Who can explain static lock on LSD

[M0nk3y]

Title explains it all.

I'm window shopping for LSDs and I'm seeing all this static lock. At first I thought I knew what I was looking at but now I just have no clue.

So, anyone have an idea on how to explain static lock, for example...OEM LSDs are 25%, but I see alot of aftermarket Clutch-Type LSDs around 40%. I came across one today that boasted 80% static lock.

If you want, you can explain ramp angles, I get that concept...1 way vs 1.5 vs 2 way...but, if you want to get on your soap box go for it

And finally, if you really want...explain clutch packs and how more clutches are better/worse for a LSD

Thanks :)

PS - I'm not interested in Torsen LSDs (Quaife, Wavetrac...etc)

[derbo]

Kyle,

From my understanding of limited-slips,

Static lock is the amount of lock that is always available. The idea is that static lock is to prevent losing traction, where as dynamic lock is reacting to the loss of traction. I'm not exactly keen on the ways of describing it, but I'm sure Jim can help you.


http://www.performancegearing.com/specifications.php


Specifications

Limited slip units that can generate dynamic lock prior to torque reaching
the rear wheels behave in an anticipatory way, since the dynamic lock is generated BEFORE
there is a need for it. With this kind of anticipatory engineering, wheel slip should (theoretically) not occur. However, that assumes that either:

1)both rear wheels are getting the same relative amount of traction, or
2) this type unit also has sustainable, static lock capabilities.

In the anticipatory category of limited slip units are the helical gear type (e.g.,Quaife, Torsen) and all the friction disc type units. The lesser of these two types - helical gear - have only dynamic lock capabilities (at a pre-determined, unmodifiable rate).

Without static lock, if both rear wheels are not getting comparable traction, the instant one wheel slips, the differential will behave like an open (non-limited slip) unit . All torque will be diverted
to the slipping wheel. But even among the various friction disc type units, any without an internal mechanism to maintain stock lock over time will act just like a helical gear type if one wheel is registering significantly less traction than the other. The more aggressive your driving, the less effective units with no static lock will be.

The Performance Gearing Limited Slip is the only comprehensive, anticipatory
differential that can:
1) provide on-demand lock (static/dynamic) in all driving situations and
2) either statically or dynamically generate more lock for any given driving situation to prevent any wheel slippage before it can even occur.

Overview
Because of an across-the-board change in differential design, BMW has eliminated the interchangeability of differential components that was common in models from 1971 to 1994. With the exception of models with Motorsport ("M")engines, limited slip has also disappeared as an option. Because of the ring-and-pinion offset and smaller confines of the main housing itself, no existing stock limited slip unit will fit any of the '95 and later models (except E36,Z3).

Since Traction Control is a safety feature and not a performance enhancement, the challenge was to design a limited slip unit that used the proper sized internal components - bevel gears, cross shafts, ramp housings, friction discs, etc. - that could withstand the torque/horsepower outputs of the 6 cylinder, V8 and V12 engines, respectively.


drawingside view of product
Schematic and actual machined limited slip unit.
Internal components include:
- five 90mm, ceramic coated friction discs (6 cylinder models),
four 106mm friction discs (twin-turbo, V8/V12 models)
- ramp housings machined for three dynamic lock generation options: 30/50,
40/60, 45/45
- opposing Bellville washers for maintaining static lock

Lock options with the limited slip differential
The comprehensive Performance Gearing Limited Slip Differential is the most flexible
with regard to configuration of any of *those available in the BMW aftermarket*. Both *static lock* and *dynamic lock* are configured independently for optimum performance in ALL driving situations,
depending upon how the car is intended to be driven.

Typical configurations would include:

street : 30% static lock , 45/45 angle ramp housings for a modest amount of
wheel slip control during cornering, with options for dynamic lock in both
acceleration and deceleration mode (from 30% to 70%)

aggressive street/driver's schools : 40% static lock (for maximum inside
wheel spin control, without generating understeer) and 30/50 ramp angles for
the most aggressive dynamic lock during acceleration (40% to 100%), yet
modest dynamic lock during deceleration (40% to 60%)

dedicated track: 25%-30% static lock , 40/60 ramp angles for more dynamic
lock governing during full-throttle acceleration (30% to 100%) and hard
deceleration (30% to 40%) to prevent rear wheel lockup.

