Below is an except from a discussion on the merits of the Clutch Type Limited Slip Differential by OS Giken compared to the ZF type Limited Slip Differential. Although this discussion compares the limited slip differentials for the Porsche, the technical information applies to the MINI as well.
Thanks to Erik Johnson from The Race Line for his permission to use his discussion on our web site.
What's the Diff? ZF differentials vs OS Giken
So there has been some pretty good questions here recently about the differences between the OS Giken Super Lock LSD and the more traditional ZF Style LSD’s that are normally found in our Porsche Cars.
First, let’s make sure that we truly understand what is going on in and how the ZF Style LSD actually works. The nature of a LSD is to essentially bind both of the rear wheels together while in a straight line, for better traction, and then to allow the wheels to slip away from each other in tighter turns to allow again for additional stability while turning. Hence the reason why we don’t all run around the track with spools, as they once did in past generations of 911’s.
So how is this accomplished? Well here is how it works. Inside the differential there are several key pieces that allow this process to work. First there are two Bellville washers on either end of the differential which exert consistent pressure on the internal clutch plates. Then there are anywhere from 2 to 8 internal friction plates, which have opposing plane plates to create a differential clutch, and lastly there are two ramps which are engaged by two pins which form a cross.
So when the car is in motion in a straight line, but not under acceleration or deceleration, there is no force being applied to the ramps and the only thing that is keeping the wheels, left and right, connected to each other is amount of preload generated by the two Bellville washers on either end of the differential. The preload will vary from differential to differential and is something that I will go into here in more detail in just a bit.
So here is what happens when you step on the gas. Power is transferred to the ramps by the cross pins in the center of the differential. The cross pins move the ramps outward from the center and apply more force to the clutches inside the differential, causing them to have a greater amount of lock. Now it is important to understand that the amount of actual lock vs slip is going to be determined by both the angles cut into the differential ramps, as well as the number of clutches inside the differential.
There are several different ramps available, but for most ZF 911 applications, a 40/60 ramping seems to be rather popular. Now their numbers are rather vague depending on the manufacture, but it is rather important that for most LSD’s these do not actually represent % of lock up, but rather the angle in degrees that the ramps have been cut at. They do have a correspondence to the overall % of lock up though. For instance a 40 degree ramp cut actually creates more lock up than say a 60 degree ramp due to the amount of torque that is allowed to transfer to the clutch plates, by the ramping affect. The steeper the ramp angle the less torque is allowed to be transferred to the clutches. So you will want to make sure that you know what you are asking for when you say you want a 40/60 differential or a 80/80 and so on. And again I cannot stress this enough, these numbers will vary in their meaning from manufacturer to manufacturer, so make sure you ask.
What makes the LSD special and so valuable in a racing environment, however, is not its ability to lock the two wheels together, but rather to allow them to slip away from each other when the car needs to go around a turn. This is again why we don’t all run solid axles, or spool differentials.
When a car goes around a corner one wheel has a greater amount of speed than the other, which varies depending on the direction of the turn as well as the tightness of the turn. In a shallow turn the speed differential is not that great between the left and right wheel, this speed differential increases as the turn tightens and turn rate increases.
So what a LSD does is allow one wheel to slip away from the other in a turn. Now again how quickly this is allowed to happen is going to be determined in a ZF Differential by the amount of preload, the number of clutch plates, the angle of the ramps, and lastly whether or not the ramps have pressure being applied to them either through acceleration or deceleration pressure generated on the ramps.
So why is this slip important anyway? Why not just keep the wheels locked to each other all the time? Okay, so as I said earlier in a turn one wheel is traveling at a greater speed than the other, or it is at least trying to. If your wheels are locked together and you try to turn the car, you will be fighting the traction of one wheel against the traction of the other. One will have to give way to allow the other to track around the corner. Now this is usually dependant on which one has more traction, although not always, but either way one tire will have to scrub against the track while the other one moves through the corner. This fight between both of the tires to maintain traction on the track, generates instability in the handling of the car, and a ton of under-steer during turn-in, unless the car is thrown into the turn and both wheels as allowed to break traction against the track, which in a 911 generates its own set of problems. So a LSD is great compromise between straight line traction, and stability during cornering.
Now that we have a general concept of how a ZF Style LSD works, which has not changed in almost 60 years, by the way, let’s get into how these differentials can be set up to adjust lock up characteristics, and what happens when a this differential has too much or too little preload applied to it, as I know that the question of preload has been brought up a time or two.
The general stock setup for these differentials is one friction plate backs up to one plane plate, which creates a clutch pack. There are anywhere from 2 clutch packs, (one on each side) all the way to 8 clutch packs, (4 on each side). Preload is setup by the Bellville washers again one on each side of the differential. The Bellville washers meant not only to keep the clutch plates from chattering against each other, but also to allow the ramps to disengage the clutch plates from each other when pressure is no longer applied to them through either acceleration pressures, or decelerations pressure. (This will be important later) Now for those who have never seen what these washers actually look like, they are basically cupped washers that have a spring rate that allows them to flatten out under pressure or return to their cupped shape once they are not under pressure. This pressure is generated by the cross pins forcing the ramps apart from each other.
