How important is all wheel drive bias?

Is it enough to simply have all wheel drive for rally cross? Or will there be a big difference in Timeslips depending on your vehicles bias?

Try driving a DCCD-equipped STi with the bias as forward as it will go, at 50:50, and as far rearward as it can go. To some people it's a revelation.

Different bias will change how you have to drive the car, but from car to car, other differences will likely be more significant than a difference in bias. 

Based on the number of times I've seen FWD cars be playing right along with AWD stuff on some courses, I'd say AWD bias is likely not a huge concern. 

I would guess it has more effect on having fun than on going fast. Often times the most fun way around a course is not the fastest anyway. But, if fun is your primary concern i would go for the largest rear bias possible on a slow car, with it fading to 50:50 as the power level of the cars in question goes up. 

It matters. On paper a first gen Audi TT should behave well enough, but in practice the front and the rear end up fighting each other and make the car way less predictable than it would be if it were FWD or RWD. 

Specifically, what car are you asking about? 

G_Body_Man said:

Try driving a DCCD-equipped STi with the bias as forward as it will go, at 50:50, and as far rearward as it can go. To some people it's a revelation.

50/50 is an open diff.  Fully locks is fully locked and the torque bias basically becomes "infinitely variable".

The STI is never 50/50 anyway, when the diff is "full open" it is 33/67 due to its being a planetary design.

IIRC, ZB always ran his STI with the DCCD set to automatic, which, if the programming isn't regretful, really should be the way to go anyway.  You should only ever want the center diff to tighten up when under acceleration, and you want it to loosen up when you are trying to turn.

Edit:  I think I see what you meant, now.  The DCCD isn't "bias forward" or "bias rearward", it is how tight or loose the computer-actuated clutch packs are.  Full forward is as close to an open diff as it can be.  Full back is as close to a locked diff as the clutches will allow.  And then there's automatic mode, where the computer controls the diff based off of throttle position (maybe engine torque?) and lateral acceleration.  I'd like to say steering angle has something to do with it, but that could be the Evo X.

Knurled. said:

50/50 is an open diff.  Fully locks is fully locked and the torque bias basically becomes "infinitely variable".

Yep.  This is a big concept people struggle with.  An open diff is equal torque to both outputs at all times (50/50).  A locked diff is equal rotational speed on both outputs (with infinitely variable torque distribution).  And limited slips fall somewhere in between.  The common trend of referring to a locked diff as 50/50 is completely wrong. 

mazdeuce - Seth said:

It matters. Specifically, what car are you asking about? 

2017 crosstrek CVT. It’s 100%fwd until. Loss of grip, then max 60:40 at full lock. Open diffs on both ends and enough nannies to drive prince Harry insane. 

That's perfect.  All you have to do is drive the course.  Keep your head outside the car, focus on driving efficiently.  (Not fast, but efficient)  Let the computers do all the work they are supposed to do.  HAL is your friend if you don't contradict him.

If you want to pitch it sideways and do all sorts of stupid crap like that, you have the wrong car.  Get an Audi ur-quattro and lock the center and rear diffs and drive it as sideways and slow as a rear driver.  On the other hand, if you play nice with HAL, and focus on driving the course, you just may find that everyone in WRXs trying to play Colin McRae Rally in real life are giving you dirty looks as you hand them their asses.  Sliding is slow.

I'll say from experience that a locked center diff is a mixed (but mostly bad) thing in rallycross.  It sucks in tight sections.  You either pitch it sideways or it just understeers.  However, it works just fine in faster slaloms, etc. 

In reply to rslifkin :

My only AWD car had the same drivetrain as the ur-quattro.  Three open diffs, or locked center diff, or locked center and locked rear diff.  (Knob on the dashboard.  Vacuum controlled solenoids on the trans and rear diff engaged a collar to splines between the diff housing and an output, kind of like a transmission synchronizer hub.  Kinda slick in an 80s way)

I found, through experimentation, that times were fastest with the center and rear diffs locked.  I think this is because the car had a 500lb anchor in front of the front wheels and it really didn't like changing direction, but locking the diffs meant that you could at least predictably throw the car around.  It was fun, but pathetically slow.

