S2000 Talk Discussions related to the S2000, its ownership and enthusiasm for it.

Differentials

Thread Tools
 
Old 12-06-2004, 03:56 PM
  #1  
jfo

Thread Starter
 
jfo's Avatar
 
Join Date: Jun 2003
Location: allentown
Posts: 1,352
Likes: 0
Received 2 Likes on 1 Post
Default Differentials

Will someone with firsthand knowledge please edumacate me or point me in the right direction. I looked in UTH and R&C FAQ and didn't find what I was looking for. I would like to know the difference in the differentials that are on the market for our car. I know the stock unit is a Torsen and I have heard of Kazz and Cusco 1-way,1.5-way, and 2-way. What is the diff? Plus what is rec. for road racing?

Thank you,

John
Old 12-06-2004, 04:38 PM
  #2  

 
JSR_AP2's Avatar
 
Join Date: Dec 2004
Posts: 1,139
Likes: 0
Received 0 Likes on 0 Posts
Default

I have a similar question regarding the differences, pros, and cons of the different types of differentials. I apologize if this is hijacking your thread, but I feel our questions are quite similar (and answering one may answer the other).
What are the differences, pros, cons, etc. of the various types of differentials: Torsen, Viscous, and Clutch?
Is a Torsen (mechanical/gear) differential the same in terms of overall design and operation as a Quaife?
Is one type better in all circumstances than another?...or are there areas of use that are strong for some and weak for others?...(e.g. I know viscous diffs don't react until there is actual slippage, but still work when 1 drive wheel is up in the air whereas Torsen(Quaife?) diffs are somewhat "torque-sensing" and react before actual/full slippage occurs, but are useless if 1 wheel is up in the air).
An example of concepts I'm looking for further explanation on is the TQ multiplication from a Torsen diff, say gears are designed to provide a 5:1 TQ multiplication to the wheel w/ grip, thus if the wheel w/ less grip is up in the air and has zero TQ, there is nothing to multiply and thus the wheel w/ grip doesn't get the additional TQ...my question is: If the gears are always connected somehow, why would no multiplication/TQ transfer occur?...even if 1 wheel is in the air, that wheel is still spinning, which turns some gears in the diff connected to that axle, which should turn the other gears. Perhaps it's my limited knowledge of diff designs.
Thanks.
Old 12-06-2004, 09:10 PM
  #3  
Registered User
 
The Reverend's Avatar
 
Join Date: Oct 2000
Location: Studio City, CA
Posts: 2,560
Likes: 0
Received 0 Likes on 0 Posts
Default

I'm almost certain I did a full write-up on this a few years back, but I'll go ahead and teach this class again...

What does a differential do? Your pinion gear turns the ring gear, which is mounted to the differential. Each of your axles goes into one side of the differential and then the differential turns the axles. Starting with a basic open differential, if you put the car in the air and put the transmission in first (which will prevent the ring and pinion from turning), you can turn one wheel and the other will turn the opposite direction. Reason is, when you plug the axles into the diff the splines for each axle are now locked into a side gear. Between the two side gears you have a planetary gear which basically reverses the direction. Turn one wheel forward, the other turns backward. BUT, if you try and move both wheels in the same direction, you are going to rotate the entire diff... which is connected to the ring gear, then the pinion, etc. So When the car applies power to the differential, what happens is you turn the entire diff housing and usually all the gears in the diff will not move, but the wheels turn.

When you try to turn the car, think of the motion of the car as being circle around a central point. The radius of the circle at the wheel on the inside of your turn is shorter than the radius at the outside wheel. Which means the circumference of the circle is smaller for the inside wheel than the outside wheel. Which means the outside wheel has to turn more times than the inside wheel to go around the circle without tire squeel. Thinking back to when you had the car in the air, you can turn one wheel forward and the other moves backward the same amount. The average speed of the two wheels is therefore zero - the same speed as the transmission was turning. Now, if the transmission is spinning fast enough to move the car 20 mph, when you turn the car, you can now have the inside wheel spin at 15 mph and the ouside wheel will spin 25 mph. The average is the still the speed of the car - 20 mph... just like when the car was up in the air.

Ok, hopefully you're at least kinda following that description of how a differential basically works.