With the Performance Gearing Limited Slip Differential, there are a multitude of
configurations to suit individual driver's needs.

[M0nk3y]

Thanks,

I emailed Jim @ Performance Gearing and got this response:


Thanks for your inquiry.

It's too bad they won't allow at least the stock automatic ratio (3.73) , but at least limited slip is an option.

The type of limited slip I use is a Salisbury ( friction discs , mechanical ramping) because it is capable of both static and dynamic lock in a multitude of configurations depending upon engine output and intended use. It's true that your M-variable LS unit has static lock ( the "at rest" resistance to rotation between the rear wheels), but for some reason BMW decided to keep the amount of static lock very low (usually in the 40 lbs/ft range, but some brand new units are only 25 lbs/ft) neither setting being very helpful because 1) 25% would be closer to 60 - 70 lbs/ft and 2) there is no mechanism to maintain the static setting by compensating for wear over time, so the M-variable static setting dissipates to zero over time. Another reason that the M-variable is not considered to be a competition LS is that the dynamic lock ( additional to static lock, throttle initiated) is viscous activated, so just like a Subaru, there is a time-lag of 1.5 to 2.0 seconds during inside wheel spin that it takes to move the viscous fluid from its reservoir to the pressurized chamber that loads the friction discs. Certainly streetable, but frustrating on the track.

The distinct advantage to the Salisbury , when setup correctly, is 1) static lock can be set at an optimal amount on a case-by-case basis, to match driving style, type of competition, engine torque curve, etc. to a "sweet spot" in the 35-40% range and 2) dynamic lock is anticipatory (not reactive to wheel spin) in that given how much you throttle the engine and what ramp angle you are using this additional lock occurs before torque reaches the rear wheels, instead of after. Ideally, you only want as much lock at you need at any given moment, so although this is more of a lock baseline (to cover extreme situations e.g., inside wheel off the ground) , the dynamic (throttle controlled) lock that momentarily adds lock depending upon the situation is fully variable/controlled by the driver and whatever separate rates for acceleration/deceleration are chosen during the build.

Since the 128i never had limited slip, converting will definitely change the handling/performance characteristics, so you will need to establish a baseline and determine, over time, what if any handling issues need to be enhanced or modified. It is up to you where you want to start but my recommendation would be:

- four 90mm friction discs
- either 45/45 or 40/90 ramp angles
- 75-80 lbs/ft static lock

And even though a race configuration, it is perfectly streetable because the transitions from accel/decel and static/dynamic lock and back are seamless/quiet.

The unit I would ship would be a complete final drive with the limited slip and 3.23 gearset installed/setup, so that it only needs gearlube added before installing into the rear subframe exactly like your stock final drive is removed.

[ZHPRegistry.net]

I can try.

Static lock by itself is a bad thing. Ideal LSDs have 0 static lock. These are the high end units that use a viscous silicon solution that "kicks in" when the difference in wheels speeds is detected, or the newest M locking diff that uses an electric motor to essentially create pressure on the clutch packs when needed and fully loosen them when the diff needs to "open".

Now, let's get back to the simplest and most cost efficient LSD, and that is the conventional clutch pack / ramped lsd. The idea behind ramp angles is that when you apply force or torque to the input shaft of the differential, it puts force on the pinion that is sitting in between the ramps, this in turn pushes the ramps laterally which put extra pressure on the clutch plates.. more pressure you put on clutch plates more friction they provide, more friction means more force that fights difference in wheels speed. As you can imagine, the harder you press on the gas pedal in a turns the more your rear wheels lock together.

The key takeaway here is that the amount of locking you would like to have in the rear wheels at let's say 30% of throttle input is typically more at the track vs street. So how do you change the diff to provide more locking at that speed? One way is to make the ramp angles less steep. Think of this from the point of view of the pinions between the ramps.. if the angle is less steep, it is easier from the pinion to spread the ramps and therefore the same torque exerts more force onto the clutch plates. Changing ramp angles is not easy because it requires machining, though a good LSD shop will recommend going down this route. What else can we do? Well, you can increase the amount of static lock, you now have more locking under load since locking under load builds up from static lock (pretty trivial right, static lock what you build up from, so if you start with more for the same throttle input you'll have more locking).

