one and it does not budge.

The left stub axel of all e46 m3s
'S engage 6 clutch discs and 7 steels; the stub runs through The viscous top cap.

On the left also are traditional clutch disc friction type just like a e30 standard clutch disc. The axle is stubby short and engages the spider cluster only.

[The Dark Side of Will]

I dont see any pump inside; the top cap has a sealed viscous unit inside that does not come apart with normal means; I've sent 20 tons of press through a scrap one and it does not budge.

That's the pump... if the hub turns relative to the case, the piston should move out to squeeze the clutch packs. The viscosity of the silicone fluid means that it should take a bit of time for the fluid to flow back to the reservoir and thus take time for the assembly to compress again.

[z122643]

I was going to get the bearings from BMW. I assume there is a better way?

[Wanganstyle]

I was going to get the bearings from BMW. I assume there is a better way?

Better way is to drop it off at a rear end shop close to you that has actually built a bmw differential before;

stabbing in the dark and not knowing from experience what to do is really bad when it comes to engines/differentials.

[z122643]

Better way is to drop it off at a rear end shop close to you that has actually built a bmw differential before;

stabbing in the dark and not knowing from experience what to do is really bad when it comes to engines/differentials.

So where do you get your bearings from? I noticed your from the bay. I'm in the East Bay. (Sunol)

Cheers