So in order to adjust a ZF Style differential you have to either change the configuration of the clutch plates, the angle of the ramps, or the spring rate of the Bellville washers. These can all be changed at the same time or in conjunction with each other depending on the desired results. There is also something that should not be forgotten during this adjusting process. There is only so much space inside a differential. This is also sometime referred to as the stack height. You have to maintain a set amount of space in the differential otherwise you will run out of room and the ramps will not have the needed space to spread apart from each other and operate correctly. So I guess what I am saying here is that if you are going to make some modifications to a ZF differential, it would be in your best interest to have a professional builder make these modifications, rather than try to do them yourself. This way if any changes that are made that might have an adverse effect on the stack height this can be addressed before it becomes an issue. This is especially important if you are rebuilding your differential and are going to utilize plates that are aftermarket, as these very often will have a different thickness and friction coefficient than the OEM plates. One thing that can be done however is to place two friction plates back to back instead of one friction plate against one plane plate. This will allow two clutched act as one, as the friction plates will not normally slip against each other once placed together. This would be used if for some reason you were not getting enough slip in the corners, and needed more.
So, let talk a bit about preload, now. Okay, so, preload as we have already spoken about, is the amount of consistent pressure being applied to the clutch plates. What this means is that every time you go around a turn, even when the ramps are not being forces apart by acceleration or deceleration, the clutches are asked to slip against each other to allow each wheel to operate independently, at differing speeds. Every time this slipping takes place a small amount of both friction material, and metal from the opposing plane plate is shed. This is what causes a differential to wear. So the higher the preload the more force it will take for the plates to slip against each other, and subsequently, the more wear will take place inside the differential. Also, something to think about, there more preload there is the more friction that occurs and thereby more heat is also generated inside the gearbox. I think that we can all agree that HEAT is the biggest factor to overall gearbox wear and tear, as well as differential wear. If you get too much heat generated, the clutch plates will actually warp, which causes all kinds of problems. So why is preload important to ZF Differentials? Why not simply lower the preload to avoid these potential issues? Well this is what Porsche thought too, and why in the last generation of LSD’s they actually did lower the preload settings. But here is what they ran into as an unfortunate side effect of not having enough preload.
In a differential with only 8 plates, and a low preload setting, there is not enough clutch material to actually maintain lockup, and you get what I call runaway. Runaway is what happens when one wheel runs away from the other and the plates inside a differential slip when they are not supposed to. Basically there is not enough force generated by the ramps to keep the clutch plates engaged and locked against each other. This happens quite a bit when one wheel has less traction than the other.
It is important to remember that a differential will always want to transfer power through the path of least resistance. If your right wheel is in dirt and your left wheel is on solid pavement, the LSD will try to transfer power away from the left wheel and send it to the right wheel. (The path of least resistance.) What a higher preload does to help keep this from happening is generate more overall torque applied to the clutches when the ramps again engaged. Think of it this way. Let’s say that your differential ramps generate 300 ft-lbs of pressure when they are fully engaged, all by themselves. Now add in another 120 ft-lbs of preload and you get 420 ft-lbs of overall torque against the clutches now instead of just 300. This means, that it will take more force to overcome this pressure and allow the plates to actually slip against each other.
Ah, but there is also a down side to having too much preload besides the wear and heat factors. Because your car is always having to counter the force generated by the preload, what will sometimes happen in a tight turn is that the clutch plates will slip, grab, slip, grab, and slip again. This happens because the car is actually fighting the differentials desire to maintain a constant preload. This can cause the car to exhibit a rather undesirable stability issue. You can actually see this from time to time when a 911 either enters into a turn or tries to exit with a lot of power. Keep in mind that whenever the torque being applied to the differential is higher than the amount of torque needed to make the differential slip, the differential if given the opportunity to transfer power from more resistance to less, will continue to try to do so. So if it takes say 420 ft-lbs of pressure to force the clutches inside a differential to slip, and you have a car that can generate 500 ft-lbs of pressure against the differential, and your wheels do not have the same amount of traction on each of them, the differential will slip until the torque being applied to the differential is lowered, allowing the differential to again maintain its lock up again. Many people will think that it is actually tire slip rather than differential slip, but this is not always the case.
One more scenario to consider, and if you have not experienced this for yourself yet you will probably have seen this happen in some of the in car footage on some race cars. You are going full bore down a straight that has a S-turn complex in the middle of it with FIA barriers. You know that the fast line is to drive over the FIA barrier, (let’s just say) but when you actually get one of the rear wheels off the ground you notice the car is momentarily unsettled. Now this would be expected to some degree right, as you don’t have all of your wheels on the ground any longer. But what also happens is that for that moment in time the differential, especially if it is a bit worn out, will transfer all of its power to the wheel that is in the air and away from the wheel that is on the ground. (The path of least resistance, and more torque generated against the plates than can be managed is the reason for this.) Now when the wheel in the air comes back down to the ground there is a shock load generated against the internals. Do this enough times and I am sure it is not difficult to imagine what happens next.
