Silver4ringer

First, some background--After getting stuck in my car where the right side was in a snowbank and only the wheels on the right side seemed to be getting power, except when I manually locked the rear, I have researched how the system works.

Thanks to the wealth of knowledge of the people in this forum, I learned that the quattro II system, which all quattros between 88-95 except for the V8 have, has open front and rear differentials and a torsen center differential.

What this means is that power should be transferred between the front and rear differentials to the axle WITH THE MOST torque up to a maximum split of 75/25. However, the front and rear differentials transfer power to the wheel with the least torque with no device to limit slip between the two wheels on the front and the manual lock to prevent slip in the rear.

Okay, now the test--So I jacked up the right side of my car and put it in gear. As expected, the open differentials transferred all the power to the two wheels with no torque (the ones in the air). Unfortunately almost all of that power was transferred to the front wheel. When the car was idleing in first gear I could easily stop the back tire with my foot.

Next, I let both back wheels down and put both front wheels in the air. The torsen center differential should transfer 75% of the power to the wheels with the torque, right? WRONG! The front was spinning fast, but the rear didn't receive enough power to even sway the jacks.

So I was thinking that there must be something wrong with my center differential since the torsen differential "transfers power to the axle with the most torque". So to see if this was the case I did the exact same tests on my brother's 1995 S6 with the same results. When the rear was on the ground all the power seemed to go to the front.

What is the deal? Can anybody give me an explanation as to why a properly functioning torsen differential wouldn't transfer most of the power to the wheels on the ground, as opposed to the wheels without torque in the air? I guess both of the cars I tested could be malfunctioning, but it's kind of unlikely. I would be interested in the results of a similar test on one of your cars.

Another interesting point is that when I would hit the brakes briefly as I was doing this test the differentials would transfer power briefly to the other wheels. This is basically an imitation of traction control. This result is to be expected for the front and rear differentials, but not for the torsen differential. When the back wheels were on the ground and I hit the brakes the center differential would transfer power to the rear. Meaning it would take power from where it thought there was MORE torque (the braking front wheels) and sent it to where it thought there was LESS torque (the rear). The torsen differential should already have been transferring power to the rear, but it seemed to act like an open differential in this case. The power would quickly return to the front as I was only tapping the brake briefly.

I just want to get a better understanding of how the quattro works and why I got the results I did. Maybe my simulation of no torque had some error I don't understand. I would think that when a wheel is in the air it has no torque, but hey I'm not an engineer!

So if anybody has any comments, observations, or answers for why the two cars I did these tests to performed the way they did I will be very happy to learn from you. Also, I would encourage anybody who is interested to do a similar test and see what happens. Thanks in advance.

QCRAZY

..of quattro. One of the drawback is that if you get two wheels to lose traction the car can get stuck. If that is not good enough go for new version, quattro IV. By the use of EDL in quattro IV, the car needs 4 wheels to lose traction before the car can get stuck. This would have helped you out.

Jacking a car off the ground is not a true test of a system. The only way this will get replicated in real-life is if you get highsided....like a snowbank.

Audi's are not SUV's and are not designed to get you out of the snowbank....they are designed keep you from going into the snowbank.

To understand what is going on read this...<A href="http://www.sonic.net/garyg/zonc/TechnicalInformation/TorsenDifferential.html"><B>Torsen</b></a> Torsen is not something simple that can be checked with NO traction. It takes traction to make it work properly!!

DavidFeygin

Even though the front wheels are spinning fast, and the rear wheels are not spinning at all, there may well be 25% of the power going to the rear wheels, it's just that it takes very little power to spin the front wheels when they are off the ground. That would explain why you get power to the rear when you step on the brakes. Most of the breaking force goes to the front wheels, it takes more power to spin the front wheels and therefore you see more power at the rear. Makes sense to me.

By the way, I don't have a quattro.

dave

QCRAZY

...wrong with your statement.

