S3 how much power to rear wheels?

[Sammyfat]

Hi
I would like to know how much power/torque can go to the rear wheels in the s3 when the fronts slip ? I googled it and it said 50/50 but I thought It was more? Or is it different again in normal Quattro models?

 

[Lewis583]

On haldex type AWD. Max that can be sent to the rears will only ever be 50%. Owing to the drive for the rears is taken from the offside output shaft from the gearbox.

The torsen Quattro as found in longitudinally mounted Audis (a4/a6/a8 etc) uses a completely different system.

 

[Sevv]

lastest gen haldex goes up to 70% alongside the Golf R

 

[Sammyfat]

Thanks guys.

 

[Sevv]

Thanks guys.

your welcome mate

 

[veeeight]

This old chestnut again


Before we start, its important to note that in the same way some Hi-Fi manufacturers refer to "peak" watts in marketing, the true measurable and comparable value should always be RMS Watts, so Haldex refer to "Torque" in marketing terms as "applied torque against a surface", rather than the more generic definition of Torque.


The Haldex unit is a multi plate clutch device, capable of locking together the front (axle) drive and the rear (axle) drive. It is not a differential. Conceptually, it just joins the front half of the propshaft to the rear (propshaft).



1. When engaged, can the Haldex ever change the 50:50 drive (front:rear) to be biased more to the rear?
A: No. The Drive will always be, 100:0 (in the case of Haldex disengaged, FWD), or 50:50 (in the case of Haldex engaged, AWD).


2. Can the Haldex ever cause the rear wheels to turn faster than the front (assuming equal friction surfaces for all tyres)
A: No. Haldex can never make the car into a power-oversteering monster (assuming equal friction surfaces for all tyres)


3. Can the Haldex "send" more than 50% of the torque to the rear?
A: Yes (in Haldex marketing terms) eg: Front tyres are on a skating rink, rear tyres are on grippy tarmac.



In Haldex world, you can't "send" torque. Torque in Haldex speak is what is being transmitted from tyre to ground.

In the skating rink example above, the rear wheels will not be rotating faster than the front (ignoring the cross-axle diff), as the drive front:rear will be 50:50. But as the rears are gripping more than the fronts, then more torque is being transmitted by the rear wheels, than the front, and this will give you movement.

In other driving situations and scenarios, you can set up the car to get all 4 wheels spinning and drifting, and yes, through driving and momentum, you can hang the back of the car out (for a period of time) - but on a dry grippy road, assuming equal traction at all 4 corners, most of the torque will be transmitted by the front tyres before it even thinks of going to the rears.

Thats not to say that the car will always feel like a FWD car. Through careful setting up of the corner, you can utilise the fact that the rears are going to take a shorter path than the fronts (and turning slightly slower than the fronts), and you manually engage the Haldex to be ON, you can force the rears to turn at the same speed as the fronts, and you will "feel" the rears pushing.


So. If you go by the definition that you define torque by what is being transmitted between tyre and ground, then yes, with a Haldex Unit, more torque can be available at the rear tyres than the front, depending on the friction (mu) of the surface.

However, with the Haldex engaged, you can never have more than 50:50 drive (front:rear), and the rear wheels can never (in a straight line, equal mu surfaces) be made to turn faster than the front wheels.



The following video illustrates torque transmitted at the tyres, as seen by Haldex.

 

[Sammyfat]

I did try to search for the answer.... Thank you very much though, totally understand now.

 

[Sevv]

i have first hand experience with that taking the Golf R to the race track how would you explain that in drags, we were chewing through toyo r888s' (rear tires) at a quicker rate in fact they were being chewed 3x as quick as the front and thats with the stock AWD system, just saying... here is a much more thorough and detailed view of the Golf R audi S3 haldex setup and the quattro set up in longitudinal A4 etc..

have read you, i can assure you'll be interested

Can someone explain the FWD Haldex system to me? - VW GTI MKVI Forum / VW Golf R Forum / VW Golf MKVI Forum / VW GTI Forum - Golfmk6.com

 

[veeeight]

I don't know how much more simply I can put this for you:

It is impossible for the Haldex system in the VW Group vehicles, to deliver more than 50:50 drive, front:rear, when it is engaged.

If my rear tyres were wearing faster than my fronts, I would consider suspension set-up, camber, toe in/out, pressures, etc., before jumping to the conclusion that the car suddenly became a power-oversteering rear wheel drive car.

Besides which, as I alluded to above, if you drive around a track with the Haldex permanently switched on with soft tyres, due to the fact that the rear wheels are naturally going to take the shorter path around corners, but you are forcing them to drive at the same speed as the fronts, you will get an amount to tyre slip/scrub, and yes, this will lead to increased wear on the rears.