(Notice how little turn-in there is, can't turn at all when on throttle, need to lift for the steering to work, and the car NEEDS to be thrown at every corner for it to turn at all.  And also notice how little effect the steering has for course correction.  You needed to aim the car very far in advance in order to go where you wanted)

Plenty of youtube on Subaru AWD vs everything else suggests bias (or lack thereof) is important. 

Michael

(full disclosure: my dd is a Forester)

In reply to mlwebb :

Lancia figured that the drive bias should be roughly equivalent to mass bias, so the Integrale had a 67/33 drive bias.  They only dominated Group A from the beginning to the time they left stage rally, so they probably knew what they were doing.

I will allow that, after 700 miles in the R, I am fully involved with the concept of front wheel drive until rear drive is necessary.  It drives completely unobtrusively when driving like a normal human being.  When the push comes to shove, drive shuffles to the rear wheels as necessary and it's like God Mode was applied to traction.  I may not actually ever rallycross this car, though, so I couldn't say how it works in competition.  (The car is just too damned wide)

I don't think the initial bias is that important. It's the amount that the torque split can change, how fast it can change, and what controls the change that is important. Front or rear bias shouldn't make much difference- if the system reacts quickly enough and effectively enough, there should not even be a perceptible bias. If a FWD biased system reacts too slowly or does not shift enough torque to the rear, it will behave more like a FWD. The same is true for a RWD biased system, although that is often seen as desirable to allow the car to behave more like a RWD until the AWD sends more power to the front, so torque available to the front is often delayed or limited. 

Let's take a simple but effective AWD system that I'm familiar with- from a DSM. It's a 50/50 with  a viscous limited slip in the center,  open diff in the front, and an LSD in the rear. Cruising along, all four wheels have plenty of traction and they get equal torque. But on a rally cross course, that can shift back and forth dramatically. When accelerating on dirt (or any surface, but loose surfaces make it more apparent) the front tires have more grip due to more weight over them. That quickly changes as weight transfers to the rear. The front wheels begin to slip, and the viscous coupling begins to lock up and send more power to the rear- really quickly, in less than one rotation of the front wheels. At that point, the "dumb" 50/50 system is very much rear biased. It's even more pronounced exiting a corner. The unloaded inside front will try to spin. Since the front diff is open, the center diff reacts by sending the power to the rears, which now get most of the power. The car exits the corner with a rear drive bias. It's very effective for such a basic mechanical system. Newer cars use computer controlled clutch packs to split the torque, and can anticipate the required result instead of react. If they are designed as such. Many of the basic AWD systems on the market are limited to pretty much just getting moving in the snow, and revert to FWD once moving. So I guess the answer is, "It depends."

Personally, from a how it drives perspective, I prefer things like viscous couplings between front and rear that transfer smoothly and predictably (and have some resistance across them to start with, avoiding a wheelspin and catch type event at one end).  Most of the clutch pack systems don't feel as good to me, as they seem more reactive (no power transfer until slip occurs).  

Boost_Crazy said:

I don't think the initial bias is that important. It's the amount that the torque split can change, how fast it can change, and what controls the change that is important. Front or rear bias shouldn't make much difference- if the system reacts quickly enough and effectively enough, there should not even be a perceptible bias. If a FWD biased system reacts too slowly or does not shift enough torque to the rear, it will behave more like a FWD. The same is true for a RWD biased system, although that is often seen as desirable to allow the car to behave more like a RWD until the AWD sends more power to the front, so torque available to the front is often delayed or limited. 