From here you ask yourself, how do I increase static lock? The clutch pack in the LSD is under pressure due to Belleville washers (google that). One route is to put thicker plates or dog ear washers and essentially increase the pressure on discs, thus increasing static lock, thus increasing overall locking capability under load. This is a shitty way of doing it, though I know people who have taken this route. Why is it shitty? Well, your clutch plates now are constantly being pushed together under higher tension, which means every time you take a slightest turn in calm driving conditions you plates are shaving away the special Moly coating. Pretty soon your LSD will be shot.

What else can we do? Well... have you seen the 2 to 3 clutch pack conversion often mentioned in the rebuilds? Turns out that many BMW LSDs come with a special spacer inside, this spacer can be replaced with another clutch disc. So now you have the same amount of tension inside your LSD but and extra clutch disc that provides friction. In essence you've increased your static lock without putting any more pressure on any of the individual discs. The reason why you have more friction is because you have more friction surface.

The 3 clutch setup will give you the 40% lock. Don't concern yourself with this number too much... it's not overly scientific. It basically means that in case of loss of traction in one wheel 40% of the input torque will be sent to the wheel with traction. The reality is not this precise because LSD at different stages of their life have different real locking rates.

Any rebuilt that gives you more than 40% locking is probably runs the clutch packs at higher tension, I personally consider this a bad thing unless you plan on rebuilding your LSD every 25-40K miles.

Was this at all helpful?

[M0nk3y]

Yes, very helpful. Thank you so much for writing at up.

So, with that being said, do you agree with Jim's recommendation for a LS?

- four 90mm friction discs
- either 45/45 or 40/90 ramp angles
- 75-80 lbs/ft static lock

How much does a typical rebuild cost, and how often does it need to be done? (Can it be measured by raw track time, or mileage?) Or does daily driving still contribute wear and tear on a LS?

I've been heavily looking into a clutch-type LS mainly because a torsen LS basically fails on ice...and goes open diff. As well, it resorts to an open diff under braking and off throttle. Do you agree with these statements?

One person made a comment on 1addicts:


The performance gearing LSD in my 330i is absolutely silent as is the quaife in my 135is. Also there is no harsh engagement. It does behave completely different though than the quaife. All clutch based LSDs will tend to lock on hard acceleration whether there is actual slip or not. The affect of this is that if you are reasonably hard on the gas coming out of a corner the diff lockup will add some understeer tendency. This affect at the track was my primary reason for switching from the clutch type on my 330 to the quaife when I bought my 135is.

To be sure though, Jim at Performance Gearing makes an absolutely silent and top notch LSD. I've had mine on for 6 years with multiple track sessions per year with no issues. He even added a drain plug to my diff as part of the build.

The lockup concerns me in turns that will contribute to understeer. However I still don't grasp how the Quaife or any other will create lock-up, since there is really no quantifiable way, at least in my searching there isn't

[ZHPRegistry.net]

Yes, very helpful. Thank you so much for writing at up.

So, with that being said, do you agree with Jim's recommendation for a LS?

- four 90mm friction discs
- either 45/45 or 40/90 ramp angles
- 75-80 lbs/ft static lock



75ft/lb of BAT (Breakaway torque) and 4 clutches tells me that the clutches are not under high pressure, which is good. Consider that the BAT of stock LSD diffs on BMWs tends to be ~50 ft/lb with just 2 clutches. This sounds like a solid track setup. What is your realistic mode of usage for the car? Are you basically going to attend 1 track event a year? or 10.

As for the ramp angles, 40/90 is a great setup, Shallower angle on accel should give you higher locking under acceleration and the 90 degree cut for decel will essentially reduce under steer off throttle.

The other thing to keep in mind is that understeer and oversteer is often evaluated as a static property of a car, and it's not. Higher lock LSDs will offer more underteer under braking but oversteer under hard acceleration. Think drifters... they weld their diff (100% static lock) so they can swing the back side of the car with ease... those cars are also very hard to turn when you are rolling off throttle.




How much does a typical rebuild cost, and how often does it need to be done? (Can it be measured by raw track time, or mileage?) Or does daily driving still contribute wear and tear on a LS?