You state that when the fronts are raised off the ground that 75% is going to the fronts and 25% is going to the rears. You go on to state that 25% is transferred to the rears but it is not noticed because the fronts turn faster. This is just plain WRONG

The Torsen differential is designed to do the opposite of what you state. If the fronts are "slipping" then the rears get 75% of the power and the fronts get 25%. This is how Torsen works (with traction). In that previous statement is where the the answer to Blue's "test" is.

We'll let it slide David since you don't own a quattro

Having corrected David, here's what's wrong with Blue's "test" (for those that read, and understand, the article):

Two common mechanical differentials that we know, and love, are: Open and Torsen

An Open differntial allows power to go the wheel that "slips"...which is undesirable. This happens because an open differential canNOT support axles turning at different speeds. ANY differentation between drive axles will cause the low traction wheel to slip. In terms of torque an open differential can be stated as: "The maximum amount of torque conveyed by the drive axles <b>collectively</b> is limited to <b>approximately</b> twice the amount of torque supported by the drive wheel having the least traction." Any torque above this amount will cause slippage of the drive wheels. This is why in a FWD car when your fronts slip you back off the power (lower the torque) until the wheels regains grip.

A Torsen differential has a limit, at which, it can support differentiation bewteen drive axles. It will not support traction beyond this limit. With Torsen differentials this is known as the "bias ratio". This ratio is set by gears selected by Audi (for our cars). On our cars (quattro) this ratio is 3:1. This means the car can send 3 times the amount of torque to the other wheel than which can be supported by the lower traction wheel.

As you may be able to see now, by jacking the car off the ground you defeat the entire purpose of quattro. The system is dependant on the LOWER traction wheel to determine the available traction. With the wheels off the ground the LOWER wheel has ZERO traction (torque transferred). ZERO*3=ZERO at the rear wheels!!

Torsen allows differentiation between wheels and an Open differential does NOT allow any differentiation. Once this limit (bias ratio) is passed, the Torsen differential will behave just as an "open differential" until it falls back within the limit. This diagram gives a good illustration:

<img src="http://www.sonic.net/garyg/zonc/TechnicalInformation/images/Figure2.gif">

The first clue that you violated the system is when you could stop the rear wheel with your foot. At this point you had obviously exceeded the bias ratio and it was behaving as a differential. Obviously??...Because you had 100% of the power at the front wheels when you stated the system is 75% to 25% MAX.

Also, there are no ABS sensors tied in with quattro in quattro II...at least for traction benefits.

If you are ever in the same situation (stuck in the snowbank) you may be able to trick the car into thinking you have traction. This would only work if the rears had lost traction. PULL the e-brake. This will give a "false" torque for the Torsen differential to work against. If your lucky it will be enough torque to pull yourself free!! Give it a try!!

I hope this answers all of your questions. I can try to clear up any hazy points!!

Silver4ringer

This doesn't explain why the back tire was hardly spinning and could easily be stopped with my foot when the both right wheels were in the air. The rear was definately not receiving 25%.

QCRAZY

You read my post below right???

Silver4ringer

I made that other post before I had read your explanation. Your comments explained well the reasons for what I observed in my test.

Thanks for that link. I must admit that most of that article was over my head, but your explanation was easier for me to understand and it makes sense to me why my test wasn't effective in testing the torsen differential. When I read that the maximum torque split was 75/25 I assumed that it was mechanically limited to never giving less than 25% of the power to either the front or rear. In actuality it seems that in a no traction situation the output from the torsen can be 100/0. The discussion has uncovered a few additional questions.

First, you stated in your first post that with the quattro IV system three wheels would have to lose traction. With four-wheel traction control(EDL) wouldn't all four wheels have to lose traction? I think that is what it says in that acticle by Eliot Lim on "Introduction to All Wheel Drive Sytems" in "About Audi" under the "Tech" link on the Audiworld home page. But then again, with my new understanding of the center torsen differential if both wheels in the rear lose all traction the most power the torsen will send to the front is zero, right? Therefore unless the EDL works not only between the left and right sides on either axle, but also between the front and rear of the car then if both rear wheels were to completely lose traction on a quattro IV system then no power would be sent to the front, right?