Again: It is impossible for the Haldex system in the VW Group vehicles, to deliver more than 50:50 drive, front:rear, when it is engaged.
(This is different to saying that up to 100% torque can be available at the rear driven wheels when the front wheels slip)



Here are some graphs from the HPA Haldex Performance Controller. In every setting (Stock, Sport, Race etc.) - the line showing the power at the rear wheels NEVER exceeds the front wheels. Click on the graphs - they show a Golf on a AWD dyno. Bit more scientific than heresay and forum chatter.


HPA Motorsports, Inc. - Haldex



(Yes I did read the link you posted - if confirms what I am saying)

 

[Sevv]

the set up was perfect so there is no other explanation, i wasnt talking about torque transfer either whilst its only 50:50 in relation to power transfer, all four wheel will be spinning the exact same speed but actually more torque to the rear is being achieved. and thats what that post talks about

im talking of drags not going around corners either.. anyway its a good informed discussion for this forum so they know exactly what the haldex system is not quattro..

 

[steeve]

What is the effect of the front tyres being slightly more worn than the rears which is the normal case? Would this not cause some slippage of the rears as wheel rotational speed will not be the same?

 

[veeeight]

The majority of the time, the Haldex clutch pack will be disengaged - so the car is normally in Front Wheel Drive.

Only when the clutch pack is engaged, will the front tyres/rear tyres diameter matter, then yes, if you have got differently worn tyres, some slip/scrub will occur.

 

[scherzkeks]

This old chestnut again


Before we start, its important to note that in the same way some Hi-Fi manufacturers refer to "peak" watts in marketing, the true measurable and comparable value should always be RMS Watts, so Haldex refer to "Torque" in marketing terms as "applied torque against a surface", rather than the more generic definition of Torque.


The Haldex unit is a multi plate clutch device, capable of locking together the front (axle) drive and the rear (axle) drive. It is not a differential. Conceptually, it just joins the front half of the propshaft to the rear (propshaft).



1. When engaged, can the Haldex ever change the 50:50 drive (front:rear) to be biased more to the rear?
A: No. The Drive will always be, 100:0 (in the case of Haldex disengaged, FWD), or 50:50 (in the case of Haldex engaged, AWD).


2. Can the Haldex ever cause the rear wheels to turn faster than the front (assuming equal friction surfaces for all tyres)
A: No. Haldex can never make the car into a power-oversteering monster (assuming equal friction surfaces for all tyres)


3. Can the Haldex "send" more than 50% of the torque to the rear?
A: Yes (in Haldex marketing terms) eg: Front tyres are on a skating rink, rear tyres are on grippy tarmac.



In Haldex world, you can't "send" torque. Torque in Haldex speak is what is being transmitted from tyre to ground.

In the skating rink example above, the rear wheels will not be rotating faster than the front (ignoring the cross-axle diff), as the drive front:rear will be 50:50. But as the rears are gripping more than the fronts, then more torque is being transmitted by the rear wheels, than the front, and this will give you movement.

In other driving situations and scenarios, you can set up the car to get all 4 wheels spinning and drifting, and yes, through driving and momentum, you can hang the back of the car out (for a period of time) - but on a dry grippy road, assuming equal traction at all 4 corners, most of the torque will be transmitted by the front tyres before it even thinks of going to the rears.

Thats not to say that the car will always feel like a FWD car. Through careful setting up of the corner, you can utilise the fact that the rears are going to take a shorter path than the fronts (and turning slightly slower than the fronts), and you manually engage the Haldex to be ON, you can force the rears to turn at the same speed as the fronts, and you will "feel" the rears pushing.


So. If you go by the definition that you define torque by what is being transmitted between tyre and ground, then yes, with a Haldex Unit, more torque can be available at the rear tyres than the front, depending on the friction (mu) of the surface.

However, with the Haldex engaged, you can never have more than 50:50 drive (front:rear), and the rear wheels can never (in a straight line, equal mu surfaces) be made to turn faster than the front wheels.



The following video illustrates torque transmitted at the tyres, as seen by Haldex.

There is some misinformation here. First, there is always about 15-20% of torque at the rear wheels, it is never 0%, unless the rear is on glaze ice and the front has grip. This factoid used to be found on the Audi site too, under the lexicon section.

