Very loose rear end (Handling)

[Gforce]

Originally Posted by Tcoat

Thank you! Glad to hear there are others that understand how they actually work.
This was the whole statement from the beginning. The original question was will it cause harm with "extended" use. Not will it cause harm.

Now there are two posters who seem not to understand how different a Torsen is.

Clutch type LSD react only to wheel speed, that's how they engage. Torsen do not react to wheelspeed, ever, not in any way.

Torsen do not wear when biasing torque although it is tempting to think that they must due to the internal friction. Friction reduces wear if of it is static friction, which it generally is inside a Torsen. Only sliding friction causes wear and there is very little of that when a Torsen is biasing torque and not very much when it is differentiating.

But, I concede it is difficult to understand how this ingenious device actually works. Thinking it gives progressive limited slip is the biggest misconception and affects how you drive these cars. You either know, or you don't.

[Gforce]

Originally Posted by ZDan

The same is true of most limited slips, including clutch-types.

Wheelspeed doesn't dictate the amount of "lock" or torque bias, but it's not totally unimportant.
(FTR "wheelspeed" doesn't dictate how much a clutch-type locks or biases torque, either.)

In both the clutch type and the torsen limited slips, torque is transferred from one output to the other via friction. In the clutch-type the friction is between clutch plates, in the Torsen T2 it's between the helical gears and the casing resulting from axial force in these gears.

If you have different rear tire sizes, rear outputs will be rotating at different speeds and this friction will generate heat and wear parts at a greatly increased rate.

Running a smaller-diameter space-saver spare on the drive end of the car with a either type is not a good idea. Tremendously greater heat and wear due to the different rotational speeds of the outputs. They're only intended/designed to operate like that briefly while cornering, not continuously.

A Salisbury clutch-type with low breakaway torque doesn't behave terribly differently from a Torsen. Can be smoother, even...

There is no waiting with a Salisbury/ramp clutch-type either. In both cases the application of torque increases "lock" or bias, so they are both "progressive" in this sense. And in both cases friction is the mechanism for transferring torque from the wheel that wants to spin to the wheel with grip.

Torsens can and do WEAR. And driving on one for a significant amount of time/distance with different diameter tires will accelerate this wear. But I don't think 50 miles normal driving would do anything catastrophic...

I do not think you are correct. I am sure Torsen would disagree with just about everything in your post.

I also point out that the space saver spare, of any size, would be completely worn out and down to the belts long before the slightest wear could have occurred inside the Torsen.

Clutch type LSD use a centrifugally activated clamping system of clutch plates, which is why electronically controlling these is preferred. Edit, the clamping system relies on the wedging action of the spider gears or of the spider drive pin in a shaped slot. The effect is as if centrifugal force pushes the side gears apart. Wave type use interlocking circular ramps to achieve the same effect, again as if there is centrifugal force pushing the side gears against the clutch plates. These wave type are speed sensitive in that there must be a faster moving axle to activate the clutch plates.

These LSD are only wheelspeed sensitive and do not engage unless a difference in wheelspeed is experienced. They react to that. They do not limit slip until a preset wheelspeed difference is achieved. Then they lock pretty firmly and any differentiation results in wear and heat. Edit. Clutch type spin up one wheel as they lock. Torsen never do.

Clutch type LSD "bias torque" by braking the spinning wheel against the wheel with grip. This is why electronic systems for limiting slip operate just like a clutch type LSD until both drive wheels lose grip. Torsen do not do this and cannot do this.

Most of the heat inside a open differential comes from the gear speed, nothing to do with limiting slip. Clutch type LSD heat up because they must allow the clutches to slip to permit differentiation. Torsen do not.

Torsen bias torque immediately and are not delayed by waiting for the lower grip wheel to turn more quickly. They react to the torque that the wheel with less grip can actually handle. Clutch type only react to wheelspin. Like haldex, they limit slip but only when there's slip to limit. Torsen do not limit slip, and cannot. They transfer torque away from the tire that might slip but only do so until it does slip. This is the complete opposite of the clutch types.