I've been heavily looking into a clutch-type LS mainly because a torsen LS basically fails on ice...and goes open diff. As well, it resorts to an open diff under braking and off throttle. Do you agree with these statements?

One person made a comment on 1addicts:



The lockup concerns me in turns that will contribute to understeer. However I still don't grasp how the Quaife or any other will create lock-up, since there is really no quantifiable way, at least in my searching there isn't

For the setup you mentioned above I would think you can easily go 70K miles before needing to rebuild it, really depends on how hard it was used. Clutch discs cost ~60 bucks each. I am not a mechanic and I can take a apart a diff and replace the clutches in about 3 hours. Funny enough I just rebuilt my E30 M3 diff.

[M0nk3y]

75ft/lb of BAT (Breakaway torque) and 4 clutches tells me that the clutches are not under high pressure, which is good. Consider that the BAT of stock LSD diffs on BMWs tends to be ~50 ft/lb with just 2 clutches. This sounds like a solid track setup. What is your realistic mode of usage for the car? Are you basically going to attend 1 track event a year? or 10.

As for the ramp angles, 40/90 is a great setup, Shallower angle on accel should give you higher locking under acceleration and the 90 degree cut for decel will essentially reduce under steer off throttle.

The other thing to keep in mind is that understeer and oversteer is often evaluated as a static property of a car, and it's not. Higher lock LSDs will offer more underteer under braking but oversteer under hard acceleration. Think drifters... they weld their diff (100% static lock) so they can swing the back side of the car with ease... those cars are also very hard to turn when you are rolling off throttle.



For the setup you mentioned above I would think you can easily go 70K miles before needing to rebuild it, really depends on how hard it was used. Clutch discs cost ~60 bucks each. I am not a mechanic and I can take a apart a diff and replace the clutches in about 3 hours. Funny enough I just rebuilt my E30 M3 diff.

Would the higher locking under accel (40/90) cause those understeer issues under a track-out type of situation? For example, I'm use to my Z4M with the stock diff being able to track out with slight oversteer, with little understeer under acceleration.

So with the 90 degree decl, I'll still be able to rotate the car off-throttle and help reduce push?

Realistic mode for this car is a Autocross, Track and DD car. It will see 10-15 autocrosses per year, 2-3 track weekends per year as well. End goal is to make it a dedicated autocross and time trial car in a couple years.

For reference, I have already ordered (but not installed yet) TCKline Dual Adjustable Suspension, APEX 17x8.5 ET40 Wheels, M3 Subframe Bushings, Eibach 28mm Sway Bar, Hawk HP+ Pads for AutoX and HT-10s for Track, and some sort of Tire (either hankook or dunlop) in a 245 size, squared.


Good to know clutch replacements aren't going to cost thousand +...which was a concern.

Good example of oversteer tracking out at the beginning of this video:


http://www.youtube.com/watch?v=QoXqTzrCx2g

[ZHPRegistry.net]

I think 90 degree cut ramp with 75 ft/lb BAT is probably close to 45 cut ramp (stock) and 50 ft/lb BAT. So in short, my guess you will see very little change off-throttle. With throttle on, all else being equal you will probably have more overstreer. Looking at the video you posted I would classify your car as under steering (not excessive) so the new LSD should work well once you get to use. There were couple of the turns that you had to slow down more than ideal to make the car go around, that's a typical sign of a car that pushes.

... stay closer to the cones :)

[M0nk3y]

I think 90 degree cut ramp with 75 ft/lb BAT is probably close to 45 cut ramp (stock) and 50 ft/lb BAT. So in short, my guess you will see very little change off-throttle. With throttle on, all else being equal you will probably have more overstreer. Looking at the video you posted I would classify your car as under steering (not excessive) so the new LSD should work well once you get to use. There were couple of the turns that you had to slow down more than ideal to make the car go around, that's a typical sign of a car that pushes.

... stay closer to the cones :)

That's the Z4M, with just camber shims and 3200 lbs...so yes, it is an understeer pig. The LSD is for my 128, which will have DAs and actual camber...

As far as staying close to cones...I'm good ;)

https://fbcdn-sphotos-c-a.akamaihd.net/hphotos-ak-prn1/t31/858095_4691691458479_1063005098_o.jpg