Second, it is possible to apportion more than 50% of the torque to the rear axle even when the car is on a dry road with good grip at all 4 wheels. Haldex responds to differences in shaft speeds between the front and rear in order to eliminate slip. Dynamic weight distribution plays a huge role here. When you accelerate, the weight shifts rearward and the front axle slips imperceptibly (to the driver, but not the system). In most acceleration scenarios about 65/70% of torque goes to the rear axle, as anytime there is slip on the front diff, the power flows rearward. This also happens in cornering, as, in addition to the weight shift that goes on, the front wheels take a longer path than the rear. In many cornering scenarios the car will also have a rear torque bias.

Furthermore, up to 100% of torque could be sent rearward if the front axle is on a very slippery surface and the rear isn't. The most important takeaway here is that there is almost ALWAYS slip on the front axle whether the road is wet or dry. So about the only time your car has 15-20% of torque at the rear is in steady state cruising at a constant speed, or in very low-speed parking maneuvers.

A lot of this info. was provided by Ulf Herlin of Haldex on a different forum when a member wrote in to ask about the Haldex in his S3 8P.

Lastly, the Gen. 4 Haldex on the facelift S3 pre locks the clutch pack unit, meaning that the car predicts when it will need to send large amounts of torque to the rear, and this makes the system proactive and capable of transmitting up to 50% of torque at any time. The difference between a Torsen and a Gen 4 (or the newer Gen 5 ) on the road is minimal, and the Haldex is actually faster to shift torque since it is electro mechanical and not purely mechanical.

The only real difference on the road for the Torsen system is that the latest versions have a more rear-biased feel in some situations since the center diff now has 40/60 base configuration, but this is not really much different than the old 50/50 quattro or the latest Haldexes in driving feel.

I have owned them all and the Gen 4 and 5 Haldex drive so similarly to the Torsens that it isn't even worth worrying about.

 

[scherzkeks]

i have first hand experience with that taking the Golf R to the race track how would you explain that in drags, we were chewing through toyo r888s' (rear tires) at a quicker rate in fact they were being chewed 3x as quick as the front and thats with the stock AWD system, just saying... here is a much more thorough and detailed view of the Golf R audi S3 haldex setup and the quattro set up in longitudinal A4 etc..

have read you, i can assure you'll be interested

Can someone explain the FWD Haldex system to me? - VW GTI MKVI Forum / VW Golf R Forum / VW Golf MKVI Forum / VW GTI Forum - Golfmk6.com

Regarding rear tire wear, the reason your rears wore faster than the front is that with Haldex 4, the rears get about 70% in acceleration scenarios. Also the clutch pack can lock up independent of slip, so rear wheel engagement is much higher than on previous versions. My old S3 8P Sportback also wore through rear tires faster than the fronts by about 2mm on average, with even flat wear across the tire. See my post above for more...

 

[veeeight]

There is some misinformation here.

There sure is. Here we go again. This isn't Pistonheads

First, there is always about 15-20% of torque at the rear wheels, it is never 0%, unless the rear is on glaze ice and the front has grip.

There will be some drag due to the viscosity of the clutch pack oil, this is where the 10-15% figure that Ulf Herlin talks about comes from. If you have a look at a graph of engagement, it is not proportional at all to the front over the entire rev range, this is simply oil drag, and not driven power.
There is definitely no driven rear with the clutch pack disengaged, otherwise we would not be able to use the 2 wheel rolling road for brake tests at the MoT stations in the UK.

Furthermore, up to 100% of torque could be sent rearward if the front axle is on a very slippery surface and the rear isn't.

Even in Ulf Herlin's words, "near 100% torque transfer" (he nevers says 100%) - can happen, hypothetically. If your front wheels and no grip (ice rink) - AND the front of the car is suspended off the ground (no weight loading). As soon as the front end of the car is on the ground, with the weight of the front and engine on the wheels, that magical "near 100%" figure disappears. Even accounting for acceleration which lifts the front, you are not going to magically eliminate the weight off the front wheels. Therefore "near 100%" torque transfer to the rear is not possible in reality. Only for marketing purposes.

Unless, of course, I happen to find myself with my front wheels on an ice rink AND suspended, whilst my rear wheels are on grippy tarmac

The most important takeaway here is that there is almost ALWAYS slip on the front axle whether the road is wet or dry.

I have no idea what you are talking about here

This also happens in cornering, as, [.....].the front wheels take a shorter path than the rear

Not in my world, they don't

I have owned them all and the Gen 4 and 5 Haldex drive so similarly to the Torsens that it isn't even worth worrying about.

In that case you can probably answer why all S3 8P's can always power oversteer out of corners?