And Torsen do not wear when they are biasing torque. On the contrary, they simply differentiate as would an open differential while transferring torque. Friction does not necessarily cause wear. Only sliding friction can cause wear and not much sliding goes on inside a Torsen when it is biasing torque.

[ZDan]

Originally Posted by Gforce

Now there are two posters who seem not to understand how different a Torsen is.

Well, there's at least one!

Clutch type LSD react only to wheel speed, that's how they engage.

Viscous and Gerotor types react to wheelspeed, but not the clutch type.

Clutch type will have some breakaway torque, so it's already locked to some degree before anything even happens. The clutch type then locks up more with applied torque. Even a "plain" clutch-type without ramps does this due to the spider gears being forced apart harder with increased torque at the carrier, which increases clamping force on the clutch pack. A Salisbury clutch type will do this to a greater degree and is tunable with ramp angles.

So without any wheel motion (in unison or differential), the clutch-type already has some lock, and with torque applied it has even more lock.

Clutch type does not rely on wheelspeed to apply bias or lock!

Torsen do not wear when biasing torque although it is tempting to think that they must due to the internal friction. Friction reduces wear if of it is static friction, which it generally is inside a Torsen. Only sliding friction causes wear and there is very little of that when a Torsen is biasing torque and not very much when it is differentiating.

You seem to be saying that the Torsen is fully locked (up to bias ratio limit anyway) when biasing and fully open when differentiating.
However I think the whole idea of the "TORque SENsing" Torsen is that it acts open when no torque is applied, and applies increasing lock or bias with greater applied torque. So when you're accelerating while cornering the diff can be both differentiating and biasing. So there is sliding friction while under load. Which generates heat and causes wear. Granted this wear may be slow enough that the diff doesn't wear out within the design lifetime of the vehicle.

Anyway, that's been my understanding of it...

But, I concede it is difficult to understand how this ingenious device actually works. Thinking it gives progressive limited slip is the biggest misconception and affects how you drive these cars. You either know, or you don't.

Again, I thought progressively more lock or bias with applied torque was the whole idea...

Originally Posted by Gforce

I also point out that the space saver spare, of any size, would be completely worn out and down to the belts long before the slightest wear could have occurred inside the Torsen.

Disagree. There's gonna be heat and wear with the thing constantly differentiating while drive torque is applied while driving great distances in a straight line.

Clutch type LSD use a centrifugally activated clamping system of clutch plates, which is why electronically controlling these is preferred.

That is NOT how a clutch type limited slip works at all! See above ^^^
There will generally be some breakaway torque which gives some lock or bias right off the bat. Then you get increased clamping with increased *applied torque*. Wheelspeed has nothing to do with it!

These LSD are only wheelspeed sensitive and do not engage unless a difference in wheelspeed is experienced. They react to that. They do not limit slip until a preset wheelspeed difference is achieved. Then they lock pretty firmly and any differentiation results in wear and heat.

Good God, totally wrong!

AGain, viscous and gerotor types *are* activated by differential wheelspeed, but definitely NOT the clutch type, whether "plain" or Salisbury.

Most of the heat inside a open differential comes from the gear speed, nothing to do with limiting slip. Clutch type LSD heat up because they must allow the clutches to slip to permit differentiation. Torsen do not.

IF you run a very low breakaway torque, clutch type lsd clutches can last a VERY long time. I ran one in my 240Z with only ~23 lb-ft breakaway for probably 80k street miles and 80 or so track days and it worked great the whole time, never had to replace clutch plates.

Clutch type only react to wheelspin.

Not true.

And Torsen do not wear when they are biasing torque. On the contrary, they simply differentiate as would an open differential while transferring torque. Friction does not necessarily cause wear. Only sliding friction can cause wear and not much sliding goes on inside a Torsen when it is biasing torque.

Here's where the friction happens, guaranteed that wear is also happening there.

Again, it may last the life of the vehicle but wear and heat do occur under load when there are different output speeds.,

[Gforce]

Got it.

