LSD Shims

[visionaut]

Hi all,

Anyone know where I can buy the shims (aka discs) to alter the % lockup of my 3.91 ZF Limited Slip Diff? I'd like to go from the stock 25% lockup to 40% by adding additional shims. I've heard they're NLA from BMW. I've also read shims from a ZF diff that was used in some Ford Escorts use the same parts (anyone know for sure?)

I beleive the shims come in pairs - one round and one fingered per pair. I also think I need to install them symmetrically, i.e if I use 2 sets on the right, I need 2 sets on the left of the gearing in the center of the LSD unit. (that's the way I read the diagram, right?)

The parts are identified as #s 8 & 9 in this realoem diagram. Per the parts listing, it shows three different Outer Shim sizes, 1.9, 2.0 & 2.1mm - any idea which combo I need?

Any help is appreciated. I'll be rebuilding my LSD soon, new bearings/sels, etc and would like to do this mod at the same time.

TIA!

Tom (aka visionaut)

[wurlyvalve]

To increase the locking of your diff you need increase the number of friction faces in your pack and/or reduce the ramp angle. The latter requires some machining. The former may require machining.

Build the pack up with equal number of plates side to side. Your ramps (part 10) are keyed to the diff case so you can use an inner drive plate against this (but it will wear your ramps out eventually, although in my experience it is acceptable wear). Pack an outer drive plate against every inner plate to maximise the number of friction faces. To pack more friction faces in fit the thinnest plates you can. The bellevile spring on the end is used to adjust preload on the diff and alter the progression zone of the locking. Whilst the spring is still compressing the diff won't lock aggressively. So try not to crush the spring with extra plates. Instead either get new ramps cut (shallower angle for more aggressive locking) with deeper cuts to increase the space for more plates. Or surface grind the ramps to thin them and increase the pack space that way. Tuftride the ramps afterwards. If you reprofile the angles ensure that the cross pin is matched in profile (or is round).

You may want to make your own plates. Would have thought EN24 should be good enough, treated. Texture the surfaces though. Inner plate with a spiral and outer with spokes. Failure to do so increases hystersis of diff unlocking due to the perfectly smooth discs being "vacuum stuck" together once compressed.

Whilst I haven't worked on the BM's ZF diff, I build for a living torsen plate diffs for SWRT. Hence my particular knowledge of the cross pins is lacking.

Cheers,

Glen.

[visionaut]

Thanks for the tech info. I'll try to find the thinnest Outer shims I can find. However I know that it's possible to get from 25% stock to 40% lockup with just the addition of shims - no machining, and no additional clutch plates. I know another alternative is to go after the ramp angles, but I'd prefer not to do that. I also researched that MM increases lockup by adding clutch plates - but that requires machining too - again not my preference.

Still the original Q - where to purchase shims/discs? Doesn't anybody still stock those part numbers?

[Lee]

Maximillian has them. Or if in a hurry, Jim Blanton may help.

[visionaut]

Thanks, Lee.

Any idea of the combo needed for 40% lockup?

[wurlyvalve]

I, and I'm not alone, really don't like the ZF notation of %lockup. A locked diff requires an infinite amount of torque to make it slip. So what is 40% of infinity?

Placing thicker shims to achieve 40% will only effect the static preload. This may infact be what ZF means by 40% ie the breaking torque is 40% of fully loaded torque. If that is the case is 40% an improvement? It'll reduce turn in and that is about it.

Ok, take one driveshaft output flange and place in a vice, put diff onto it. Place the other driveshaft stub axle (or o/p flange) in the diff. Weld a 1/2" drive socket to a steel plate and bolt it directly to the stub axle that is not in the vice. Using a torque wrench in the 1/2" socket slowly increase the torque until the torque wrench stops "clicking out" and the diff body starts to turn. Assuming that your 25% diff is in good nick then this represents 25% of locked. So multiply by 4 to give the fully loaded locked torque. Then work out what 40% is (ie loaded torque divided by 10, multiplied by 4). Then add thicker shims to achieve this new figure.

We never build rear diffs with more than 60lbft of preload - as a guide.

Still a little arse about tit, most other diff manufacturer quotes in ramp angles and preload.