 

[scherzkeks]

There sure is. Here we go again. This isn't Pistonheads



There will be some drag due to the viscosity of the clutch pack oil, this is where the 10-15% figure that Ulf Herlin talks about comes from. If you have a look at a graph of engagement, it is not proportional at all to the front over the entire rev range, this is simply oil drag, and not driven power.
There is definitely no driven rear with the clutch pack disengaged, otherwise we would not be able to use the 2 wheel rolling road for brake tests at the MoT stations in the UK.

It is clutch pretensioning, and it means that 15% of the torque is on the rear wheels. The Haldex fully disengages on braking as it is electronically triggered to do so, and there is no reason for it to stay engaged. It does not disengage fully when cruising on a motorway.

Even in Ulf Herlin's words, "near 100% torque transfer" (he nevers says 100%) - can happen, hypothetically. If your front wheels and no grip (ice rink) - AND the front of the car is suspended off the ground (no weight loading). As soon as the front end of the car is on the ground, with the weight of the front and engine on the wheels, that magical "near 100%" figure disappears. Even accounting for acceleration which lifts the front, you are not going to magically eliminate the weight off the front wheels. Therefore "near 100%" torque transfer to the rear is not possible in reality. Only for marketing purposes.

It is possible in reality, just unlikely. The more important figures are found in the dynamic weight distribution scenarios I mentioned, as this is where AWD is needed most.

I have no idea what you are talking about here

There are a variety of factors that cause slip in the front diff, and there is rarely a state in which you are driving where the front is not registering some degree of slip.

In that case you can probably answer why all S3 8P's can always power oversteer out of corners?

Since the system matches axle speeds, this would be impossible. Just as it is with a Torsen 50/50 or any system that does not overdrive the rear. In some regards Torsen cars are also more prone to understeer than Haldex cars, but it depends on what section of the corner you are talking about, and what the driver inputs are.

I'm not exactly sure which of the 3 PH stalkers you are, but I do find it amusing that you are now integrating information that I brought up in past threads (that you challenged) on that site and passing it off as information that you always knew and accepted as fact. At least you appear to be learning something.

 

[veeeight]

It is clutch pretensioning, and it means that 15% of the torque is on the rear wheels. The Haldex fully disengages on braking as it is electronically triggered to do so, and there is no reason for it to stay engaged. It does not disengage fully when cruising on a motorway.

This is clutch drag, not engagement. It is not a constant 15%. it is only 15% at a specific rev, tapering to less than 10% at 5000rpm. Look up the engagement graphs for yourself, always better than forum heresay. Once again, if it is engagement, use of the 2 wheel brake test rolling road would be impossible as the car would be trying to jump off the rollers. This does not happen.

Additionally - clutch pretensioning only possible on Gen 4 & 5. Ulf wrote that email back in Gen 2 days, before pretensioning. That blows the 15% engagement out of the water.

It is possible in reality, just unlikely. The more important figures are found in the dynamic weight distribution scenarios I mentioned, as this is where AWD is needed most.

Exactly when will I find myself with my front wheels with no weight on them, and rears on grippy tarmac?
"Near !00%" torque transfer to the rear is a marketing scenario, not reality.

There are a variety of factors that cause slip in the front diff, and there is rarely a state in which you are driving where the front is not registering some degree of slip.

These are your words - "The most important takeaway here is that there is almost ALWAYS slip on the front axle whether the road is wet or dry". Demonstrate this in a straight line, with equal high friction surface on all tyres please.


I think you didn't detect the sarcasm of the oversteering S3 scenario. It was a reference to how you found Haldex so similar to Torsen that it wasn't worth bothering with the difference. Nevermind.

I'm not exactly sure which of the 3 PH stalkers you are, but I do find it amusing that you are now integrating information that I brought up in past threads (that you challenged) on that site and passing it off as information that you always knew and accepted as fact. At least you appear to be learning something.

I have no idea what you are talking about. I am not a member of PH, and have never posted there. I just occasionally read some of the far fetched scenarios on there.
Why, have you got previous form for rubbish on there ?

 

[scherzkeks]

This is clutch drag, not engagement. It is not a constant 15%. it is only 15% at a specific rev, tapering to less than 10% at 5000rpm. Look up the engagement graphs for yourself, always better than forum heresay. Once again, if it is engagement, use of the 2 wheel brake test rolling road would be impossible as the car would be trying to jump off the rollers. This does not happen.

Additionally - clutch pretensioning only possible on Gen 4 & 5. Ulf wrote that email back in Gen 2 days, before pretensioning. That blows the 15% engagement out of the water.

