Effect of negative camber on lateral load capability
 

Since there is an ongoing fascination with negative camber, I am curious if anyone has any actual data showing max lateral load capability vs. camber for a typical 50 series street tire? From what spotty "data" I could find in Milliken&Milliken, a degree of camber does not appear to have a humongous effect - a few percent of lateral capability, no more.

http://forums.pelicanparts.com/uploa...1233352192.jpg

My own experiments with accidentally setting the LF camber to +0.5 degrees as opposed to -0.5 degrees (those pesky level indicators read 89.5 either way :rolleyes: ) also suggest 1 degree has little effect when using street tires, at least as far as this definitively non-expert driver can tell.

As another thought, according to a=v^2/r, a 6% increase in lateral capability (.85 to .9 g) would only result in a 3% increase in cornering velocity. Not a whole lot for a 60 MPH corner unless my technique was absolutely flawless (it's nowhere close).

Obviously for a race car, every percent matters. It's gotta be optimum.
I would also think the wider the tire and the lower the aspect ratio, the more camber matters as well. No idea how much more or less.

Any thoughts? Am I missing something? Is it more subtle and complicated? Am I just too crappy a driver to notice? I look forward to getting edumacated a bit more...

Good subject. Practical experience says that it makes a big difference, even on the street. Ten + years ago I had a stock car w/ original suspension set to factory specs, alignment alone transformed that car and 90% of it was increased negative camber.

Submitted mostly just because I like this picture. The tires are 235/40/17 & 255/40/17 RA-1s at -2.5 & -2.0 respectively.

http://forums.pelicanparts.com/uploa...1233353524.jpg

It's going to depend very much on the
tires
wheels
geometry
track
other use besides track

In my experience, the factory settings are great for street driving.

Once I started autocrossing, I was amazed what a difference a competition alignment made, even on the same street tires.

I went from battling for 5th place, to a top contender in the class.

The main benefit of negative camber on a factory 911 suspension geometry is on the track or when aggressively carving canyons. The negative static camber helps to level the contact patch in the turn when undergoing side loads and suspension/tire deflection. This evens out wear and heat, providing more useful tire life in addition to longer heat cycles and greater grip. Proper warm operating tire pressure also prevents over-working the tire.

Of course, when driving leisurely, the large negative camber does just the opposite.

On a stock 911, the contact patch ain't going to be level ... no way enough static camber can be dialed in.

The even wear and heat thing makes sense though - so maybe for a few corners it won't make that big a difference (which I assume is the Milliken graph scenario), but once the tires start overheating on one edge because of continuous high load, high slip angle operation, the difference in wear and adhesion becomes more obvious?

The graphs are very short on information framing the tests.

This was likely performed on old school bias ply tires that weren't sensitive to camber the way modern radials are. I don't have my copy of miliken here to check the original publication date, but I suspect it's old enough that bias plys were used.

Further, a few percentage points difference it grip IS a big difference.

Further still, Ryan's picture is pretty awesome. It ought to be in that 'Essence' thread.

The main purpose of negative camber is to counter the body roll of the car in a corner. When I did the suspension upgrades on my own car I did some calculations on effective camber on the front suspension to see what was necessary to do.

The solid lines in the below diagram show camber variation as a function of suspension travel, and the dash dotted lines show the effective camber. To calculate the effective camber I've assumed that the car rolls symmetrically, so that the suspension extension on the inner wheel equals the suspension compression on the outer wheel. From that I've calculated the body roll of the car as a function of suspension travel. The effective camber is then calculated as body roll subtracted from the static camber. Compression is chosen as negative values of suspension travel, and zero suspension travel is when the car is at rest on the ground.

http://forums.pelicanparts.com/uploa...1233395014.gif

I've looked at three different front strut geometries.

1) Stock
2) 35 mm raised hub
3) 35 mm raised hub, and 1 degree of extra negative camber added to the hub

Geometries 1 and 2 have about -1.6 degrees of camber at rest, and geometry 2 has about -2.6 degrees of camber at rest.