Now, here's the problem with an open diff: when one wheel loses traction, it becomes easier to turn than the wheel with grip. So say you are still going 20 mph around your turn, but the inside wheel loses grip. Since the differential will allow you to turn the two wheels at whatever speed you want as long as the average works out to what the ring and pinion are spinning at, your inside wheel can continue to spin 25 mph while the engine revs way up, spinning the inside wheel to maybe 55 mph. The average speed of the two wheels is 40 mph, so your engine might be revved all the way up to redline in 1st gear now. Problem is, that outside wheel is still going 25 mph, so you're still only moving forward at 20 mph, you're just spinning the inside tire uselessly.



To counteract this effect, you can make a mechanical limited slip differential. Basically, the design is much the same as the open differential, but now you've got some clutch plates between the side gears and the differential housing. Those clutches make it difficult to spin the gears inside the differential. So now the differential basically is less willing to allow a difference in the two wheel speeds. That means that when the car is doing 20 mph around a turn, the differential will fight the wheels to keep them both closer to 20 mph. If you set it stiff enough, it will lock and both wheels will do exactly 20 mph, regardless of the fact you're turning (like a go kart). This will make the tires squeel real loud on a road car, so you don't usually do that.

The amount of pressure on the clutches (and the amount of friction they provide in turn) can be varied. Basically, you put the shaft that goes through the housing and holds the planetary gears in the middle of the side gears so that it rests in two wedges cut into these cam mechanisms that go around the side gears. When you accelerate, the housing will shift one direction, pushing the shaft against one V and spreading the cams apart. There are spring washers on each side of the differntial pushing the cams together, but the torque from the engine will over-power the springs, pushing the cams harder against the clutches. When you're coasting, you don't apply pressure to the clutches, so you don't get much lockup. Under accerlation and deceleration, it locks right back up. Using different angles on this V, you can change the ratio of lockup under forward acceleration and deceleration. You can use varying amounts of friction modifier to control how much overall friction the clutches provide. Lots of friction modifier and the thing will slip almost as much as an open diff. No friction modifier (just plain diff fluid) and it will act like a welded diff.

A 2-way diff has the same type of V cut into the side gears for both acceleration and deceleration. This will make the wheels lock together at the same speed under both acceleration and deceleration. You mainly use this for drifting events on a RWD car. A 1-way diff has a V cut into the side gears for acceleration, but is flat on the deceleration side so it doesn't let the clutches lockup when you're slowing. This is what you use on a FWD car. A 1.5-way diff has a somewhat flattened V on the deceleration side so you still get some lockup on deceleration, but not as much. This is what you use on a RWD car for road racing. I have a Kaaz 1.5-way on my car and use a mixture of Redline 75-90 NS and Redline friction modifier to get the amount of lockup I want. I have about 60k miles on it so far and the clutches barely even look broken in - so don't worry about having to replace the clutches "all the time," like some people say. Just make sure to keep fresh fluid in there.

Kaaz's explanation of the mechanical diff: http://www.kaazusa.com/technology.asp

BMW puts a solenoid in the M3 differential to allow a computer to control just how much it locks up and when, instead of relying on a preset mechanical cam system like a Kaaz.

A Torsen T2, like the S2000 comes with, is basically identical in design to a Quaife, but the casting is of rather low quality on the stock S2000 unit, IMO. Unlike an open diff or clutch-pack, this type of diff has the teeth of the side gears cut along the same axis as the rotation of the diff. It then has 8 planetary worm gears along the outside of the diff to allow it to spin the wheels in opposite directions like the open diff. When you apply torque and the wheels are turning at different speeds, the worm gears fight each other a little and push outward from each other, binding on the housing and locking the diff up a little. But you've gotta have both tires gripping for this to work. so if you lift a wheel around a turn, it will act like an open diff, sending all the power to the inside wheel. Then the wheel touches back down, suddenly there's grip at both wheels and a huge difference in wheel speeds from where they should be. The gears lockup suddenly and can strip at the meeting surface with the housing, breaking the diff.

Here's Torsen's description and picture of a T-2: http://www.torsen.com/files/Torsen%20T-2%2...cal%20Sheet.pdf

I've never taken a viscous diff apart, so I won't try to explain it, but they're not very worthwhile for sports cars, IMO. Too slow to react.

How'd I do?
Old 12-07-2004, 12:05 PM
  #4  
Registered User
 
slick rick's Avatar
 
Join Date: May 2002
Location: la
Posts: 1,633
Likes: 0
Received 0 Likes on 0 Posts
Default

Rev,

you have way too much time on your hands.