So it is either 15% or 15% that can tapers down to 10%. Even assuming you are right (which I know you are not, as the 15% figure has been published by Audi AG), I don't see the issue.

As for the rest, you clearly don't even understand how your own MOT RR dynos work for brake testing.

Exactly when will I find myself with my front wheels with no weight on them, and rears on grippy tarmac?
"Near !00%" torque transfer to the rear is a marketing scenario, not reality.

What exactly is your point? The mechanical ability to do so is there, and this extreme example illustrates the greater concept -- that under slip, more than 50% of torque passes rearward. Do you actually have trouble comprehending this?

These are your words - "The most important takeaway here is that there is almost ALWAYS slip on the front axle whether the road is wet or dry". Demonstrate this in a straight line, with equal high friction surface on all tyres please.


I think you didn't detect the sarcasm of the oversteering S3 scenario. It was a reference to how you found Haldex so similar to Torsen that it wasn't worth bothering with the difference.

1. The key word is "almost." Too subtle?

2. In my opinion the systems are virtually interchangeable as sold in current Audi products. You aren't doing power oversteer with a 50/50 Torsen, nor with the 40/60, unless you have a sport diff -- and even then it is effort. I should know, I have one.

I have no idea what you are talking about. I am not a member of PH, and have never posted there. I just occasionally read some of the far fetched scenarios on there.

3/10

 

[veeeight]

as the 15% figure has been published by Audi AG

Then you can provide a link, or document. You must be getting confused with Gen2 Haldex.

1. The key word is "almost." Too subtle?

Yes. Especially seeing as Gen 2 Haldex required 15 degrees of wheel slip to couple. "Almost" doesn't cut it.

2. In my opinion the systems are virtually interchangeable as sold in current Audi products.

One is a full time all wheel drive system, the other is a part time all wheel drive system. Too subtle? Or just difficulty comprehending?
You would be a genius in the marketing department, actually you could save them a shed load of money by fitting Haldex to the RS6 if it's "virtually interchangeable".

3/10

Again, I have never posted on PH. I'm not one of your so called stalkers. Paranoia?

 

[scherzkeks]

Then you can provide a link, or document. You must be getting confused with Gen2 Haldex.

For Gen. II, the amount of rear wheel engagement varied by manufacturer and even among models in the range. Audi previously stated 15% on their old website under the lexicon. Other than that, you have a link already -- Herlin confirms between 10-15% for the Gen. II and points out that this amount of torque is enough to stabilize the car and also reduce fuel consumption when large amounts of rear torque are not needed.

A side note though, the Gen IV is completely different. It uses the same pretension on start up of about 15%, but uses a seperate electronic pump to keep the accumulator full , so when one accelerates or when the system deems otherwise necessary, full locking torque is available immediately, rather than the system having to build pressure to effect more clutch lockup.

Yes. Especially seeing as Gen 2 Haldex required 15 degrees of wheel slip to couple. "Almost" doesn't cut it.

False. Wheel spin is not the same as slip on the axle. In a normal acceleration scenario with good grip on all four tires and a dry road, we know for a fact that more than 50% of torque can and does go to the rear axle with no wheel spin, due to slip from dynamic weight transfer. The lag from detection to engagement is the reason why Gen. II would sometimes exhibit a small amount of wheel spin, as the system did not have enough instantaneous pressure available to fulfill the amount of toruqe requested by the ECU.

One is a full time all wheel drive system, the other is a part time all wheel drive system. Too subtle?

No, just flat out wrong. Full time does not denote a 50/50 or any torque split. It denotes that the system is working at all times, which all Haldex generations do. The word you seek is proactive, which Gen. II was not, since it could not always supply enough torque instantaneously to meet ECU demands -- an issue that sometimes manifested itself to the driver in "seesaw" like transitions from under to oversteer in on-the-limit handling.

 

[veeeight]

Audi previously stated 15% on their old website under the lexicon. Other than that, you have a link already -- Herlin confirms between 10-15% for the Gen. II

Fact is, there is no official Audi document stating a 85/15 split. This is just forum heresay, and that is where you appear to have picked it up from.

Herlin does not speak for Audi calibration, he is/was VP Marketing.

So are you accepting that Gen 4 has no 85/15 split ? And is actually 100/0 (or 95/5) ?

Once again, you have failed to prove any doumentation from Audi regarding this fictitous 85/15 split.

we know for a fact that more than 50% of torque can and does go to the rear axle with no wheel spin, due to slip from dynamic weight transfer.

You ae confusing fact, with your opinion or supposition.
Provide calculations of the weight distribution of the car while static, and dynamically under acceleration. I doubt that it changes such that over 50% torque transfers to the rear on an equal friction surface.


