Are mixed-wear tyres damaging to an AWD drivetrain?

[johnny g-pipe]

Hi amigos.

Just bought some 'turbo' wheels for my impreza 'sport' off good ole fleabay. Two have 50% worn pirellis, two have pretty new uniroyals.

I'm aware of possible handling surprises with mixed tyres, but I thought I heard once that you can damage the awd system like this because the effective rolling diameter is different between the different wheels?
I reckon the scoobie is pretty tough, but I thought I'd better check..

TIA

[Mr Whippy]

Depends how much out they are.

Though to be honest transmission wind up won't occur with such a small difference in rotational speeds across the diff in normal straight line driving, because the viscous coupling LSD's (pretty sure thats what Impreza has in the centre diff) has such a low differential between it's outputs.

It's no different in reality than driving along an back road and following curves where one end of the car follows a slightly smaller arc.

If overheating your diff is a worry with same tyre size's but slight wear variations then I'd worry about going around roundabouts several times in a row, or driving along very twisty roads at speed because the differentials then are HUGE in comparison...

Dave

[GavinPearson]

I really wouldn't worry about it from the drivetrain's perspective.

But I would make sure that 4 of the same tyres were on the wheels to give consistent handling.

[johnny g-pipe]

Thanks!

[combemarshal]

I remember reading somewhere that the difference should never be more than 2mm between any 2 tyres, can't remember where or why though!!

[GreenV8S]

The center diff normally won't see any differential speed at all. If you have different rolling circumference front/back the diff will see a continuous speed differential which will cause more wear than normal. Whether this is enough to cause a problem depends on how much speed difference there is, and how much of your time you speed in high speed cruising. If you tend to work the transmission hard anyway, then a small amount of extra wear when you're cruising probably doesn't matter. On the other hand if the car spends its life cruising up and down motorways the extra heat and wear could be significant.

[Mr Whippy]

GreenV8S said:

The center diff normally won't see any differential speed at all. If you have different rolling circumference front/back the diff will see a continuous speed differential which will cause more wear than normal. Whether this is enough to cause a problem depends on how much speed difference there is, and how much of your time you speed in high speed cruising. If you tend to work the transmission hard anyway, then a small amount of extra wear when you're cruising probably doesn't matter. On the other hand if the car spends its life cruising up and down motorways the extra heat and wear could be significant.

Surely going round any bend generates front/rear differential by varying amounts though? And since the centre diff is the same viscous LSD as the rear one (or front maybe, not sure on his scooby), then those in theory would wear with excessive cornering??

Accelerating mildly could even generate enough rolling radius change as the tyres load/unload front to rear and vice versa.

Not sure here, but is it *really* that important? I'm sure the same effect could be had by putting 3 people on the back seats with regard to rolling radius. Do 4wd car owners alter their tyre pressures constantly with varying load to optimise the changing rolling radius?


In my view at least it's negligible with same brand tyres at spec pressure's but with varied wear... but I'm happy to be proven wrong here in the name of learning more

Dave

[GreenV8S]

Mr Whippy said:

Surely going round any bend generates front/rear differential by varying amounts though? And since the centre diff is the same viscous LSD as the rear one (or front maybe, not sure on his scooby), then those in theory would wear with excessive cornering??

Yes diffs also have a speed difference across them when you're cornering, the difference is that you don't tend to drive round corners for hundreds of miles at a time at high speed.

[Mr Whippy]

GreenV8S said:

Mr Whippy said:

Surely going round any bend generates front/rear differential by varying amounts though? And since the centre diff is the same viscous LSD as the rear one (or front maybe, not sure on his scooby), then those in theory would wear with excessive cornering??

Yes diffs also have a speed difference across them when you're cornering, the difference is that you don't tend to drive round corners for hundreds of miles at a time at high speed.

But what about varying loads as per my previous post? Ie, passengers, fuel, accelerating/braking, then the cornering factor ontop. And the fact that stock the cars (scooby) have ~ 60% front load, so the front tyres may well sit with a lower rolling radius anyway vs the rears.

Surely after all those, tread is one of only many many factors that will influence potential front and rear axle differential speeds at the centre diff? Is it significant next to the other factors? Should owners be pumping up their front tyre pressures or letting down the rears so they have the same RR?

For example, my 306 Gti6 with 34psi all round in the tyres had a frontal tyre to rim measurement of 55mm vs the rears which had 70mm. That HUGELY outweighs the tread variation of maybe 5-10mm... knowing the F/R masses also allowed me to calculate the tyre rate from that, which was nice Then I could do 30psi all round and get the rate at that lower pressure. Yes it's only two points, but to get the rear RR the same as the front would need around 40psi.

Again, not saying I'm right, just is the tyre tread significant next to these other factors?

Dave

Edited by Mr Whippy on Saturday 19th August 22:12

[GreenV8S]

Mr Whippy said:

But what about varying loads as per my previous post?

You're thinking about how differences in sidewall deflection affect the rolling circumference?