If you were working instead of typing, you would be making more money. Therefore, you would afford to do 'proper' racing which means wheel to wheel.
Old 12-07-2004, 02:54 PM
  #5  

 
JSR_AP2's Avatar
 
Join Date: Dec 2004
Posts: 1,139
Likes: 0
Received 0 Likes on 0 Posts
Default

Thanks Reverand. That's explains some...I need to read it again to understand more (and probably again...and again...and again...). I should just swing by your place to have you show me in person...I've always been a pictoral/3D learner and like to see things while getting explanations. I asked about the viscous as my friends G35Cs come w/ viscous diffs.
Old 12-07-2004, 03:51 PM
  #6  
jfo

Thread Starter
 
jfo's Avatar
 
Join Date: Jun 2003
Location: allentown
Posts: 1,352
Likes: 0
Received 2 Likes on 1 Post
Default

The Reverand has spoken!

Thank you very much, very good explaination! I am just a lowly grashopper and I bow to your knowledge oh wise owl.

I should have said that I do understand how differentials work but it is good you put the basic info in to bring everyone up to speed.

My vote is for this to go into FAQ (even though it is more a tutorial) if the mods are listening.

I have 2 follow-up questions:

1) Is there a difference in the amount of power it takes to operate the stock Torsen compared to the 1.5 Kazz and is it significant? Everyone knows about power at the crank and how much you lose in the drivetrain. Do you lose more or less with the different diffs?

2) Can I have an example of when you would be on the power with one rear wheel off the ground. Yes on autocross situations but this is more road racing and I don't think the car will be turned that sharp. I can think of 2 situations one is bouncing over curbs through a chicane like F1 (I don't think I want to try that). Two is the beginning of an accident if you are going so fast and change directions quickly (assuming enough traction) you pick up both the inside wheels.

From what I know the limited slip diff can be used to help the car turn more or help the car go "straight" depending on the amount of lockup. Of course there are many other ways to setup the car that are not as complicated. I think for the experienced racer the Kazz(limited slip) will put the power, and more of it due to less losses, to the ground better once the amount of friction modifier is optimised. I also think that I am not that good of a driver to notice the difference right now so I don't think a different diff is on my list, for now.

Rev, thanks again and with that "spare" time you have maybe you should teach a tech class and get paid.

John
Old 12-07-2004, 04:53 PM
  #7  
Registered User
 
mikegarrison's Avatar
 
Join Date: Apr 2003
Location: Covington WA, USA
Posts: 22,888
Likes: 0
Received 3 Likes on 3 Posts
Default

Originally Posted by jfo,Dec 7 2004, 04:51 PM
2) Can I have an example of when you would be on the power with one rear wheel off the ground.
Inside rear wheel can lift when going through a tight corner. That's a function of your suspension choices. I've watched corners where many of the more stiffly sprung cars lifted the inside rear wheel.

I noticed that Torsen has a Type 2R that uses a spring preload to ensure that there is never an "open diff" situation.

AFAIK I've never had any problems with rear wheel lift on my stock suspension setup and non-R-compound tires.
Old 12-07-2004, 06:36 PM
  #8  
Registered User
 
The Reverend's Avatar
 
Join Date: Oct 2000
Location: Studio City, CA
Posts: 2,560
Likes: 0
Received 0 Likes on 0 Posts
Default

1. No. Keep in mind that when you're going in a straight line, the gears inside the diff don't move at all. The entire differential spins and the axles spin with it. It's only when you turn that the gears inside the diff start to move to allow one wheel to turn faster than the other.

2. If you do a search, you'll find plenty of pictures of people's car's with the inside rear wheel in the air. Inside rear wheel lift is a common problem with road racing the S2000. DavePK and Krazik have had the most problems in this regard because they're running 18" slicks with more grip than the rest of us on R-compounds. But I've even lifted the inside rear on plain ol' mountain roads. It happens.

DavePK and Krazik are running no rear sway bar to try and keep the outside wheel from lifting the inside wheel and using stiff rear springs to counteract that. We've gone back and forth about this, but my thought on the subject is that they're not running stiff enough springs for the amount of grip they have. I think they need to run much higher front rates (to keep the weight from pitching forward so much) and leave the rear sway bar. But that's an entirely different discussion altogether.