Haldex is not, and will not be, including Gen V, a full time AWD system. If this was the case, the clutch packs would be burnt out and overheated in no time at all with a constant slip on the clutches. It may be working all the rime, but it certainly is not engaged all the time. You are plain wrong and misinformed.

 

[scherzkeks]

Fact is, there is no official Audi document stating a 85/15 split. This is just forum heresay, and that is where you appear to have picked it up from.

Herlin does not speak for Audi calibration, he is/was VP Marketing.

So are you accepting that Gen 4 has no 85/15 split ? And is actually 100/0 (or 95/5) ?

Once again, you have failed to prove any doumentation from Audi regarding this fictitous 85/15 split.

Herlin clearly states a rough figure of 10%, depending on the manufacturer. And as mentioned to you, the figures for Audi used to be on the homepage under the lexicon. Here is an old German press release for the last TT coupé confirming that under normal circumstances 85% of the power is on the front axle (the final paragraph under Antrieb).

Audi TT Coupé

Worth keeping in mind is that the calibrations are more aggressive for S and RS models.

You ae confusing fact, with your opinion or supposition.
Provide calculations of the weight distribution of the car while static, and dynamically under acceleration. I doubt that it changes such that over 50% torque transfers to the rear on an equal friction surface.


Haldex is not, and will not be, including Gen V, a full time AWD system. If this was the case, the clutch packs would be burnt out and overheated in no time at all with a constant slip on the clutches. It may be working all the rime, but it certainly is not engaged all the time. You are plain wrong and misinformed.

This is just trolling at this point. Permanent means exactly that. It is working full time and sending at least 15% of power to the rear under light load. As mentioned to you before, the key thing to remember with Gen. 4 (Audi models from mid-2008 on) and 5, is that this number is also nearly meaningless, since the coupling can be fully locked at any time in a matter of milliseconds by the ECU. The system does not need axle slip to react.

Furthermore, in terms of actual torque distribution, rough base numbers are (again) to be found in the e-mail from Herlin on the Gen. II system: on acceleration, 60-70 percent usually ends up at the rear, on a grippy surface with no wheel spin (axle slip is not necessarily wheel spin). Logic dictates that since the system must respond to dynamic weight distribution (as it effects axle load and thus slip), it is going to be most heavily involved under heavy load transfer, such as in cornering and acceleration. Considering that the Gen II already could shift a majority of torque to the rear under various dynamic conditions, it is quite clear that the Gen IV and V are even more effective at this, given their ability for full preemptive lock up. Rear tire wear and TüV dynamometer testing proved the theory for me when I had my old S3 Sportback.

 

[veeeight]

Herlin is VP Marketing Haldex and does not speak for Audi Calibration.

This is clearly stated in his email, that individual manufacturers calibrations are different.

You have failed to show me any documentation from Audi AG that shows the supposed split you speak of, you are merely propagating Internet Myth, as you frequently do on PH.

On Gen V there is no such permanent lock or split. The 5-10% is clutch drag, no engagement. In addition, on Gen V 8V, the coupling is deliberately not engaged to improve their fuel economy figures. By this or any definition, Haldex is a part-time system.

VCDS logging on clutch clamping pressures prove full clutch dis-enagement on the 8P and 8V.

 

[S3Alex]

This is just trolling at this point.

Furthermore, in terms of actual torque distribution, rough base numbers are (again) to be found in the e-mail from Herlin on the Gen. II system: on acceleration, 60-70 percent usually ends up at the rear, on a grippy surface with no wheel spin (axle slip is not necessarily wheel spin). ......

Trolling ehh?

As you're much better at this than myself for instance,can you explain how the Haldex can apportion 60-70% to the rear,without there being a large amount of front axle slippage.

You're telling me one thing,but the car,the TC and prior experience of cars with torque sensing diffs and rearward torque bias is telling me otherwise.

 

[Jtaylor]

My brain is melting reading this.

 

[S3Alex]

My brain is melting reading this.

It's possible that someone else's melted writing it.


The point is that whilst the Haldex clutch is a clever idea,and generally works well,it is not the omnipotent,all-conquering God of 4WD.

It's clear that there will always be jokers who will make wisecracks about anything that doesn't fit into their take on things.