If you imagine a foam tyre with a rolling radius of say 2 feet, and put enough weight on it to squish the rolling radius down to 1 feet, it'll behave like a rigid wheel with a radius of 1 foot.

Tyres don't behave like that though. The reason is that they have a very stiff belt underneath the tread which forces the overall circumference to stay the same. If you put weight on the tyre and squish the sidewall the tread gets closer to the axle, but one revolution of the axle still moves you the same distance along the ground. Imagine a tank track connected via very squidgy springs to an axle, they both move in all sorts of directions but one revolution of the axle always equals one revolution of the tread regardless of how high or low the axle is.

There are some effects that vary with weight - mainly in the amount of slip that the tread generates, but they're quite small.

[Mr Whippy]

GreenV8S said:

Mr Whippy said:

But what about varying loads as per my previous post?

You're thinking about how differences in sidewall deflection affect the rolling circumference?

If you imagine a foam tyre with a rolling radius of say 2 feet, and put enough weight on it to squish the rolling radius down to 1 feet, it'll behave like a rigid wheel with a radius of 1 foot.

Tyres don't behave like that though. The reason is that they have a very stiff belt underneath the tread which forces the overall circumference to stay the same. If you put weight on the tyre and squish the sidewall the tread gets closer to the axle, but one revolution of the axle still moves you the same distance along the ground. Imagine a tank track connected via very squidgy springs to an axle, they both move in all sorts of directions but one revolution of the axle always equals one revolution of the tread regardless of how high or low the axle is.

There are some effects that vary with weight - mainly in the amount of slip that the tread generates, but they're quite small.

Ah yes. Thats starting to make sense now the more I think about it.

Hmmmmm, very interesting point. Now I need to go have a looonnnggg think about it so I am happy with it in my mind

Cheers for that. Thats quite a fundamental concept and property that I haven't taken account of in my visualisations!

So basically you just get the tyre effectively running flat at the CP, not actually loosing any circumferance. Yes the rolling radius is smaller, but the rolling circumferance (as silly as it sounds) isn't!



Dave

[GravelBen]

I've heard that they don't really like it, but I wouldn't imagine the difference in diff rotation caused just by tyre wear to be much. If they're different brand tyres it may be worth measuring up to make sure rolling radius is about the same, I've heard even tyres of the same nominal size can be slightly different radius from different brands. I've certainly never had any problems with my AWD Subaru Legacy from running tyres of different brand or wear levels, I make sure LF and RF are same brand/similiar wear, and LR/RR also similiar brand/wear.

As far as handling goes, I find it good to have better tyres on the front than the back to help cancel out some of the natural understeer that you get with 50/50 AWD cars. I tend to wear out rears faster, so whenever they need replacing the fronts get switched to the back and nice new ones on the front.

[combemarshal]

GreenV8S said:

Mr Whippy said:

But what about varying loads as per my previous post?

Tyres don't behave like that though. The reason is that they have a very stiff belt underneath the tread which forces the overall circumference to stay the same. If you put weight on the tyre and squish the sidewall the tread gets closer to the axle, but one revolution of the axle still moves you the same distance along the ground.

No it wouldn't if the tyre was 'Squashed' it would require more turns of the axel to match the other side, the same as putting a smaller wheel/tyre on

[Mr Whippy]

combemarshal said:

GreenV8S said:

Mr Whippy said:

But what about varying loads as per my previous post?

Tyres don't behave like that though. The reason is that they have a very stiff belt underneath the tread which forces the overall circumference to stay the same. If you put weight on the tyre and squish the sidewall the tread gets closer to the axle, but one revolution of the axle still moves you the same distance along the ground.

No it wouldn't if the tyre was 'Squashed' it would require more turns of the axel to match the other side, the same as putting a smaller wheel/tyre on

Gonna test this today, just to get it into my head in a practical way.

Get a flat tyre, push it along the floor and put a mark on it so I do one rotation and mark distance travelled.

Then do it with a high pressure tyre and do it again


I think GreenV8S is right, I was up a few hours last night thinking about it. The only problem in my mind was a curve has x distance, if you flatten that arc into a straight line it is longer than the arc in the axis of the direction of travel. It's almost like the leading and trailing edges of the contact patch would be pushed forwards and backwards, or the extra "length" is forced into the rest of the non contact patch tyre...

Either way, it makes sense. It's just like tank tracks as said. The overall circumferance is what matters, rolling radius is irrelevant. You can't in theory alter the rolling circumferance without compressing the tyre lengthways considerably!

Dave

[GreenV8S]

combemarshal said:

No it wouldn't if the tyre was 'Squashed' it would require more turns of the axel to match the other side, the same as putting a smaller wheel/tyre on

That isn't how it works. The critical point to keep in mind is that the circumference of the tyre is absolutely fixed regardless of what shape the tyre is, because the tread is rigid in tension. It's tempting to visualise the tyre as a homogenous flexible lump, but it isn't.