(Obscure Airplane! reference: "That's an entirely different discussion.")
Old 12-08-2004, 05:05 PM
  #9  
jfo

Thread Starter
 
jfo's Avatar
 
Join Date: Jun 2003
Location: allentown
Posts: 1,352
Likes: 0
Received 2 Likes on 1 Post
Default

Rev,

Thanks again for your insight. I appreciate it and I hope others do also.

Follow me on this--straight cut gears create the least friction, helical a little more friction, worm gears the most friction. Also, the more parts you have to move the more energy expended in order to move them. Conservation of energy. In a straight line I agree the power is the same because both diffs (Kaaz and Torsen) are locked. In the turn the Torsen is moving more parts in addition to moving the type of gears that create more friction. That is why I think the Torsen robs more power. What do you think?

On the rear wheel lifting. I agree it happens, I just am not sure if it is happening under power. On mountain roads with dips and rises yes the wheel will lift and it could be while you are under power. For track racing I guess I am thinking of a flatter or more gradual flowing track where you wouldn't normally have that situation.


Now if I may hijack my own thread and offer a suggestion about this rear wheel lifting. Please realize that I do not think I know more than you guys that have been racing MUCH longer than me. I just think I may have an idea that you can at least try. Now if the car is pitching forward to lift the rear wheel I don't think you are under power, otherwise the car would be transferring weight to the back not the front but if front weight transfer is the culprit I agree with the Rev--stiffer front springs should be tried. If the wheel lifted is more a function of weight transfer side to side then I would say a stiffer front sway bar should be tried. That will keep the car "flatter". Run the rear bar how you are comfortable. You guys probably went over all that stuff though. I'm thinking more along these lines...........The rear wheel coming off the ground is it a momentary problem like the racing surface is not smooth or is it a "steady state" problem like when you are going through a corner and you have the tire off the ground for a resonable amount of time. If it is a momentary problem you are correct you need stiffer rear springs to push that inside tire down to the ground quickly and also a higher (faster) rebound on the shocks would help. If it is a "steady state" problem again a faster rebound should help. Did you ever try experimenting with the suspension "droop"? I'm sure you guys have race/custom shocks and can hopefully change this. If the attitude of the car is so much that the inside wheel is off the ground it may be that the shock is fully extended and cannot touch the ground no matter what spring is on that corner or the fact that it is disconnected from the other side. I don't propose raising the ride height just increasing the stroke of the shock. That way if you want to run the car the way it is the increased stroke will allow the inside wheel to drop lower and touch the ground. Am I bringing up something that has already been done or way off base? What do you think?

John
Old 12-08-2004, 06:48 PM
  #10  
Registered User
 
The Reverend's Avatar
 
Join Date: Oct 2000
Location: Studio City, CA
Posts: 2,560
Likes: 0
Received 0 Likes on 0 Posts
Default

The gears inside a differential turn at such a slow rate (even during a turn) compared to the ring/pinion or transmission gears. Any "loss" of power is going to be nominal. The difficulty in turning those gears is exactly what makes it send the power to the wheels more evenly. If it were easy to turn the gears, it would be just like an open diff. Think of a go kart - no differential. Yeah, there is some inefficiency in turning both wheels at the same speed during a turn, but in the end it's the effect on the handling nature of the car, not on how much power you get to the wheels that makes the difference. If a diff takes .01 hp less to turn its internal gears, what doese that matter if it's sending that extra power to the wrong wheel?

The rear wheel lifting is steady-state, happening under braking on turn entry and continuing until as late as the apex. Like you said, not usually happening under power application. The last thing Rylan needs is more droop. His car has a ridiculous amount of droop with his JRZ dampers - probably 7 or 8 inches. He simply needs much more spring up front. He doesn't believe me, but he should look at the type of rates the cars that his tires come off of are using. He gets Dunlop slicks that are taken off of ALMS cars. ALMS cars will typically run spring rates around 1500 lb/in. I saw an ad for a Ferrari 355C up for sale the other day. 2200 lb/in springs all around. I understand the concept that you want to run the lowest spring rate you can get away with, but Rylan is NOT getting away with the spring rates he's running. 800 lb/in springs with an 18" Dunlop slick from an ALMS car is most definitely UNDER SPRUNG.

I just bought a set of JRZs myself. I haven't seen them in person yet, but I have seen pictures of them and I'm afraid they might not have installed the droop limiting tubes I asked them to put in. I'll have to see when they're actually here, but I don't really want to have as much droop as Rylan and Richard have.


Quick Reply: Differentials



All times are GMT -8. The time now is 02:32 AM.