Sponsored

Thoughts on replacing Torsen for a clutch type LSD 15 GT PP

EricSMG

Well-Known Member
Joined
Aug 22, 2016
Threads
3
Messages
361
Reaction score
106
Location
San Diego
Vehicle(s)
2017 GT w/PP, 2004 BMW M3 Coupe
Well, to be fair, the expressed purpose of a "Performance Pack" model is to emphasize track performance. By definition, that makes it a focused package for the task at hand. But, with that said, some of this is on Ford; they specified the performance criteria they wanted the Torsen to meet, they also were responsible for the chassis controls and suspension integration. That also includes sorting the wheelhop issue (and its cascade effects).

Part of where I think Ford might've fallen short was in scope of imagination - they envisioned a low-volume package that would be purchased by a handful of people autocrossing and doing track days, etc. Not something that would be adopted by the general consumer. In retrospect, that seems kind of silly, because most people buying the performance version of a sport coupe are probably going to opt for the most performance that is practical. The take rate of that option seems to have been a bit higher than was anticipated. That may be why we've seen indication of a PP2 option on the '18MY. It might be to push the overall performance level up, while also making more of the performance content more readily accessible.

In any case, it does remain true that the limitation of a typical helical-gear differential is the one-wheel unloaded situation, but in my 20-year experience using these types of products, I can't say its been a significant concern. I've not come across the wheelhop-induced unloading as described, so I wonder how relevant it is to the typical buyer. Obviously, it must happen if some folks are reporting it, but on the whole how widespread is it? There are certainly ways around the issue, but all have their tradeoffs. The Wavetrac design seems novel, I'd like to get my hands on one at some point to see how it performs...
Great convo, Rick. It's refreshing to exchange ideas without being rude or dismissive.

Calling the PP focused is not taking a jab at it - it's calling it what it is. I bought it for precisely the reason(s) you state - to get the most practical performance. I always buy/drive the 'sportiest' version of what ever it may be.

Is this "issue" really a problem? Not really. Is it something that's both extremely annoying, and, unique (in my experience) to this car? Absolutely, to me. I'm 40 years old and have driven nothing but front engine rwd cars with traditional and viscous LSDs - none behave this way, which is the point - this doesn't happen in my experience and so it jumps out like a sore thumb. I'm sure others don't even notice it.

I like my cars to be precise, effective and well mannered. Something about the Torsen/IRS combo in this car is none of those at very slow speeds while turning and accelerating gently - if it's wet, forgettaboutit - total and complete mess. I can only imagine ice and snow. Is it the Torsen or the IRS or the combo? Not sure, but it's a nasty little characteristic in the S550 GTPP.

Edit - poly diff bushings inserts have helped this about 25% or so in the dry by highly limiting the wind-up/snap-back of the differential (better wheel control, more constant force). Still very bad in the wet.

I think the issue is people trying to induce wheelspin...
Although this is entirely not the case, think about what you're saying. The entire point of a good diff is to prevent single wheel spin no matter what the driver does. Inducing spin should always result in both wheels spinning for maximum control and forward push.

In other words, saying that the single-wheel-spin is driver-induced essentially proves that the diff/suspension aren't doing their job(s).
Sponsored

 
Last edited:

TexasRebel

Gearshifter
Joined
Mar 19, 2016
Threads
27
Messages
2,500
Reaction score
835
Location
between the mustard and the mayo
Vehicle(s)
2016 YZ GTPP - PP2
That's not how a Torsen works.

The low traction wheel spins, but that spin requires a torque. That torque is multipled by the bias ratio to spin the high-traction wheel.

It is not a 100%-100% posi (increased wear), detroit (clicks), or C-lock (BOOM!).

The entire point of a good diff is to allow the wheels on a common axle to spin at different speeds as needed. The point of a limited slip differential is to get torque to the high traction side when the low traction side is too low.

Which LSD is best? That depends on the application. On pavement I'll take the Torsen T2 any day of the week. However, you might want to seriously consider a T2R. The T2R has a friction preload to eliminate the no-traction wheel slip. ...at the cost of tire wear.

Also, inducing a spin is losing. There's no reason to intentionally turn static friction (control) into kinetic friction (good luck!). The place you want to be is right on the edge of breaking free.
 

