PDA

View Full Version : Diffs and wheels response from chassis Engineer working for Audi check last paragraph



ernysp76
03-11-2010, 07:28 AM
Thought I would attach a response I got from a person who is a suspension and mechanical engineer with Audi:

"I was having a look at the type of differential you were talking about and I could not remember the exact type the Quaife differential is, so looked it up. It is essentially the same as the Torsen T2 differential. Similar to what my Audi has as a centre differential. You were thinking of putting it in the front to replace the open differential, so had a think about the other differentials in the car. The existing front differential is open conventional and will split torque 50:50 even if a wheel is spinning both will have the same torque, so torque transfer to the surface is still happening. Car weight force may not be equal and the torque will be equivalent to the dynamic friction times the weight force on the wheel. Dynamic friction is the lowest and the weight force on the wheel is also the lowest. Speed times torque is power so the spinning wheel is sapping power from the drive train as a whole. If you notice the front wheel spin reducing overall acceleration, it is the centre differential not being up to it, or rather the limited slip part. The centre differential requires the slip from the front to allow increase in torque to the rear [this is why most AWD systems have open front differentials]. This system is reactive to a wheel spin situation, as in it can only act after the spin has occurred to transfer torque rear ward. There is a reaction time for the fluid to heat up and bind the plates, plus the binding is a gradual process then a locking to defined slip ratio. There are many variations on the locking method and degree it is difficult to know which. Also this system can wear out resulting in near open differential capabilities. You will always have a 50:50 torque split f:r no matter what and if one wheel spins all wheels suffer and only to a lesser extent with the viscous centre differential. I have driven many different Subaru AWD systems and over uneven surfaces you can feel the difference in the centre differential and the noticeable superiority of the Audi quattro with Torsen centre. Although the difference is marginal in on road driving over the best Subaru system I have driven [not driven the active Subaru centre differential] the difference come when putting torque to the road to accelerate, corner, and negotiate uneven surfaces. The better centre transfers torque and removes the jerkiness when accelerating over uneven surfaces. The additional smoothness comes from torque transfer happening pro actively without wheel spin. If one wheel on an axle reduces or loses traction the torque is smoothly moved to the other one without wheel spin. The difference is that previously one wheel's loss of traction compromised the drive train, now one wheel on each axle must concurrently be compromised to compromise the drive train.

Now the difference in traction between front and rear is altered by weight transfer. Meaning the front may only have 20% weight force during a first gear acceleration worst case. Then add cornering affects and it may be possible to get under 10% traction to the inside front. The front will experience weight forces during acceleration from 20% to 60 %. The 60% assumes neutral weight balance at the initial stages of acceleration. This means the rear has 40% to 80% of the acceleration duty and most of the time will be in the 60% range. Since and a standard differential has a torque split of a fixed 50:50 modified slightly with limited slip it makes more sense to put the Quaife torque sensing differential in the centre. This is because it actively transfers torque and can do so to the maximum ability of the torque bias ratio after which it behaves like an open differential! A 20:80, f:r torque split without going to open differential mode is 4:1. Of the total 5 torque units one goes forward and 4 go rearward. TBRs of 5:1 are available in Torsen T2. Audi use 2.5:1 to 3:1 on most of their normal cars, on my car they used 4:1. On the next version six years later they used a Torsen T3 with 35:65 f:r biased split and a 3:1 TBR. The pre-set f:r bias improved the car feel for more RWD handling. On the latest Audi S5 and RS5 they use a crown wheel differential that is different to the differentials discussed in arrangement but not in function. It still has a passive f:r bias of 35:65. I believe it offers computer control of the centre so it can shunt torque when it thinks it is required but still has the passive split.
Caveat on the above: if you are running on loose surfaces all the time, a locking or locked centre differential is probably a cheaper choice and will do a similar job albeit the front must slip to send torque to rear and rear must slip to go around corners, because the front always travels further than the rear in a corner. Typical TBRs for front differentials 1.4:1 to 2.5:1, and rear differentials 2:1 to 3:1. Think of TBR as if one wheel can handle 1 unit of torque the other wheel or axle will get the ratio times more. If a wheel has zero capability then 0 times TBR is still 0!. This style of differential in the centre is compatible with left foot braking, but am not sure what the results would be. Most likely for shutting down the front spinning wheel rather than getting the back to slide. With more torque in the rear you may not need left foot braking.
With the Quaife centre the front will not spin until the TBR is exceeded and more torque will be sent to the rear improving acceleration. The rear viscous differential will handle the inside wheel spin.
The next differential to replace with a Quaife is the rear. It will send more torque to the outside wheel improving steering response by putting more torque outside and pushing the car around. You would move the rear viscous LSD to the front. The ultimate solution is to have TBR differentials all round and maybe a pre-biased centre. Hope this helps on your decision on how to fix the front end wheel spin problem.

If after the modifications you still have problems lowering the centre of gravity will help reduce weight shift. Maybe the front roll bar needs to be less stiff so as not to lift the inside front. Lower mass wheels help with ground contact and acceleration. Most of the time low profile tyres hinder acceleration and ground contact by increasing the mass. Only supplying the benefit of improved turn in response and feel, which in itself is no real gain to a racing driver who can easily compensate for these problems. This means anything less than 45% profile is a waste of time literally. A few Kg saved in wheel / tyres combination is equivalent to 10s of Kgs on the chassis. Where the mass resides on the wheels radius is also important. Spending money on wheels has the greatest benefit of any power/performance modification you can buy. It will out perform any chip or exhaust modification by more than 2:1."

zero
03-11-2010, 11:04 AM
Interesting read, good stuff!

Blackstar
03-11-2010, 06:24 PM
I can attest to the superior benefits of a quaife.

Completely transforms a car.

lathiat
03-11-2010, 06:31 PM
I can attest to the superior benefits of a quaife.

Completely transforms a car.

Ditto.. I honestly don't think I could live without it now. If I got another FWD car it'l be my first mod I think... in an AWD setup well thats a bit more complex :P

ernysp76
05-11-2010, 08:55 AM
Further to the above information I recieved this today:

I did notice that the Quaife web site does not mention torque bias ratio. It is a very important factor in choosing the correct ratio for the torque distribution capabilities.

I have equations that will work out the weight transfer if you know vehicle mass, weight distribution front to rear and the centre of gravity.

Lf = dG + Bh / w, Lr = (1 - d)G - Bh / w

d=weight distribution
G=weight force, Newtons, mass times g [9.81]
B=Brake force
h=height of G
w=wheel base


With this you can work out the Load front and rear. Make B negative for acceleration. Force is axle torque times the distance to the ground.

There are some interesting YouTube videos with Quaife rear differentials, one with two nearly identical BMWs in the snow. The good thing about the torque sensing differentials is they are compatible with open differential stability control programs offering a slightly improved torque to the ground, especially on differential friction coefficient surfaces between left and right. Not so for some LSDs!

Disciple
05-11-2010, 09:20 AM
I love reading this sort of stuff. It's like figuring out how ACD/AYC works, lol.