Torsen never locks up but it also biases torque immediately. The torque bias line is straight. Clutch type torque transfer line has two slopes, shallow until lock up and steep after lockup.

We have the same understanding of the differences, at least I think we do.

We differ about the friction and heat aspects.

Torsen themselves, and Quaife, and Eaton all claim no internal wear occurs when their diffs are operating. The key is to appreciate just how little gear tooth sliding is actually going on and anynthat us occurs at low face loads and low speeds. The hypoid bevel pinion gear is making much more heat, has very high tooth loads and a lot of continuous gear face sliding. There's nothing going on in the differential gears of a Torsen that can come close to the heat loading delivered by the pinion engaging and driving the crown gear.

We may have to just agree to disagree on this point about different tire sizes presenting an unusual wear risk to the Torsen type diff. If the manufacturers of the devices themselves do not warn against this that is conclusive enough for me. As I have previously said, there is nothing going on inside a Torsen that can generate significant additional heat from either differentiating or torque biasing. That is not true of clutch type which generate a lot of heat until they actually lock, at ehich time they can no longer differentiate. Viscous couplings, which are speed sensitive as you say, can quickly be damaged by forced differentiation. However, generally speaking other components first bear the brunt of a heated up viscous coupling because they can cease differentiating, one of the main drawbacks. Eventually the heat buckles the drive plates inside the coupling.

[Tcoat]

Originally Posted by Gforce

Got it.

Torsen never locks up but it also biases torque immediately. The torque bias line is straight. Clutch type torque transfer line has two slopes, shallow until lock up and steep after lockup.

We have the same understanding of the differences, at least I think we do.

We differ about the friction and heat aspects.

Torsen themselves, and Quaife, and Eaton all claim no internal wear occurs when their diffs are operating. The key is to appreciate just how little gear tooth sliding is actually going on and anynthat us occurs at low face loads and low speeds. The hypoid bevel pinion gear is making much more heat, has very high tooth loads and a lot of continuous gear face sliding. There's nothing going on in the differential gears of a Torsen that can come close to the heat loading delivered by the pinion engaging and driving the crown gear.

We may have to just agree to disagree on this point about different tire sizes presenting an unusual wear risk to the Torsen type diff. If the manufacturers of the devices themselves do not warn against this that is conclusive enough for me. As I have previously said, there is nothing going on inside a Torsen that can generate significant additional heat from either differentiating or torque biasing. That is not true of clutch type which generate a lot of heat until they actually lock, at ehich time they can no longer differentiate. Viscous couplings, which are speed sensitive as you say, can quickly be damaged by forced differentiation. However, generally speaking other components first bear the brunt of a heated up viscous coupling because they can cease differentiating, one of the main drawbacks. Eventually the heat buckles the drive plates inside the coupling.

Well Uber, I posted 3 documented examples of cars using Torsion diffs that do indeed say it could cause damage and/or operate improperly and you summerly dismissed them. You can agree to disagree all you wish there are documented statements from car manufacturers not to do it.

[Gforce]

Originally Posted by Tcoat

Well Uber, I posted 3 documented examples of cars using Torsion diffs that do indeed say it could cause damage and/or operate improperly and you summerly dismissed them. You can agree to disagree all you wish there are documented statements from car manufacturers not to do it.

I don't think you did, none of the makers you mentioned fit Torsen.

Clearly I am referring to the makers of the diffs in this post.

[ZDan]

Originally Posted by Gforce

Torsen never locks up but it also biases torque immediately. The torque bias line is straight. Clutch type torque transfer line has two slopes, shallow until lock up and steep after lockup.

Clutch-type biases torque IMMEDIATELY also, and starts from a positive value instead of zero.

Torque bias of a clutch type starts at a positive value because of the breakaway torque. How much steeper it gets is dependent on tunable factors like ramp angle, active clutch plates, and in the case of the OS Giken I got for the FD, preload springs. I had mine specifically tuned to have minimal breakaway torque and a shallow slope to avoid rapidly increasing lockup with torque.

[Gforce]

I think it's worth one more try at making my point.