 

[nluk100]

Some more information here ... AUSmotive.com » Frankfurt: Volkswagen Golf R -specifically;

As standard equipment, the Golf R transfers the TSI’s power to the road via the latest generation of Volkswagen’s 4Motion all-wheel drive system. Compared to the version implemented in the Golf R32, the system underwent significant advanced development. Above all, power transmission between the front and rear axles – especially the all-wheel differential that operates in an oil bath – exhibits clear advances compared to the previous generation. The most important one: Activation of the all-wheel differential no longer requires a difference in speeds between the front and rear axles.

That is because, different than on the previous generation, for the first time an electric pump is used to build pressure. The electric pump supplies oil to a hydraulic reservoir whose working pressure is 30 bar. A control module computes the ideal drive torque for the rear axle and controls, via a valve, how much oil pressure is applied to the working pistons of the multi-plate clutch. The contact pressure at the clutch plates rises in proportion to the desired torque at the rear axle. The amount of torque that is transferred can be varied continuously with the magnitude of the pressure applied to the clutch plates. Compared to the previous 4Motion generation, the system operates independent of slip, since the system’s working pressure is always available. When starting up and accelerating, this prevents spinning of the wheels at the front axle more effectively, since the control module regulates the torque distribution based on dynamic axle loads. In extreme cases, nearly 100 percent of the drive torque can be directed to the rear axle. This results in further gains in active safety and dynamic performance.

 

[Brodster]

Have just put my feet up for an hour and read all of this and watched the videos recommended..........Im all ears for more. Very informative veeeight...... I now kind of understand the working physics of locked/unlocked diffs and how the Haldex system operates and it was never intended to be the King of 4WD as S3_Alex put it. If that was the case then just pop 2 locked diffs in and away we go along with alot of accidents as I have driven afew Legend cars with solid locked rear diffs and they will spit you off the track if you get it wrong......well it did to me anyway.

 

[S3Alex]

As mentioned and explained previously,the only extreme circumstance in which 100% of the torque can be directed to the rear is if the front axle is slipping completely I.e no grip.

Otherwise even a locked up Haldex cannot apportion all of the torque to the rear.

 

[nluk100]

I don't understand why everyone is getting hung up about wanting to have 100% torque fed to the rear wheels? Surely if you wanted that you would just buy a BMW? The article I posted said 'nearly 100%' and 'in extreme circumstances'. The fact that it prevents wheelspin (I had a stage 2+ Ed 30 previously) and offers better economy than a Torsen system (by disconnecting the rear wheels) works for me.

 

[veeeight]

As mentioned and explained previously,the only extreme circumstance in which 100% of the torque can be directed to the rear is if the front axle is slipping completely I.e no grip.

Otherwise even a locked up Haldex cannot apportion all of the torque to the rear.

I would go even further than that.

The only extreme circumstance where "near" 100% of torque can be directed to the rear, is when the front has no grip, AND has no weight on the wheels (i.e.: Up in the air).

As long as the front wheels have weight on them, even if they are slipping on ice, *some* torque will be transmitted onto the ice (scrabbling, producing friction) - therefore you are not going to get anywhere close to 100% torque to the rear.


The other laughable suggestion, is that a Haldex car is as good as a Torsen equipped car. Torsen with specific TBR's are far superior to Haldex cars, for a start Haldex is part time AWD, Torsen is full time AWD, Haldex Gen 4&5 disengage when you brake, Torsen does not.

If Haldex were truly as impressive as Torsen, Audi, Subaru etc. would fit them to all their cars, Haldex is a good compromise for transverse engines, and to save weight/space as you don't have a centre diff.

 

[S3Alex]

I don't understand why everyone is getting hung up about wanting to have 100% torque fed to the rear wheels? Surely if you wanted that you would just buy a BMW? The article I posted said 'nearly 100%' and 'in extreme circumstances'. The fact that it prevents wheelspin (I had a stage 2+ Ed 30 previously) and offers better economy than a Torsen system (by disconnecting the rear wheels) works for me.

Hi nluk.

The comment was aimed at scherzkeks who appears to be very hung up on this.

A quick trawl through a lot of other forums will confirm that.

 

[Lima]

2. Can the Haldex ever cause the rear wheels to turn faster than the front (assuming equal friction surfaces for all tyres)
A: No. Haldex can never make the car into a power-oversteering monster (assuming equal friction surfaces for all tyres)

On dirt roads and/or wet roads I've had the **** of my S3 out several times. True, it's not as tail happy as a RWD car and you need a fair degree of safety margin (on wet roads, dirt not so much) but to say it can't be done just means you haven't tried hard enough.

 

[S.]

i'll just leave this here..
BMW E92 Audi S3 drift - YouTube

 

[veeeight]

^ Very nice effort


Did you notice:


1. S3 has to invoke the handbrake to get the tail out (once you get the tail out and have inertia, things become easier).