Cheers,

Glen.

[winstontj]

Thanks for the tech info. I'll try to find the thinnest Outer shims I can find. However I know that it's possible to get from 25% stock to 40% lockup with just the addition of shims - no machining, and no additional clutch plates. I know another alternative is to go after the ramp angles, but I'd prefer not to do that. I also researched that MM increases lockup by adding clutch plates - but that requires machining too - again not my preference.

Still the original Q - where to purchase shims/discs? Doesn't anybody still stock those part numbers?

Dave Varco @ Aardvark also does this for a reasonable price. He has also been very willing to talk about this stuff every time I've asked.

He knows the torque #'s and it was something like one clutch pack is equal to exactly (ir very close to) 25% so one = 25%, two = 50% three = 75%

The factory manual has all the specs or anyone you can buy the discs from will probably know.

[bmw02rally]

Glen,

Hows the car coming?

This Mark who you bought the Ansa off!!

I'm doing my first rally with the Quaife in on the 25th of March so will tell you then which I prefer.

Regards

Mark

[Guest Anonymous]

. .....However I know that it's possible to get from 25% stock to 40% lockup with just the addition of shims - no machining, and no additional clutch plates. I know another alternative is to go after the ramp angles, but I'd prefer not to do that. I also researched that MM increases lockup by adding clutch plates - but that requires machining too - again not my preference.

You are saying that you don't want to add clutches.

But the "shims" you point to _are_ the clutches.

wurlyvalve's contributrions above are some of the best technical posts this board has seen. Thanks Glen, for the great write-up. Here in the USA, the diff builders I know produce their own plates, as you had suggested.

[wurlyvalve]

Cheers dude. I think information should be distributed - whole point of the internet! I try my best, any mistakes (and there are many) are unintentional. I build diffs, driveshafts, gearlevers and gearboxes for a living. Driveshafts bore everyone, gear levers are trivial, gearboxes bore me but diffs..... rocks my boat!

Mark. Good to hear from you buddy! Keep in touch pal. The car is coming along now. Got fibreglass boot and bonnet ready to go on. Put on a 4-2-1 but an equal length one (tuned length about 40") - goes like a rocket now!!!!! My offer still stands - whichever diff you don't want I'll pay the right money for. If you want your ZF sorting just let me know I can get contacts from work to do the machining (with P diddy discounts).

Also if you want a spanner crew for your rally let me know, a lot of the guys get bored with working on those crappy Scoobs and for a shade more than beer money will happily come out for a weekend to help out. Looks good on CV's you see And if you knew how little the WRC crews get paid for day money.....

Regards 40% diff. Why copy ZF's settings? Fix the diff for your cars requirements. If you need less turn in tigthen the preload. If you want more on throttle tail out pleasure change the accel ramp to a lower angle. Want more tail out on throttle and more stability in engine braking increase FF (friction faces). The only restriction is the BM's diff casing is minute (check out Quaife's website for tech drawings on diffs) so space for more FF is VERY limited!

My opinion is that the BM needs a torque sensing (which the ZF is) diff not a speed sensing one. Torsens produce less oversteer at low speeds (when inside rear wheels are off (or nearly off) the ground) and more turn in at high speeds (and frankly my BMW understeers at high speed). So more preload is not an option I'd personally use with 2002. Although if you lower the front RC (roll centre) and raise the rear RC?????????

Seriously, anyone need guidance with slippers I'm more than happy to give what info I have. If mine is lacking I can ask around - I work with ex F1 diff designers!!!!!!!

Cheers,

Glen.

[PatAllen]

i sucessfully modified mine and increased the sliping factor up to 75%.

I added 2 set of clutches and modified/machined all the necessary parts.

I bought the clutches from Max import.

hth.

[visionaut]

Thanks, guys. I've followed all the tech info - great stuff.

(I wasn't belittling any of the input, BTW, I don't know how that got construed J/. Clearly the input/info was from someone way more knowledgeable/experienced that I.)

OK, so there are no shims that are different than the clutch discs, it's just clutch discs one can add...