Torsen will not torque bias just because one wheel is smaller. It does not and cannot torque bias just because one wheel turns faster then the other. It just differentiates.

If traction differs left to right it doesn't matter what size wheels are fitted the Torsen will bias torque in exactly the same way as it would if the wheels were the same size. I mean this is just the physics of how they work.

Clutch type using preload springs always wear the clutch plates a little if only because they are preloaded. Non pre loaded disc type LSD shouldn't care if the rear wheel sizes are different either.

[Tcoat]

Originally Posted by Gforce

I don't think you did, none of the makers you mentioned fit Torsen.

Clearly I am referring to the makers of the diffs in this post.

Your denying that they are torsens does not make that fact. The best example was the Civic Si which does indeed have a torsen. I don't suppose you even bothered to look at the list I linked for you since you just "know".

[Gforce]

Originally Posted by ZDan

Clutch-type biases torque IMMEDIATELY also, and starts from a positive value instead of zero.

Torque bias of a clutch type starts at a positive value because of the breakaway torque. How much steeper it gets is dependent on tunable factors like ramp angle, active clutch plates, and in the case of the OS Giken I got for the FD, preload springs. I had mine specifically tuned to have minimal breakaway torque and a shallow slope to avoid rapidly increasing lockup with torque.

Understood. I'm not disagreeing with you. Fact remains they slip until the ideal designed in clutch pack load is achieved. Torsen cannot and do not ever slip.

[ZDan]

Originally Posted by Gforce

Understood. I'm not disagreeing with you. Fact remains they slip until the ideal designed in clutch pack load is achieved. Torsen cannot and do not ever slip.

No, clutch type don't slip until differentiating torque overcomes lockup torque. Lockup torque which starts at breakaway torque and increases with applied torque.

Torsens DO slip, if they didn't they wouldn't be in OEM applications. The whole idea is that they slip and act like an open diff under low load, and progressively lock up with applied torque.

[Gforce]

Originally Posted by ZDan

No, clutch type don't slip until differentiating torque overcomes lockup torque. Lockup torque which starts at breakaway torque and increases with applied torque.

Torsens DO slip, if they didn't they wouldn't be in OEM applications. The whole idea is that they slip and act like an open diff under low load, and progressively lock up with applied torque.

Well, we disagree. Torsen don't limit slip at all. They bias the same torque regardless. Torsen act as an open diff all the time as far as differentiation goes.

So, yes, technically Torsen "slip" but that misuses the word slip. Torsen always differentiate, they do not lock. However, if the wheel with less grip still grips then the Torsen "locks" completely, transferring all the "spare" torque. Torque transfer is immediate. They are not progressive, they bias torque at a fixed ratio all of the time.

Torsen are unaffected by different wheel sizes. Completely unaffected.

[Gforce]

Originally Posted by Tcoat

Your denying that they are torsens does not make that fact. The best example was the Civic Si which does indeed have a torsen. I don't suppose you even bothered to look at the list I linked for you since you just "know".

Helical and Torsen are not the same. Don't you ever get tired of constantly being wrong?

[ZDan]

Originally Posted by Gforce

Well, we disagree. Torsen don't limit slip at all. They bias the same torque regardless. Torsen act as an open diff all the time as far as differentiation goes.

C'mon, if it acts as an open diff at all times, it wouldn't do anything to transfer torque.

So, yes, technically Torsen "slip" but that misuses the word slip. Torsen always differentiate, they do not lock. However, if the wheel with less grip still grips then the Torsen "locks" completely, transferring all the "spare" torque.

So, first you say that the Torsen does not lock, then you say that it "locks" completely? WTF?!

Torque transfer is immediate. They are not progressive, they bias torque at a fixed ratio all of the time.

The Torsen does lock up PROGRESSIVELY with applied input torque. Also, the clutch type locks up progressively with applied input torque.

Torsen are unaffected by different wheel sizes. Completely unaffected.

Of course with different wheel sizes there is constant slip, and the faster you drive the greater the torque, and consequently the greater the heat and wear due to friction, which is the mechanism by which the Torsen works in the first place...