The black rings mark the amount of effective camber at a cornering acceleration of 1 g, or about 9.81 m/s2 for two different setups

1) 21 mm / 30 mm front and rear t-bars with stock arbs.
2) Leda coilovers and custom made arbs

Setup 1 is used for the blue and red lines, and setup 2 is used for the green line. As can be seen from these graphs, it is almost impossible to achieve negative effective camber in a 1 g turn with a t-bar setup unless you use very stiff arbs or extreme amounts of static camber.

The importance of the effective camber increases with wheel size. A 16" wheel with 50 series side wall is rather forgiving since the side wall will deflect and help keep the contact patch parallel to the ground. On an 18" wheel with 30 series side wall on the other hand, the side wall will not deflect much resulting in a situation where you just use a part of the tire width to generate lateral grip.

/Peter

Chuck, I agree that the graphs are short on info. Thus my post, though it looks like noone has anything better.

Fig 2-25 is for a round section bias-ply. fig 2-26 is for a Goodyear Eagle, which I'd assume to be radial.

A few % change in grip might be huge to you - you drive a race car around a track, and are presumably quite skilled and experienced. I drive a street car, and I'm a lousy driver to boot. Would I even notice? As I stated, due to a track day with an accidental misalignment, I couldn't tell the difference between +0.5 (LF) and -0.5 (RF) degrees. My instructor also didn't seem to notice anything amiss...

Peter, I came up with the same conclusion - no matter what I do, I'm stuck with positive camber in a corner. Most of it is due to body roll. (man iti is hard to type with a crabby squirmy tired 2 year old on yer lap!)

Since most of it comes from body roll (camber gain of 0.7 and 0.75 deg/deg Front / rear by my figuring), I see 3 ways to reduce it:
1) Use huge torsion bars to get the car closer to an ox cart with no suspension
2) Use huge sway bars to get the wheels coupled together more like a solid axle (non-independent suspension)
3) Drive a swing axle suspension car, which has near 0 deg/deg camber gain (perfect camber compensation) - but other well know issues, of course.


So far Flieger has offered up a theory why a bit more negative camber might make a difference in actual track usage, and Chuck suggested that a few % is indeed significant (give me a few years - maybe I'll get there! :) ). But I've heard no fundamental disagreement that the difference is more than a few % ... so I guess I'll keep my occasional DE car close to stock alignment so I don't go broke buying tires.

No hard tech or fancy graphs, but maybe some practical thoughts.

A stock 911 in good condition is a very good handling car. Any single improvement is going to be fairly small, except for going to R compound tires.

Theory is fine and can be a lot of fun, but what do you want your car to do differently or better? There are plenty of people here who can point you in the right direction.

If you have understeer in slow corners, there are several things you can do (not least would be slower entries;)); Tail wagging? Same deal. If you have no specific complaint(s), perhaps you are not going fast enough to find them. That's not a criticism. But a car's handling faults don't really show up until you are approaching its limits (Okay, the wife's Infiniti shows much understeer at ridiculously slow speeds, but your Porsche won't).

Can't tell +.5 degrees from -.5? Pffft. Little wonder, -.5 is way conservative. IMO -1.5 degrees in front (with maybe -2.0 in back) would be a good starting point for a street car with occasional track use. That won't kill your tires, unless you drone around the freeways endlessly.

What will kill tires is toe-out. Why would anybody run toe-out? Better turn-in. If your turn in is fine, move on.

Yes, even with quite a bit of negative front camber, most cars, especially those with stock T-bars and sways, will have positive camber at full tilt. But, it will be a lot less that if you did nothing. So, making static camber more negative is useful, up to a point (a tire pyrometer can help find that point).

In auto-cross (my own obsession), wins are all too frequently by hundredths of a second. A change that improves lateral grip by a few percentage points is huge. Since there are no trophies in DE, not so huge there.

