The argument for/against FM in a rear diff + LSD

[dparm]

Come spring when my M3 is done with winter storage, it'll be due for a rear differential fluid change. The rear diff comes with a GKN Visco Lok, which is a speed-sensitive LSD. Factory fill is Castrol Syntrax Limited Slip 75w140 with additional friction modifier -- more on that below.

When the E92 M3 was first released in 2008, people complained about a noisy rear differential. Specifically, there was some chattering at low speed when the steering wheel was at full-lock (think parking lot maneuvers). Obviously the fluid had friction modifier already, but BMW decided to up the FM content to address customer complaints. If I recall correctly, the service bulletin had a recommended amount of FM to add, but also mentioned it should be done "to customer preference". If you go buy a bottle of it at the dealer, it's actually private-labeled for BMW and has the additional FM in it.

My question is this: doesn't additional friction modifier actually reduce LSD locking? If that's true, isn't the best performance gleamed by NOT using it? Are there any other downsides to less FM other than the noise? I understand BMW has to balance performance and driveability to keep owners happy, but I do track the car so performance is important.

There are a few fluids that do not contain any FM whatsoever, and of course the vast majority of differential fluids contain the "normal" amount of FM. Which avenue should I take?

(FWIW, the fluids I'm considering are Renewable Lube, Gulf Competition LS, Torco SGO, and the non-BMW Syntrax LS)

 

[rslifkin]

It depends on the type of LSD. If it's a clutch type, then yes, more FM makes it less grabby (which can be used as a tuning tool). For other types of LSD, it shouldn't make it any less effective.

Adding too little FM can make the LSD's engagement rough and chattery, which isn't necessarily good for performance. I'd add the minimum amount it takes to keep it from being chattery, except for maybe in full lock turns on a grippy surface with sticky tires.

If the car doesn't need the tighter locking to put down power without spinning the inside tire in a corner, making the LSD a little less aggressive won't hurt performance (and may actually help by reducing understeer slightly on corner entry).

 

[dparm]

Good info, thanks rslifkin. It is a clutch-type LSD. You can read about it here:
http://www.gkn.com/driveline/about-us/Documents/datasheets/ViscoLok-engl.pdf

That being said, is it best to start with no FM at all and then add to taste (tricky, but not impossible), or is a pre-mixed differential fluid the better option?

One problem is that all of these fluids are slightly different viscosities, and the amount of FM is not disclosed.

 

[Doog]

There is no argument.....if you need xxx amount you need it

 

[rslifkin]

Most of the pre-mixed fluids don't have a ton of FM in them, so you're probably fine to use one of those (and then add more if you determine it's needed).

 

[JHZR2]

I recall reading on here long ago (and admittedly Ive propagated it since) that the FM adds can actually "harden" the gear faces, which was a good thing. (Mechanism/reality of this unknown).

Id speculate that for a stock LSD on a fairly common car, that the FM levels already included in most decent lubes like M1, Amsoil, Redline, Fuchs, Castrol XJ (which I believe should be OEM) should be sufficient. If you've tuned the LSD then all bets are off...

If you street use the car as much as track it, Id probably err on the side of longevity over absolute performance.

 

[440Magnum]

Originally Posted By: dparm

My question is this: doesn't additional friction modifier actually reduce LSD locking? If that's true, isn't the best performance gleamed by NOT using it? Are there any other downsides to less FM other than the noise? I understand BMW has to balance performance and driveability to keep owners happy, but I do track the car so performance is important.


I wouldn't say that it "reduces locking." It makes the jump between the static (holding) and static (slipping) coefficients of friction smaller, so you don't get into a vicious grab/release/grab/release cycle that results in chatter. It shouldn't significantly reduce the actual holding capacity of the LSD clutches enough to really matter, and any loss is more than offset by a more controlled onset of locking.

Its been my experience that the "limited-slip ready" gear lubes really don't have enough LSD modifier in them, but I guess it all depends on the unit. My experience is with the SRT Getrag LSD (modern) and the old Dana/Spicer (clutch pack) and Auburn (cone clutch) units used in Mopars.

 

[MolaKule]

Quote:
My question is this: doesn't additional friction modifier actually reduce LSD locking?


LSD "modifies" the friction coefficient to reduce chatter.

I always recommend adding about 1 ounce increments of LSD additive at a time and then doing figure 8's in a parking lot with a warm diffy.

If it still chatters, add another ounce.

Most people DO add too much of the additive such that the LS may not lock up when needed.

