...the 4:1 bias center Torsen will absolutely help your car's cornering performance. Most of the time, you won't be able to tell the difference between the 2:1 and 4:1. But, during heavy throttle corner exit, it is very apparent.

We've run two autox's and one track event at Sears Point with a 4:1 center diff in our car (we actually got ours from Champion Racing, but what Stasis provides is identical). The thing I noticed immediately was that the car had much less corner exit understeer when heavy on the throttle. You know when you get a really good run out of a corner, and you're fighting the steering to keep it on the track (or from hitting a cone)? With the 4:1, it wasn't even a challenge. And, what that means is that you can get on the throttle a lot earlier, and carry more speed out of the corners.

I never tried a locked center like 4kq mentioned, but I've heard some good things about them. Although, I'm sure their advantage is limited to certain traction situations (like if it's dry). Since the Torsen is dynamic, it ought to work in just about every situation (except if you lift wheels).

Chassis/suspension tuning is very important when looking at this whole package. Changing the bias ratio won't do much for getting your car to rotate while off throttle or at neutral throttle, but huge gains can be had while on throttle. For off-throttle stuff, I'd concentrate on spring rates, swaybars, shocks, and alignment.

Now, lifting a wheel anytime while on the throttle will take away from the Torsen's performance. Just as all Torsen documentation states, it will not transfer torque anywhere when one side of it sees zero torque resistance. Depending on how you use the car, what your grip levels are, and your state of suspension tuning, this could affect the value of a higher bias center Torsen. In the case of Sharon's S4, it is very critical, since autocross probably sees more weight transfer and faster weight transfer than any other road race event. We also run sticky Hoosiers, and see up to 1.3g's of cornering force, therefore increases the chance of lifting a wheel somewhere. To compare to how the car does on a racetrack, with the same setup, we will see very little (if any) wheel lift issues.

To get more detailed, the stock 2:1 Torsen can split up to 67%/33% to either the front or rear of the car. And, anywhere in between. It all depends on how much traction you have at each axle which determines the split. There is no nominal split, it is simply dynamic, 100% of the time. If you don't have enough power to break any tires loose (let's say you have a 4000lb car with 100hp), then it'll probably be somewhere near 50/50, but if any wheel has a reduction is traction, then the split will change. I will only refer to corner exit acceleration, even though this applies to acceleration while cornering at any time, because that is the most difficult time to get power on the ground.

Ok, let's back up a little more, when braking just before corner entry, you transfer weight to the front, therefore making the weight bias of the car worse (stock is around 60/40, now you could be around 80/20).

Template:
LF-RF =total front
LR-RR =total rear
L%-R%

30-30 =60
20-20 =40
50-50

to a:

40-40 =80
10-10 =20
50-50

As you turn left into a corner and get to steady state cornering, you gradually go from a 80/20 front/rear split, back to a 60/40 split, but at the same time your left/right split goes from 50/50 to perhaps 30/70.

18-42 =60
12-28 =40
30-70

Ok, now let's say you transition from this steady state cornering to full acceleration corner exit. Now your weight balance might look like 30/70 F/R and 35/65 L/R.

10-20 =30
25-45 =70
35-65

I must admit that these numbers are fudged, but I don't think they are that far off. I am basing them off the power of the S4, however. Ok, just from looking at the total weight distribution front/rear (which is 30/70 right now), you can already see that you can utilize more than a 2:1 bias Torsen. And that's not the end of it. Look at that LF corner of the car, it only has 10% of the car's weight on it. So, the Torsen sees that you only have 10% of the total traction on the front axle (due to the open diff), and can multiply that by 2 and put enough for 20% traction in the rear. BUT, you see that you have the ability to handle 25% traction in the LR. Conclusion, at this given instant, a 2.5:1 Torsen would be sufficient. But why stop at 2.5:1 when there are most likely other case where the traction limit bias from front/rear could be higher? Might as well put the 4:1 in there to handle those situations, cause they do occur with extreme driving.

This whole example was greatly simplified, and assumed that you have an open differential in the front and rear of the car (which all A4/S4's have). I actually have no experience with just changing the center Torsen, since we did that after we changed our rear diff to an LSD type diff, but in theory all this should work out. We have been able to use these basics and successfully apply them to our car. On the 3min autocross we ran with it, we probably saw a 2 to 4sec improvement in lap times compared to the competition. The car still doesn't throttle oversteer in high traction situations, but the throttle understeer was hugely reduced. As we increase our HP levels this year, I think the higher bias center Torsen will help us even more.

Corey S