I had assumed they were something different, as I'd read you can add shims and not need machining, but adding clutches involved machining the housing (?). As I'm not looking to do any machining - either to change the ramp angles (which seems to be the best approach if one was looking to really get in to tailor/tune the engagement/disengagement performance of the diff), or to add additional clutches - maybe I should just leave my LSD stock. It seems to work fine - I was just looking to get a little additional perfromance if it was an easy mod while 'in there'. I thought it was just a parts addition one could do in the rebuild - and if was as simple as adding a few shims, I'd go for it. (Also wasn't looking to fab my own parts, but to buy 'em).

The info I had gathered prior to my post told me I could add shims to increase/alter preload (and that too much preload would increase understeer). It said the shims come in pairs - one being a Belleville spring 'as big as the discs', and are approx 2mm thick and the other being the dog-eared plates to shim up/increase the static locking. I saw that BMW/ZF uses the same angle for the coast and power ramps to drive the pinion shaft (arrow in pic), and shallower ramps and/or altered ramp geometry would increase the mechanical advantage increasing the lockup %. (and that increasing lockup this way, vs. just altering the static preload is really the better way to go.)

As for the differences I'd get between 25% and 40% lockup, I looked into the engineering to try to figure out what it might mean as far as driving the car. (I grok the ZF way of citing the LSDs perfromance - 25% LSD meaning 25% of the torque applied to the faster wheel is applied to the slower wheel). With the 25% lockup LSD the high torque side provides as much as 62.5% of total, while low side could be as little as 37.5%. This yields a bias ratio of 1.67 (1.67 times more torque is applied on the high side). A 40% LSD would have a 2.33 bias ratio. It would increase rear tire wear on the inside wheel during cornering. It would increase understeer in tight corners in the dry, and increase oversteer in slippery conditions. It would get me less inside wheelspin accelerating out of corners and more power being translated to the ground in that situation. This seemed like a reasonable tradeoff - and if I could get it 'for free' (cost = an additional set of "shims"), it looked like something to go for.

If I can't increase the % lockup without machining (i.e. if adding any additional discs beyond the default included in the stock 25% LSD config would necessiate machining the housing to fit them) I guess the 25% limited slip will do as is.

I appreciate the info - I learn from the knowldege of others here, and try to share things I know too. It's great to know if I wanted to add discs, I could get them via Maxmillian or Aardvarc. Also good to learn that if I really was interested in altering/tailoring my diffs performance, machining and the use of an expert diff builder is required.

Tom (aka visionaut)

[wurlyvalve]

Cool, try the thicker bellevilles. That will help a little. It'll increase preload and also make the diff a little tighter in the progression zone. Worth a go. You may need thinner plates/shims to unlock the preload a little bit. Give it a try dude, play time!!!!!

Glen.

[visionaut]

Glen - I might try that.

Do you happen to know the number of disc pairs each side in the stock configuration, to achieve the 25% lockup, for this diff? (i.e. is it one #8 & one #9 each side as shown in the realoem diagram, or is it possibly two or three sets per side? I'm uncertain, as whenerver one sees a pic like the last assembly pic I posted, it looks like multiple sets per side.). My diff was originally from an '83 320i Sport.

Secondly, should I assume the "Outer Discs", #9 in the realoem diagram (which look to be the dog-eared parts, and not the Belleville springs, BTW) are the 2.0mm size in the stock configuration? If the Bellevilles are the #8s, then it looks like there's only one thickness (unspecified), true? So rebuilding the diff with the thicker 2.1mm #9s is what you're suggesting could be done without machining to tighten/increase the preload/static lockup some?

Let me know - and thanks for the experienced input!

Tom (aka visionaut)

[wurlyvalve]

Tom mate, no idea how many plates you'll get in per side. Sorry, never worked on a BMW ZF diff. Outer disc are dog eared shaped. However, just to add major confusion some bellevilles are also dog eared - others aren't. Looking at that diagram no 6 is the belleville. Thicker plates (ie no 9's) will increase the static preload, yes, and can be done w/o machining. It will help you when you've lifted an inside rear wheel. But it will reduce the turn in (or increase turn in understeer). I have seen bellevilles doubled up before. And also we have played with having the bellevilles re heat treated to change the springing. I do have a selection of bellevilles available to me but you probably wouldn't want wait for them in the post.

Cheers,

Glen.