 

[dparm]

Thanks MolaKule, the figure-eight thing is very similar to BMW's procedure of driving clockwise 20 times, then counterclockwise 20 times.


Seems like the safest bet is to get a fluid with FM already mixed in, and then just add more if the chatter/whine is bothering me. The LSD probably wasn't engineered for a non-FM fluid, meaning the behavior might be unpredictable. Per the above PDF, the locking characteristics are "tuneable" by the OEM.

 

[jaj]

When I had my 2008 E92 M3, it came with Castrol SAF-XJ in the differential (75W-140 with FM) and it never chattered or caused any noise. Mine was a very early car, and I know that the later ones were covered by a TSB about adding more FM, but it was never an issue for mine.

So, I'd track down some SAF-XJ and start with that. It may be all you need.

 

[dparm]

Originally Posted By: jaj
When I had my 2008 E92 M3, it came with Castrol SAF-XJ in the differential (75W-140 with FM) and it never chattered or caused any noise. Mine was a very early car, and I know that the later ones were covered by a TSB about adding more FM, but it was never an issue for mine.

So, I'd track down some SAF-XJ and start with that. It may be all you need.


SAF-XJ is now Syntrax LS, per my original post.

If your car ever got a rear diff fluid change (every 3rd oil change), you actually have the new stuff with extra FM. BMW no longer stocks the normal version, but there's obviously places to get it.

 

[dailydriver]

Question;

Besides the reduction (or elimination) of chatter, does the FM fluid actually save measurable wear and tear on clutch packs in a mostly street driven LS differential??

 

[Shannow]

My Nissan, without FMs, you hear the diff "bang" on slow speed corners, skid the inside wheel, and plough understeer at walking pace in wet carparks.

Not nice.

 

[asand1]

With a name like "GKN Visco Lok" I would think it was a viscous (VLSD) limited slip differential, and not need any FM at all?

 

[Gasbuggy]

Originally Posted By: dparm

Seems like the safest bet is to get a fluid with FM already mixed in, and then just add more if the chatter/whine is bothering me. The LSD probably wasn't engineered for a non-FM fluid, meaning the behavior might be unpredictable. Per the above PDF, the locking characteristics are "tuneable" by the OEM.


That is what you should do. You've probably got some miles on your diff and that wear could make it less likely to be an issue for you.

 

[meep]

In my 'yota, I usually need between 1/2 to 2/3 a tube of FM on top of a full D/R of amsoil. Otherwise it does bind/release/bind when pulling out of my street into traffic.

 

[dparm]

Originally Posted By: asand1
With a name like "GKN Visco Lok" I would think it was a viscous (VLSD) limited slip differential, and not need any FM at all?



Well, BMW puts it in from the factory.

The viscous term is derived from this key point (in the PDF up top):

Operating Principle
> A self-contained silicone-fluid filled reservoir
and shear pump generates a hydraulic
pressure proportional to a speed difference
across the axle
> The speed difference between the feed disc
and the fluid filled channel of the pump disc
and the resulting fluid shear forces moves the
fluid from the reservoir to the apply piston
> The hydraulic pressure creates an axial force
via the apply piston, compressing the clutch
pack

 

[turtlevette]

I really hate the term friction modifier. Its not descriptive enough. Is the substance causing a slippier surface such as in diffs or a grabbier surface such as in auto trans clutches. There should be a term for each. Laymen get confused all the time about this. There is a tendency to believe its just one or the other.

We've argued many times about adding snot to diffs.

A. If you drive an appliance and are worried about every little noise then run the stuff.

B. If you drive a performance car 1, why are you worried about noise? 2. The goal is for the clutches to grab. Its not smart enough to know you're doing a slow speed turn and grabs, so what? 3. If you want to lube it so it lasts longer why not chip the car to limit it to 100hp then it'll last forever.

 

[Shannow]

In wet brakes and auto trannies, the FMS are there to stop the grabbiness...just like an LSD.

As to the "noise" just being noise, it's actually the winding up of the axle like a spring against the sticky clutch, and the sudden release of that stored energy.

It adds another level of unpredictability in the instantaneous torque at the wheel, so it's desirable that the diff doesn't clunk in that respect.

 

[turtlevette]

You're saying all FM by definition decrease the coefficient of friction. If so I've been misunderstanding it.

Because I thought for example F type transmission fluid had a modifier that increased COF. So you're implying it lacks a FM?

In any case, the term FM in my opinion is used too loosely.