Brazenbuck

Well-Known Member
Joined
Dec 24, 2014
Threads
4
Messages
187
Reaction score
40
Location
United States
Vehicle(s)
Mustang
EricSMG said:
The entire point of a good diff is to prevent single wheel spin no matter what the driver does. Inducing spin should always result in both wheels spinning for maximum control and forward push.

In other words, saying that the single-wheel-spin is driver-induced essentially proves that the diff/suspension aren't doing their job(s).
This is wrong.

Texas Rebel said:
The entire point of a good diff is to allow the wheels on a common axle to spin at different speeds as needed. The point of a limited slip differential is to get torque to the high traction side when the low traction side is too low.
This is right.
 

EXP Jawa

Well-Known Member
Joined
Jun 9, 2014
Threads
3
Messages
1,011
Reaction score
205
Location
Rochester, NY
Website
www.torsen.com
First Name
Rick
Vehicle(s)
1999 Cobra Convertible, Electric Green
Both EricSMG & TexasRebel are right and wrong, to a degree. Yes, the idea is to prevent single wheel spin. I think maybe that the implication wasn't that driver was trying to induce single-wheel spin, but rather spin in general. At least that's how I read it. Obviously, with this much torque, its easy to do on demand. Generally, you'd want that to be both wheels, but depending on circumstances, you'd want that limit to be approaches progressively, holding one tire to its limit while the other is being also being loaded to its limit. This is opposed to an all-or-none approach, which can behave very abruptly and lead to a snap-oversteer behavior.

The best way I've found to sum up how a limited slip diff work (any type) is this: they generate internal friction to provide resistance to wheel spin. When that friction is overcome, they differentiate (or spin), allowing the side to side speed difference to occur. The differences between the various types of product amount to how they produce friction and what the characteristics of that friction build are (steady, linear progression, how much total, etc).

So it is fair to say that the Torsen works to try to prevent one-wheel slip from occurring. By building friction, it can support a traction imbalance from side to side, up to a ratio described by the TBR level. So, if its 3:1, it will allow that much difference (as a ratio) of traction difference between drive tires before differentiation can occur. This is where it "sends" 3X the torque to the higher traction tire - by restricting the spin of the low traction tire, allowing torque to flow to the high side. It isn't so much sent from the low to the high, but the low is restrained at its own limit so that high side can have more delivered from ring gear (if that makes sense). If the driver attempts to send more torque to the axle than the combination of the low traction tire's limit plus that number X TBR (going to the high side), that's when the low traction tire will slip, not before. But in 75-80% of cases (depending on TBR), that enough to prevent any noticeable slip from occurring - and when it does occur, the differential continues to bias torque to the high traction side.

This does come back to the limitation of the low-traction tire - its important to maintain some reaction torque on the wheels for the diff to work against. This usually means keeping the tire on the ground, but can be supported momentarily with the brake system since both provide a negative, reaction type torque to the system. TCS usually does this pretty well, but parking brake can work, as can simply touching the brake pedal briefly (brake modulation).

I'm not sure if any of that clears things up or muddies the waters. However, Eric, I'm glad we can have a fair conversation about this, rather than simply claiming that Torsens cause crashes. It goes both ways, as end users can be just as opinionated as manufacture reps. I do hate it when a vendor or rep is dismissive. That basically describes the conversations I had both with a Wavetrac vendor as well as Wavetrac's rep when I asked specific functional questions at the PRI show. I know its annoying when someone tells you "oh, its stronger than you need" or "it works better just because", etc. Not that those where the actual words spoken, but you get the point - it is, frankly, insulting to an extent, so I try not to do that.

Anyway, TexasRebel, the Torsen in the GTPP is already a T-2R (whereas the GT350 has the Type-2). The difference between them is not a system preload, but rather enhanced torque bias ratio. The only (production) differentials that Torsen has produced that use a preload have been standard Type-2s that added a spring pack, whereas preloaded T-2R models where only built in small numbers for off-road racing.
 

Brazenbuck

Well-Known Member
Joined
Dec 24, 2014
Threads
4
Messages
187
Reaction score
40
Location
United States
Vehicle(s)
Mustang
The purpose of a differential is to differentiate axle speeds when negotiating a turn. It's not to "prevent single wheel spin no matter what the driver does.". That's simply incorrect.
 