Next, do I understand correctly that neither of the two wheels on an open differential will lose traction until one of the wheels has less than half the torque of the other wheel? And the torsen is similar except that one wheel would have to have less than 1/3 of the torque of the other?

If the torque of the axle with the most traction is limited by the torque of the axle with the least traction does this mean that giving full throttle on snow and causing a wheel to spin will reduce the torque of the axle with the most traction to whatever 3 times the amount of torque the spinning wheel has, which may be zero in an extreme situation? Thus, even if the front axle has traction if the rear loses traction then 100% of the power could be sent to the rear if there is zero torque in the rear?

Now a hypothetical, is there a reason why a system with three torsen differentials would not perform superior to any system with open differentials and even a system with EDL? The quattro IV system with Electronic Differential Locking (EDL), which I understand is basic traction control, seems to be a reactive system. Meaning more power is not transferred to the wheels with traction until after a wheel slips. Wouldn't a system with torsen differentials in the front and rear perform better?

In the article you posted under section 2.4 it says that a torsen differential with a 4:1 bias ratio can deliver 2.5 times the torque to the drive axles collectively than an open differential. I don't understand how they come up with this. To me it seems that drive axles on an open differential collectively would receive 3 times the amount of torque supported by the axle with the least torque since the other drive axle can receive twice the amount of torque as the axle with the least torque. In other words, where x is the amount of torque in the LOWER axle the total torque output in an open differential would be (x+2X=3x). In a torsen with a 4:1 bias ratio wouldn't the total delivery to the drive axles collectively be (x+4x=5x)? Two and half times 3 is 7.5, not 5. What am I missing here?

I realize that this is very long, but any explanation is greatly appreciated. Thanks a ton.

QCRAZY

I'll try to answer all of your questions, I'll start at the top and work down. Hopefully I won't miss any!!

First off, I'm glad you found my explaination useful. I wasn't sure if it was clear or not when I was typing it!! I see a few things that were misunderstood and I will go back and check the "delivery" of those.

EDL is NOT traction control. EDL stands for Electronic Differenital Locks. While I'm not positive on the inner workings of the system I do know a little about the basics. EDL uses the ABS sensors on each wheel to brake the wheels as necessary for optimum traction. The system is only working under 24mph. With this system it can brake ANY wheel that it needs to to create no slippage. As I stated in my last response you can create a false torque by holding the brake. The advantage of EDL is that it can brake wheels instantanoulsy as needed so that a higher "low" traction vavue is achieved. When done manually the traction will be lower and may not always create enough torque to move the car (in extreme situations). That is EDL.

By the way, ESP is Audi's traction control!!

A Torsen differential can support a different amount of torque to the drive axles. An Open differentail cannot. This is the <b>KEY</b> point. In an open differential the axles cannot support different rotating speeds and still retain traction. I will put it in variables to see if that is clearer:

Open differential:
Lowest Traction (LT axle) drive axle = "A"
Therefore the other (OT axle) drive axle must also have "A" traction. Anymore than "A" traction to the other wheel will cause the LT axle to SPIN!!

Traction Summary:
LT axle - A
OT axle - A

Total Traction = "2A"

Total traction by the axles collectively: 2A....Hence, "The maximum amount of torque conveyed by the drive axles collectively is limited to approximately twice the amount of torque supported by the drive wheel having the least traction."

<pre>If the torque of the axle with the most
traction is limited by the torque of the
axle with the least traction does this
mean that giving full throttle on snow
and causing a wheel to spin will reduce
the torque of the axle with the most
traction to whatever 3 times the amount
of torque the spinning wheel has, which
may be zero in an extreme situation?</pre>

In theory!! However you must realize that once the torque is shifted to the axle that grips the "low traction" axle will no longer have as much influnce on the traction of the system.