Now back to your regularly scheduled theory.:)

I like Peter Bull's comment that the 16inch wheels use tires with sufficient sidewall height and flexibility that positive wheel camber can be partially compensated for. Again, I am not saying the contact patch is level to the road, but it is closer than with the low profiles.

FWIW, here are some photos which may or may not have any useful visual information on camber in tight kurves with quick turn-in and weight transfer.:)

http://forums.pelicanparts.com/uploa...1233441772.jpg
http://forums.pelicanparts.com/uploa...1233441856.jpg
http://forums.pelicanparts.com/uploa...1233441867.jpg
http://forums.pelicanparts.com/uploa...1233441882.jpg
http://forums.pelicanparts.com/uploa...1233441907.jpg
http://forums.pelicanparts.com/uploa...1233441986.jpg

+1.

It became my new desktop as soon as I saw it.

Caster enters the fray when the wheel is turned off center. The inside tire adds negative camber and the outside tire goes more positive. Porsche designed in an agressive caster for this reason(amongst others). Add that to your charts.:D

I'm a tired and ready for Z's, but I think it might be the other way around - the outside tire adds negative camber, the inside tire gets more positive camber. Alas, I've been wrong often before ...

Yes, that is the correct caster/camber coupling effect. However, the outside tire is much more important in a turn because it has more grip than the inside tire due to the lateral weight transfer.

Ooops. I did say that backwards, didn't I. Oh well, I knew what I meant.:D

Here in Michigan, theory and a sub-freezing Garage is all I've got for 6 months of the year!

Ryan, that is an awesome picture!

If I'm reading it right, Milliken's second graph (the radial one) shows that radials are much more responsive to camber angle. This we know to be true.

For the 800lbs load line, it appears lateral force increased about 10% going from 0 to -2.25 degrees camber.

That is huge.

The more interesting and relevant part of the curve is of course to the left of the y-axis (ie, it isn't there), as most 911's are likely to end up with some positive camber on the outside wheel during cornering ...

A pic of back in the day when I was running about static -2 camber. Those old Victoracers liked that. It appears to me there is no lack of negative camber, particularly up front. .......http://forums.pelicanparts.com/uploa...1233513136.jpg

Just a thouthe, in autoX where the track time is so short there may even be a performance advantage to running less than idea camber on some types of tires to get at least part of the tire to operating temp.

I suspect that would not be the case with proper AutoZ compond tires on a warm day.

It looks like 70 aspect tires, soft side wall construction, much different compound than we have available currently, harrower tread, and at low air pressures relative to what we run now. I suspect these tires will be much more pron to roll over.

Remember, those tires want to roll a lot easer than what we run. It looks like there is a very steep curve as to adhesion on the 7" tire at -10 to -20 deg of camber. We may have as steep a curve on our tires from 0 to -2.5 deg.

Rough-Assumptions as to camber changes in a full turn.

1) Stock car lean adds about +4 deg.
1.1) Stiff torsion bar car leans about +3 deg.
1.2) Race coil over system (400/600?) leans adds about +2.5 deg
2) Stock rubber bushings flex adds about +.5 deg.
3) Camber gain under compression is about -.2 deg pre deg of lean.
4) Camber gain up front from positive Caster with turn is variable at ???.


Total stock carmcamber gain is about + 3.7 deg.

total full race coil over car might is about + 2 deg.

Front may be less in a sharp turn due to Caster.


---Is this close on a non-turbo 911 suspension?

Sounds reasonable to me ...
I think the effect of 4) is generally small. 90 deg of steering wheel ==> 5 deg of road wheel ==> around -0.5 deg of camber (for 6.5 deg caster angle) by my very simplistic and possibly erroneous figuring.

I am going to bet that is the origin of the suggestion of setting the front up with .5 deg less neg camber.

Probably more important on an AutoX car the sees more extreme steering angles. One a track car, depending on if the spindles have been lowered the Delta might even be able to be narrowed. :)

We still do not know what the ideal camber is for the wheel.

Do we want it square to the track, positive or negative and how much.