Sponsored

EXP Jawa

Well-Known Member
Joined
Jun 9, 2014
Threads
3
Messages
1,011
Reaction score
205
Location
Rochester, NY
Website
www.torsen.com
First Name
Rick
Vehicle(s)
1999 Cobra Convertible, Electric Green
The purpose of a differential is to differentiate axle speeds when negotiating a turn. It's not to "prevent single wheel spin no matter what the driver does.". That's simply incorrect.
Agreed, in general terms. A basic differential exists to allow the wheels to change speed while still providing drive, no other reason. Within the context of this discussion, its probably fair to assume that above statement was with regards to a limited slip differential, because, well, that's what we're talking about here. In which case, that might make an idealized description, but not accurate just the same.

There is only one class of differential that "will prevent single wheel spin no matter what the driver does". That product is a locker (locking differential). These lock up 100%, essentially having a TBR that is infinite:1. These also have a lot practical downsides, which is why they're not preferred in on-road applications, on-road applications in which good vehicle dynamics are important.

The only LSD product that might be close is an eLSD, but that depends highly on the tuning and amount of lockup it really has, and is still a reactionary (not proactive) device. The basic clutch type LSD (Traction Lock) is not going to prevent single wheel spin under every circumstance. Yes, it may an advantage here or there, but if you lift a tire off of the pavement while applying torque, it will spin. So, that idealized statement doesn't really play out in practice.
 

Brazenbuck

Well-Known Member
Joined
Dec 24, 2014
Threads
4
Messages
187
Reaction score
40
Location
United States
Vehicle(s)
Mustang
Agreed, in general terms. A basic differential exists to allow the wheels to change speed while still providing drive, no other reason. Within the context of this discussion, its probably fair to assume that above statement was with regards to a limited slip differential, because, well, that's what we're talking about here. In which case, that might make an idealized description, but not accurate just the same.

There is only one class of differential that "will prevent single wheel spin no matter what the driver does". That product is a locker (locking differential). These lock up 100%, essentially having a TBR that is infinite:1. These also have a lot practical downsides, which is why they're not preferred in on-road applications, on-road applications in which good vehicle dynamics are important.

The only LSD product that might be close is an eLSD, but that depends highly on the tuning and amount of lockup it really has, and is still a reactionary (not proactive) device. The basic clutch type LSD (Traction Lock) is not going to prevent single wheel spin under every circumstance. Yes, it may an advantage here or there, but if you lift a tire off of the pavement while applying torque, it will spin. So, that idealized statement doesn't really play out in practice.
You're right that we're talking about limited slip differentials. Limited slip by definition means some slip is allowed. You want that for going around turns.

There's no differential that meets the no-matter-what use case. Lockers wear out. Sometimes those preload springs can't hold. The only device that truly does meet this use case is a spool which binds the axles together. As you know, a spool by definition isn't a differential.
 

TexasRebel

Gearshifter
Joined
Mar 19, 2016
Threads
27
Messages
2,500
Reaction score
835
Location
between the mustard and the mayo
Vehicle(s)
2016 YZ GTPP - PP2
This is where it "sends" 3X the torque to the higher traction tire - by restricting the spin of the low traction tire, allowing torque to flow to the high side. It isn't so much sent from the low to the high, but the low is restrained at its own limit so that high side can have more delivered from ring gear (if that makes sense).
It makes sense, but the wording is contradictory. As I understand dynamics, the friction in the differential must be initiated by rotational differentiation (unbalanced wheelspin). This, however, happens at such a small degree of rotation that it is nearly undetectable by eyesight, and torque is transferred to the high traction side up to the TBR multiplier.

At this point one of four things can happen:
1) too little torque on the low traction side spins the low traction wheel
2) too little traction on the high traction side spins both wheels, albeit with different torques.
3) too little available torque stalls movement.
4) the vehicle moves.