Let's take a simple but effective AWD system that I'm familiar with- from a DSM. It's a 50/50 with  a viscous limited slip in the center,  open diff in the front, and an LSD in the rear. Cruising along, all four wheels have plenty of traction and they get equal torque. But on a rally cross course, that can shift back and forth dramatically. When accelerating on dirt (or any surface, but loose surfaces make it more apparent) the front tires have more grip due to more weight over them. That quickly changes as weight transfers to the rear. The front wheels begin to slip, and the viscous coupling begins to lock up and send more power to the rear- really quickly, in less than one rotation of the front wheels. At that point, the "dumb" 50/50 system is very much rear biased. It's even more pronounced exiting a corner. The unloaded inside front will try to spin. Since the front diff is open, the center diff reacts by sending the power to the rears, which now get most of the power. The car exits the corner with a rear drive bias. It's very effective for such a basic mechanical system. Newer cars use computer controlled clutch packs to split the torque, and can anticipate the required result instead of react. If they are designed as such. Many of the basic AWD systems on the market are limited to pretty much just getting moving in the snow, and revert to FWD once moving. So I guess the answer is, "It depends."

So since the viscous coupling for say a Volvo V70 is in the rear drive shaft, could you just add this to any AWD car?   (edit) That has IRS.

I loved how my V70 XC handled in the snow.  Hateful FWD until it got nasty or you goosed it, then the back would step out and we you go.

P.S. Volvo's that have broken AWD are really cheap and easy to fix, $65 splined coupling in the bevel gear box.

akylekoz said:

 

So since the viscous coupling for say a Volvo V70 is in the rear drive shaft, could you just add this to any AWD car?   (edit) That has IRS.

I loved how my V70 XC handled in the snow.  Hateful FWD until it got nasty or you goosed it, then the back would step out and we you go.

P.S. Volvo's that have broken AWD are really cheap and easy to fix, $65 splined coupling in the bevel gear box.

 

My WRX (have had it since 2005, and done the rallycross thing with a bit) does something similar where it understeers a lot if turning under throttle, but if you stand on the gas for a few seconds the back will start to come around gradually...   

I wonder if this is just a typical AWD car behavior ? (I know everyone says the engine being in front of the front axle is responsible for the understeer, but I'm really skeptical that it's just (or at all) a nose-weight or momentum issue.   

I almost think it's something to do with the AWD system adding some kind of drag to the drivetrain that hinders turn-in, or something to do with the chassis or suspension design).

 If a FWD biased system reacts too slowly or does not shift enough torque to the rear, it will behave more like a FWD

Another thing that people seem to deny is that right up until you achieve power oversteer it will basically handle like a FWD too. If it cant power oversteer at all it's basically going to drive like a FWD with less 'inside tire fire' AKA 'driver induced understeer'. It's definitely a selling point but people talk about it like it's magical. 

Vigo said:

 If a FWD biased system reacts too slowly or does not shift enough torque to the rear, it will behave more like a FWD

Another thing that people seem to deny is that right up until you achieve power oversteer it will basically handle like a FWD too. If it cant power oversteer at all it's basically going to drive like a FWD with less 'inside tire fire' AKA 'driver induced understeer'. It's definitely a selling point but people talk about it like it's magical. 

My freakin' RX-7 doesn't power oversteer.

Why would you ever WANT power oversteer?  You want to apply power when exiting a corner.  You want "oversteer" only ever on corner entry, which requires driving with BOTC, and you want acceleration to straighten you out on corner exit.  That way, you just drive the car and any stupid situation you find yourself in will get fixed by adding power.

If you oversteer on corner exit, you're just driving slow, since it's the corner EXIT and rotating the car won't help you through it anymore.

This reminds me of the suspension tuning difference between the FR-S and BR-Z.  The FR-S is tuned for oversteer so that Scion people feel a "Whooo!!' when driving at 5/10ths.  The BR-Z is tuned for a little understeer so that Subaru people can drive fast.

Boost_Crazy said:

If a FWD biased system reacts too slowly or does not shift enough torque to the rear, it will behave more like a FWD. The same is true for a RWD biased system, although that is often seen as desirable to allow the car to behave more like a RWD until the AWD sends more power to the front, so torque available to the front is often delayed or limited. 

A FWD biased system will never "shift power to the rear" unless it has active differentials.  Likewise a RWD biased system will never "shift power forward".  Don't think about power, ever.  Differentials do not work based on power, they work based on torque.