There is one more thing that I think should be addressed here as well when thinking about how a LSD generates and maintains its lock up. One very important consideration would be in the material used internally to generate friction in the clutches. Friction between the plates, after all is what is keeping them together during lockup.
Depending on the differential manufacturer, the material used on the internal friction plates will vary greatly. Some companies like, carbon, others like brass, and others still like a moly blend. The important thing to note here is what happens when the material begins to shed? It goes all over your bearing and gears, and just about anything and everything. Some materials, usually the coarser, will need to be removed more frequently from the gearbox’s fluid. So keep in mind that the oil change intervals will change depending on the differential.
Now we get to the part that I really like. How is the OS Giken LSD different from all of the ZF Style differentials? Where to start…….
Okay, since friction materials were the last thing that I addressed let us start here. The OS Giken’s differential is most commonly called, a steel on steel differential, which means that both the plane plates and the friction plates are made of steel. The benefit is that there is no friction material to wear away and contaminate the gearbox oil supply. This also means that there is reduced wear on the internals, and longer differential life. The last benefit is that the OS Giken LSD generates less heat, because it has less friction generated in the clutch plates. (So there is one benefit)
The OS Giken LSD also has 24 clutch plates. That is 4 to 12 times as many as you will find in a Zf Style differential. What this means is that this differential has enough clutch surface to maintain lockup under even the highest amount of torque generated by our Porsche cars. It also means that there is no possibility of runaway, which we discussed earlier. The best advantage that the additional plates offer is that this differential generates full 100% lockup on acceleration. (So there are 4 more benefits that the OS Giken has.)
The OS Giken has a progressive lock up effect that is governed by the way its internal ramps are structured. You see while the OS Giken too has Bellville washers on either end of the differential, they are there to keep the plates in place more so than to create preload. (keeps the differential quiet)
So here is where the OS Giken is really special, and quite different from other ZF-Style differentials. It is both in its cross pins, which it does not have, and in its ramps. It is also where the OS Giken is able to offer a greater ability for adjustment.
First the cross pins. The OS Giken has a solid one piece design unlike the two piece design that no matter what anyone else says has been known to fail, from time to time. Anyone ever see what happens when a differential breaks the cross pins? It is not pretty. This is also why you will find that over the years manufacturers, including Porsche have shrunk down the size of the windows that allow oil to flow in and out of the differential. To protect against a broken cross pin from leaving the housing and getting caught up in the ring and pinion. What happens here is when you are on and off the throttle the cross pins will move back and forth against each other over and over again. After enough time, they can become fatigued and fail. I will say this though, it is a rare instance when this happens. Still all and all when it does happen really bad things are close to follow, often times too fast for the driver to make any difference what so ever.
Second, ramps. The OS Giken has ramps not all that dissimilar to a ZF differential, but they are counter sprung against each other. What this means is that when no pressure is applied to them, they automatically close against each other and release the pressure on the internal clutch plates. When you don’t need the differential to work, it doesn’t. The best benefit that this setup offers is, the rate of lock is completely adjustable simply by replacing or changing out the springs and govern the lock up progression rate. So, you want faster lock up go to a softer spring rate, you want slower, more progressive lock up, then you go to a stronger spring rate. Pretty simple if you ask me, which I guess you did, eh? In addition to this, you can still change the configuration of the internal plates to increase slip if you find that you are getting too much lock up effect, which I have yet to hear from anyone as an issue.
There are two other things that I think that you might really like that no one that I am aware of has besides the OS Giken. First, all of the internal gears are made by using forged steel materials. That means the gears, the one piece cross pin, etc. Most importantly though, OS Giken is the first and only company to offer a forged steel, external case, which is then further chemically hardened for strength and durability. There quite simply is no stronger case design currently available, and I doubt there will be until someone goes WAY over the top and decides to build a case out of Ti. Once that happens, I will concede to them and immediately ask for a price, as I am sure it will be WAY up there in the stratosphere.
Speaking of price, the OS Giken is full retail priced at $1990.00 which is far better than most of the competition currently out there. There are certainly less expensive differentials, but none that offer what I know this differential does.
Lastly, and I really like this part. Ask any other company if they will give you a one year warranty on their differential. If you have a problem with the differential, OS Giken will not only send it immediately to Japan for extensive analysis, but they will also send you a completely free replacement differential.
Okay, so I just realized that I almost forgot to address something that the OS Giken has over the ZF Style differentials. It has very low preload. What this means is there is no fighting the differential for control while turning into tight corners. I could go on and on about this too. But I see that I am nearly at 4000 words now and that this has taken me the better part of the day to write. I am a fairly generous person with my time, but I need to try to make some money today too. LOL
Warmest Regards to everyone,
The Race Line
Source: The Pelican Parts Forum Pages