Anyway, TexasRebel, the Torsen in the GTPP is already a T-2R (whereas the GT350 has the Type-2). The difference between them is not a system preload, but rather enhanced torque bias ratio. The only (production) differentials that Torsen has produced that use a preload have been standard Type-2s that added a spring pack, whereas preloaded T-2R models where only built in small numbers for off-road racing.
I don't know where I got the preload thing from. Maybe I muddled the literature up in my head. It's interesting that Ford would put a tuneable Torsen in the PP cars, but not the GT350s. It's also nice to know I have a T2R. I've looked everywhere for that info and ultimately just assumed the PPs got the T2 as it was only ever advertised as "T2" without the "R".
 

EXP Jawa

Well-Known Member
Joined
Jun 9, 2014
Threads
3
Messages
1,011
Reaction score
205
Location
Rochester, NY
Website
www.torsen.com
First Name
Rick
Vehicle(s)
1999 Cobra Convertible, Electric Green
At the risk of splitting hairs (of which I don't have many left), I've always interpreted the term "limited slip" as a device that reduces undue slippage, rather than a device that allows some slipping to occur. I know, its semantics, more of a glass half-full thing.

Also, less about semantics and more about accuracy, you don't want "slip" for going around turns, you want differentiation. This means a wheel speed difference of up to, maybe, 25 or 30 RPM from side to side. Most of the time its less. Wheel slip is essentially what happens when differentiation becomes excessive, say, 100 RPM or more. Severe slip is 1000+ RPM delta-N rates. So, there's differentiation, slip, and severe slip, all distinguished by order or magnitude basically. Not trying to be a prick, but this stuff is what I do for a living... :D

But anyway, a good LSD allows ready differentiation while limiting undue slippage. Generally easier said than done, and often easier done than explained!
 

EXP Jawa

Well-Known Member
Joined
Jun 9, 2014
Threads
3
Messages
1,011
Reaction score
205
Location
Rochester, NY
Website
www.torsen.com
First Name
Rick
Vehicle(s)
1999 Cobra Convertible, Electric Green
It makes sense, but the wording is contradictory. As I understand dynamics, the friction in the differential must be initiated by rotational differentiation (unbalanced wheelspin). This, however, happens at such a small degree of rotation that it is nearly undetectable by eyesight, and torque is transferred to the high traction side up to the TBR multiplier.

At this point one of four things can happen:
1) too little torque on the low traction side spins the low traction wheel
2) too little traction on the high traction side spins both wheels, albeit with different torques.
3) too little available torque stalls movement.
4) the vehicle moves.
The friction in the (Torsen) differential is initiated by applying torque to the ring gear, which transfers it to the diff case. Torque causes gear thrust force reactions within the assembly, which push the gearing into the case and generate friction. This static friction is used to resist the onset of slippage. Think of pushing a crate across the floor. You don't need to move it to create the friction that keeps it from sliding. Its own mass does that. You have to overcome its static friction in order to slide it. Same thing in the differential.

You'd be surprised, in fact, how little differentiation may actually occur if you were lapping a track, because the car's chassis doesn't put enough loads on the tires to make them overcome the friction load generated within the diff by the amount of torque that's applied to the axle. Generally, you can get to a certain torque load for a specific chassis, referred to as "performance torque" above which, the car's chassis cannot make the differential differentiate no matter what it does. But you have to have a lot of torque on the axle to achieve that, it isn't something that happens from parking lot speeds.


I don't know where I got the preload thing from. Maybe I muddled the literature up in my head. It's interesting that Ford would put a tuneable Torsen in the PP cars, but not the GT350s. It's also nice to know I have a T2R. I've looked everywhere for that info and ultimately just assumed the PPs got the T2 as it was only ever advertised as "T2" without the "R".
Some earlier literature showed preloading as well the bias-enhancing features, sort of as a "these options are possible" situation. That literature was intended for the OEM customer, not end users, where designs would be tailored to suit. Either way, that's not how its done anymore anyway. Basically, a Type-2 has the TBR at whatever its natural friction level makes it. The T-2R model has enhanced TBR, now days typically by a helical spline coupling. It isn't tuneable or readjustable by any means, it is what it is once the geometry is set. But the basic design allows other design or TBR configurations without major retooling.

In the case of the current Mustang variants, the Mustang team wanted one level of performance for the PP, whereas SVT had different ideas about what was ideal for their version of the chassis. Simple as that...
 