Let's take a simple but effective AWD system that I'm familiar with- from a DSM. It's a 50/50 with  a viscous limited slip in the center,

Reiterating - if it has any kind of limited slip differential, it does not have a "50/50" bias.  Bias is variable depending on how much traction each end has versus how much the limited slip device is locked.

If you mean static torque distribution, then okay.

Yes, I may be being a weenie.  But differential torque distribution is one of those things that have a lot of misconceptions because a lot of people don't understand what is going on because the terminology being used is all wrong/faulty.

A tight enough limited slip can have a greater than 100% torque transfer.  The fast end can backdrive the diff and therefore drag, which is negative torque, and equal-and-opposite-reaction means that this negative torque gets applied as a positive torque to the other end of the diff.  Which is why you can hammer on a spool equipped vehicle as long as you don't try to make a tight corner at slow speed, at which point you break an axle...

Why would you ever WANT power oversteer?

Good question. I was only making a point about how people seem to think AWD has some inherent effect on 'handling'. It really doesn't unless you're applying enough throttle that you would have been spinning a tire if it wasnt awd.  Then, it will change how the car changes direction. Even then, it's debatable whether you call that 'handling'.  Up to that point it's probably just a slower version of whatever 2wd car it was based on.

Vigo said:

 

Why would you ever WANT power oversteer?

Good question. I was only making a point about how people seem to think AWD has some inherent effect on 'handling'. It really doesn't unless you're applying enough throttle that you would have been spinning a tire if it wasnt awd.  Then, it will change how the car changes direction. Even then, it's debatable whether you call that 'handling'.  Up to that point it's probably just a slower version of whatever 2wd car it was based on.

If you only coasted around corners and none of the systems connected the front and rear wheels except when you were pointed straight that would be true. I do agree that thinking AWD magically improves handling is flawed.

One interesting thing about AWD is that a good limited slip center diff (that doesn't go open off-throttle) or a locked center can slightly help braking performance on inconsistent surfaces.  Tying the 2 ends of the car together makes brake bias much less of a concern.  In the extreme case of a locked center, F/R brake bias will be effectively perfect regardless of the actual bias (because the drivetrain won't let you lock up just one end).  

MrJoshua said:

Vigo said:

 

Why would you ever WANT power oversteer?

Good question. I was only making a point about how people seem to think AWD has some inherent effect on 'handling'. It really doesn't unless you're applying enough throttle that you would have been spinning a tire if it wasnt awd.  Then, it will change how the car changes direction. Even then, it's debatable whether you call that 'handling'.  Up to that point it's probably just a slower version of whatever 2wd car it was based on.

If you only coasted around corners and none of the systems connected the front and rear wheels except when you were pointed straight that would be true. I do agree that thinking AWD magically improves handling is flawed.

You guys clearly don't drive in deep snow with hard packed snow under that much.  Admittedly It only happens a few times a year around here lately, but it is sooo much fun.  

The best (read most fun) way to go fast in this stuff is with a lot of slip angle, sometimes as much as the reverse lock will take.  Straight RWD or the Viscous coupled to the rear works great, front wheel drive just want to take you to the closest curb in the most direct way.

Oh and Subarus are the best for straight line acceleration on glare ice.  That symetrical AWD E36 M3 is for real, you can hit the gas with no hands on the wheel and it goes straight.

Maybe this belongs in the minor rant thread.

In reply to akylekoz :

I think you just summed up why I've always told people RWD > FWD in snow.  Slip angle is going to happen.  And it's much more controllable if the rear end has silly amounts of slip angle than the front end.  RWD car starts to understeer?  Just give it a stab on the throttle, step the tail out and the front end should grab.  FWD car starts to understeer and you aren't turning enough to e-brake the tail out?  Well, you're mostly berkeleyed until you lose some speed and it grabs. 

RWD just gives a better balance of control inputs that influence each end of the car.  Which means it's harder to drive well in the snow, but once you get good at it, the ability to make the car do what you want it to is much better.