<pre>Thus, even if the front axle has
traction if the rear loses traction then
100% of the power could be sent to the
rear if there is zero torque in the rear?</pre>

Yep!! When the 75/25 split is exceeded a "SPIN UP" effect will occur. It will act just as an OPEN differential until you are back within the bias ratio!! This would have to be an EXTREME (bad) situation and the system is supposed to prevent this from happaning before it does. Recall that at any given time there is a 50/50 split. Once a slip occurs UPTO 75% can be shifted to the "gripping" wheels. Once this shift has taken place the "slipping" wheels will be able to regain traction since they have been given a reduced role in the traction of the vehicle.

In theory three Torsen's would be better BUT... There would be great expense in developing THREE torsen differentials for ONE car. Torsen can create problem in cornering if the wrong ratio is used. Also, the friction loses in the torsen differential would be another issue. Using EDL is cheaper!! Audi is using ABS sensors that are alredy in the cars and using them for TWO functions!! Three Torsens would be better but it would be VERY hard to design, expensive to build, and OVERKILL!!

I discussed this ealier but I will do it again to answer your last question...although I think you know the answer this time!!

Open differential:
LT axle - "A"
OT axle - "A"

<b>Total - "2A"</b>

Torsen Differential:
LT axle - "A"
OT axle - "4A"

<b>Total - "5A"</b>

<b>5A/2A = 2.5</b>

I think I got them all. Let me know if I missed any questions!! I hope I didn't create more confusion than I alleviated

Silver4ringer

First, did we conclude that in the Quattro IV system that 3 or 4 wheels would have to lose traction to immobilize the car? I was thinking the EDL could transfer power to the one wheel with grip even when three are slipping, but you mentioned that only three would have to lose grip to immobilize the car. The third paragraph of section 5 of the article by Eliot Lim says all four wheels would have to lose traction. Is that incorrect? Also he talks about 4-wheel "traction control". He's talking about EDL and not some earlier set-up with ESP, right? What does ESP stand for?

And the final question, on the performance curve graph you posted in your explanation it looks like the Y-axis value (total tractive effort) on a torsen differential with a 3:1 bias ratio is about 4 times the x-axis value (surface coefficient). This makes sense because x+3x=4x. The question is why the curve changes approximately where the surface coefficient is .3. Is this because with this much traction on the lower torque axle the maximum torque split (75/25) is not achieved (not necessary)? I think that's right, but can you confirm that?

QCRAZY

EDL, again, is Electronic Differential Locks. In my last post I wasn't clear in stating that it simply acts as differential locks. If you can lock both the front and rear differential you force the wheels on each axle to be "connected" together. This will eliminate "spin-up" of any one wheel and forces BOTH on an axle to turn at the same rate. This means that for the front (or rear) axle that both wheels have to lose traction before the car will stop moving. If both axles require each wheel to lose traction...four wheels need to lose traction before the car will get stuck!!

Elliot is correct. EDL is NOT traction control. ESP is the new system and it IS traction control. ESP-Electronic Stabilization Program. Actually the new quattros (2000+) are equipped with EDL and ESP (optional on some models). ESP uses YAW sensors (and the like) to further manipulate when to apply ABS and to what wheel(s). I imagine the systems are quite impressive when combined!!

The 0.3 (0.29) is the lowest traction that can be split effectivly at the 3:1 bias ratio. Theoretically it would be .25 because of the 75/25 split but they are taking friction loses into account to arrive at 0.29. The 6:1 ratio shows a loss of "effective" traction at about 0.16 when idealized it is closer to 0.14. How they take into account the friction losses are discussed in Section 3.1 of the article and depends on the arrangement of invex and side gears.

The drop of in traction occurs on the graph because the Torsen differential canNOT fully transfer torque below the indicated traction coefficient. This means you will have to "let off" the gas to prevent spin-up below the indicated traction coefficients.

HTH!!