Sponsored

Brazenbuck

Well-Known Member
Joined
Dec 24, 2014
Threads
4
Messages
187
Reaction score
40
Location
United States
Vehicle(s)
Mustang
At the risk of splitting hairs (of which I don't have many left), I've always interpreted the term "limited slip" as a device that reduces undue slippage, rather than a device that allows some slipping to occur. I know, its semantics, more of a glass half-full thing.

Also, less about semantics and more about accuracy, you don't want "slip" for going around turns, you want differentiation. This means a wheel speed difference of up to, maybe, 25 or 30 RPM from side to side. Most of the time its less. Wheel slip is essentially what happens when differentiation becomes excessive, say, 100 RPM or more. Severe slip is 1000+ RPM delta-N rates. So, there's differentiation, slip, and severe slip, all distinguished by order or magnitude basically. Not trying to be a prick, but this stuff is what I do for a living... :D

But anyway, a good LSD allows ready differentiation while limiting undue slippage. Generally easier said than done, and often easier done than explained!
You're not being a prick, say what you've got to say. My feels aren't in these forums.

I get what you're saying and it's semantics. You don't WANT slip, but isn't that how you get differentiation in some diffs? I wasn't referring to wheel slip. In the case of clutch LSD or locker, once you get into a turn the differentiation (what you do want) is achieved by slip in the clutch pack or the block in the locker that receives a force of the preload springs and engages the axle. I forget the name of they section of the locker, but I haven't had my coffee yet and I know you know what I'm talking about LOL
 

gsxr1300

Well-Known Member
Joined
Nov 16, 2014
Threads
11
Messages
2,101
Reaction score
454
Location
MA
First Name
Jason
Vehicle(s)
2015 Mustang GT PP
I have a Torsen pumpkin with 5K miles on it with 3.31 gears professionally installed for sale if anyone is interested.
 

BmacIL

Enginerd
Joined
Sep 21, 2014
Threads
69
Messages
15,010
Reaction score
8,920
Location
Naperville, IL
Vehicle(s)
2015 Guard GT Base, M/T
Vehicle Showcase
1
I have a Torsen pumpkin with 5K miles on it with 3.31 gears professionally installed for sale if anyone is interested.
Now that's tempting...
 

TexasRebel

Gearshifter
Joined
Mar 19, 2016
Threads
27
Messages
2,500
Reaction score
835
Location
between the mustard and the mayo
Vehicle(s)
2016 YZ GTPP - PP2
The friction in the (Torsen) differential is initiated by applying torque to the ring gear, which transfers it to the diff case. Torque causes gear thrust force reactions within the assembly, which push the gearing into the case and generate friction. This static friction is used to resist the onset of slippage.
So the torque being transferred from the diff. case, through the parallel gears, and to the axles is what forces the parallel gears into the sides of their pockets? I thought the action of differential torque applied to the axles forced the floating parallel gears outward, away from the helical gears on the axles, into the pockets. More torque differential, more normal force, more friction (as well as more parallel pairs, and larger diameters, and different pitches, &c.)

You'd be surprised, in fact, how little differentiation may actually occur if you were lapping a track, because the car's chassis doesn't put enough loads on the tires to make them overcome the friction load generated within the diff by the amount of torque that's applied to the axle.
Nah. I painted the tops of my tires once after I reclutched an Eaton Posi on my C10. Without any weight in the bed those marks almost never moved. Everything went into increased tire wear.




Some earlier literature showed preloading as well the bias-enhancing features, sort of as a "these options are possible" situation. That literature was intended for the OEM customer, not end users, where designs would be tailored to suit.
I'm in that grey area of being and OEM customer (well, designer technically, but that's how parts end up in OE specs) as well as an end user in my free time. ;) You've still avoided my post #45 question.
 

abuch91

Well-Known Member
Joined
Nov 17, 2015
Threads
5
Messages
235
Reaction score
108
Location
NoVA
Vehicle(s)
2016 Ruby Red GT PP
Pardon my dumbass question amidst a sea of technical talk, but would the Torsen contribute to my right rear tire wearing out much faster than my left? Tire guy said it didn't look alignment related and that it was due to the diff.
Sponsored

 
 




Top