2. The rear wheels of a car take a shorter path naturally (this is why you don't hug the kerb when turning left out of UK junctions - those alloys !).

If you switch off the ESP in a S3, you're engaging Haldex, so you are locking up the rear wheels to turn at the same revolutions as the front wheels. So - you are effectively asking the rear wheels to "step out" and follow a longer path.

This is why many people say they feel the "rear wheels kick in" or feel the "rear wheels driving" when Haldex engages - the rear of the car is trying to take a longer path that it isn't natural to it! (leads to tyre scrub, transmission windup).

3. Once the S3 invokes the rear out (handbrake), has momentum/inertia, and the rear "steps out" - the driver has to do his best to keep the fronts from gripping if he wants to continue the donut. Note that as soon as he lets up on the throttle, the fronts grip, and off he shoots at a tangent!



edit:
In this video, Tiff demonstrates just how tricky it is to drift a AWD car:

 

[kanecullen89]

i'll just leave this here..
BMW E92 Audi S3 drift - YouTube

I don't think the haldex will last long doing those sorts of forced drifts. It's a fairly delicate system as it is

 

[scherzkeks]

It's possible that someone else's melted writing it.


The point is that whilst the Haldex clutch is a clever idea,and generally works well,it is not the omnipotent,all-conquering God of 4WD.

It's clear that there will always be jokers who will make wisecracks about anything that doesn't fit into their take on things.

Good. Disprove what I have said if you have issue with it. Perhaps you'll make more sense than veeeight.

1. Haldex 4 and 5 are predictive and proactive

2. The systems are faster at directing torque than a purely mechanical system

3. Haldex is generally better on mixed surfaces. Several Sport Auto tests demonstrate that it did better in sandy and snowy conditions than any mech system in their tests

4. It can transfer up to 100% of torque to the rear axle, and regularly shifts more than 50% to the rear under dynamic load shift, so statements that it is "FWD" most of the time are absurd

5. It nearly always sends at least 15% of power to the rear axle, disengaging with brake intervention

6. It can be set up to drive very much like a Torsen, distributing power evenly when needed and working to ensure that axle speeds are matched to quell slippage

I would suggest that the troll is the poster saying that this is all nonsense. The reality is that Audi's mech and electro mech systems have essentially reached parity with one another. There are slight advantages and disadvantages for both, but in reality, they both do the same job: matching axle speeds to quell slip and provide stable handling. Neither are set up to allow RWD type dynamics, because that isn't really the point. Just for the record, I'd have the EVO's system over just about anything else for use on mixed surfaces.

 

[Ron burgundy]

I think trolls are kinda cute

 

[Ron burgundy]

M

 

[S3Alex]

Good. Disprove what I have said if you have issue with it. Perhaps you'll make more sense than veeeight.

1. Haldex 4 and 5 are predictive and proactive

2. The systems are faster at directing torque than a purely mechanical system

3. Haldex is generally better on mixed surfaces. Several Sport Auto tests demonstrate that it did better in sandy and snowy conditions than any mech system in their tests

4. It can transfer up to 100% of torque to the rear axle, and regularly shifts more than 50% to the rear under dynamic load shift, so statements that it is "FWD" most of the time are absurd

5. It nearly always sends at least 15% of power to the rear axle, disengaging with brake intervention

6. It can be set up to drive very much like a Torsen, distributing power evenly when needed and working to ensure that axle speeds are matched to quell slippage

I would suggest that the troll is the poster saying that this is all nonsense. The reality is that Audi's mech and electro mech systems have essentially reached parity with one another. There are slight advantages and disadvantages for both, but in reality, they both do the same job: matching axle speeds to quell slip and provide stable handling. Neither are set up to allow RWD type dynamics, because that isn't really the point. Just for the record, I'd have the EVO's system over just about anything else for use on mixed surfaces.

Personally,I would say that having made the claims,the onus is on yourself to prove them.

I understand the differences in preload etc with the later systems.

With load shift(i.e.front end lift and slip under acceleration)it would be possible to direct more than 50% of the torque rearwards,but how a centre-mounted diff of this sort can direct 100% without the front being in a zero-load/100% slip condition I cannot see,but perhaps you can explain rather than simply saying that's it,disprove it.

Having driven cars with diffs made by Haldex,Quaife and Torsen type diffs,I can honestly say that the Haldex even with an adjustable controller,is simply not as good as the Quaifes for instance,but obviously,there is a bias in terms of feel and driving characteristics in that,although the ability of a diff such as the Quaife to find traction is better than the Haldex.