The tank track analogy is the clearest way I can think of to explain why one revolution of the axle always[*] moves the tyre the same distance along the ground regardless of what shape the tyre is.

[*] I'm ignoring the effects of slip between the tyre and the road, because in the scenario we're discussing the amount of slip is negligeable.

[Scuffers]

GreenV8S said:

combemarshal said:

No it wouldn't if the tyre was 'Squashed' it would require more turns of the axel to match the other side, the same as putting a smaller wheel/tyre on

That isn't how it works. The critical point to keep in mind is that the circumference of the tyre is absolutely fixed regardless of what shape the tyre is, because the tread is rigid in tension. It's tempting to visualise the tyre as a homogenous flexible lump, but it isn't.

The tank track analogy is the clearest way I can think of to explain why one revolution of the axle always[*] moves the tyre the same distance along the ground regardless of what shape the tyre is.

actually that's wrong...

a tyres rolling radius is the important bit, yes the tread lenght will always be the same, but that's not the bit you need to look at - using yout tank track, the radius of the drive wheel is the important bit, irrispective of the lenght of the track.

also, back to the orriginal question, it depends entirly on what 4WD/AWD system you are looking at.

Proper 4WD with a real centre diff will have no problem with a variance in tyres, however, there are *many* so called AWD cars that don't actually have a diff, they use all things from a viscous coupling to a clutch (ala Hadex etc). (edited to add - Subaru's have a diff with a viscous LSD in them, this is not what I am reffering to)

the latter type will have a significant problem with odd tyres, certainly the old Vauxhall 4x4's (Cavalier etc) used to have warnings in the handbook to this effect, if you did not heed it, the viscous pack would ususally fail in short order (sometimes with interesting effects!).

With the Hadex ones, depending on which end of the car has the 'smaller' tyres, the cluch will either never engage, or be dragging all the time.

either way, it's not good, then again, calling these 4WD is a bit of a piss-take too.

Edited by Scuffers on Sunday 20th August 18:47

[Pigeon]

Scuffers said:

a tyres rolling radius is the important bit, yes the tread lenght will always be the same, but that's not the bit you need to look at - using yout tank track, the radius of the drive wheel is the important bit, irrispective of the lenght of the track.

If you put a paint mark on the tyre where it touches the ground, and one on the wheel at the point nearest to the ground, then watch it rolling along, the wheel mark will always reach the lowest point when the tyre mark touches the ground. This remains true no matter how hard or squidgy you inflate the tyre. One wheel rotation always equals one tyre's circumference worth of movement along the ground.

[Mr Whippy]

Pigeon said:

Scuffers said:

a tyres rolling radius is the important bit, yes the tread lenght will always be the same, but that's not the bit you need to look at - using yout tank track, the radius of the drive wheel is the important bit, irrispective of the lenght of the track.

If you put a paint mark on the tyre where it touches the ground, and one on the wheel at the point nearest to the ground, then watch it rolling along, the wheel mark will always reach the lowest point when the tyre mark touches the ground. This remains true no matter how hard or squidgy you inflate the tyre. One wheel rotation always equals one tyre's circumference worth of movement along the ground.

Yep, tested this tonight like that...

Doesn't make any difference.

The problem I always tripped up here was using the word rolling radius, which ultimately means nothing. It's a useless figure really, because the tyre rolls about it's fixed circumferance, not it's variable radius!

Thanks for making it clear anyway GreenV8S, changed my way of viewing tyres quite alot which can only be a good think. So obvious yet I overlooked it so many times!

That silly word, should be banned. Rolling circumferance is where it's at

Dave

[GreenV8S]

Scuffers said:

actually that's wrong...

a tyres rolling radius is the important bit, yes the tread lenght will always be the same, but that's not the bit you need to look at - using yout tank track, the radius of the drive wheel is the important bit, irrispective of the lenght of the track.

I don't think you read my description of the tank trap analogy properly. I wasn't describing a tank track installed in the conventional way with a drive wheel. Think of a tank track connected to an axle by lot of flexible springs so that the tread is free to move in and out as far as it wants, but the tread still has to complete one revolution for each revolution of the axle. Then visualise how far the track will move along the ground for one revolution of the axle when the track is arranged into a neat circle like a giant wheel, and when it is flattened out into a the normal caterpilar track sort of shape. The distance is the sam, because the length of the track hasn't changed.

The point is that in the case of car tyres, the rolling circumference is what is important and this does not change when the sidewall deflects.

[richb]

Pigeon said:

Scuffers said:

a tyres rolling radius is the important bit, yes the tread lenght will always be the same, but that's not the bit you need to look at - using yout tank track, the radius of the drive wheel is the important bit, irrispective of the lenght of the track.

If you put a paint mark on the tyre where it touches the ground, and one on the wheel at the point nearest to the ground, then watch it rolling along, the wheel mark will always reach the lowest point when the tyre mark touches the ground. This remains true no matter how hard or squidgy you inflate the tyre. One wheel rotation always equals one tyre's circumference worth of movement along the ground.

Yep...