For reference, if the tread were a 9" flat and had not flex and we tilted the wheel 1 deg, one corner of the tread would lift off the ground about 1/8" to 3/16".

I wonder how rim width effects this. Seems if the rim width is the same as the tread width the tread would stay more square to the wheel. If the rim is narrower the tread would want to lift first on the inside???

Just some thoughts.

/Blushinghttp://forums.pelicanparts.com/support/smileys/nuts.gif

Thanks guys, it's been a long fun project. It's great to get some track time again, especially in a car that I built for it.

Yikes! My condolences. Out west we have 12 months of auto-x and 11 of DE so there always seems to be little enough time to work on the car, much less theorize about it.

Seriously, not knocking theory, you just have to test it on the track. I am reminded of an engineer friend of mine who calculated the maximum speed he could take a certain corner with on his motorcycle. He then proceeded to throw his bike away right in front of me because he forgot to take changing surface conditions into account.:eek:

One other thing to take into account is the tire in use. Hoosier, e.g., generally calls for more negative front camber than Kumho. Something to do with differences in sidewall stiffness I'd guess. Ideal static camber is not a simple question.

Take a look at a NASCAR racer sometime, they seem to run positive camber on the inside tire on the ovals - camber thrust?

Yes, different tires will have different camber requirements.

If we guess square to -1 deg negative at the wheel for max hold we need about the following static camber numbers to get there.

Stock. -3.7 to -4.7 deg rear.
Sport torsion bars and bushings. -2.4 to -3.4 deg rear
Full race suspension. -2 to -3 deg rear.

Front about .5 deg less.

This is a rough estimate only and subject to other variables like ride height, chassis stiffness, spring and sway rates...

This fits with Toyo's recommendations for there R888 and RA1 of -2.5 to -5 deg recommended static camber settings.

Just a some thoughts. :)

90 deg SW angle is a pretty tight corner for a track ... luckily the answer scales in a linear fashion :)

Agreed. I am thinking where we need our front bite the most up front is probably in those tight low speed corners on the track or when running AutoX.


Bottom line is to allow tires to work to there designed capacity on a 911 we require bushings, stiffer suspension, and some mods to allow more neg camber in front (strut top, ball joints, decambered spindle) and rear (modify lower slot for more travel).

As someone noted above, it is impossible to get enough neg camber on a stock suspension car to allow the tires to work to there full potential. This has to be accepted and worked with.

How can we maximize the handling potential of a stock suspension car knowing we can not get enough neg camber and the outside of the tire is going to have to do most the work?

Sticker tires?

Manage air pressure to get the center of the tire to work with the outside of the tire?

Do we look for stiffer or softer sidewall tires?

Do what we can to stiffen to front and loosen the rear to let the inside rear tire to do as much as possible? (Stiffer shock?) (Front v rear roll center management?) (front spoiler?)???

Take weight off the back? (headers, bumper...)

Starting to understand this graph and it is gold! Gold I tell you!

(Sorry in advance as my spell check crashed.)

I belive 4 deg lean is equal to -50 compression. His graph shows that starting at -1.6 deg camber on a stock spendle height it would end up at +3 deg or a total delta of 4.6 camber gain! Using Burgermisters info and making assumptions I estimated 4.7 (but included rubber bushing compression.)

Also showed that 21/29mm torsion bars moved from -1.6 and to about +1.5 for a total gain of 3.1 deg compaired to my est at 2.4 to 3.4.

Reseting of the spendle showed a solid improvement in camber by .8deg and reduction in roll by about 4 degrees.

The full race (coil over) set up with rased spendles started at -2.6 and ended up at -2 deg. of almost no loss in camber.

Good stuf here. It seem in the ball park of what we were thinking and showes that rasing the front spendle has a significant impact by reducing camber loss and decreassing roll.

This seems to be the standard starting point recommendation for "normal" cars.

But, in Bruce Anderson's book (at home, of course, I'll try to post the page later), he recommends nearly equal front and rear camber for the Turbo/Factory Turbo-Look cars. These cars have altered rear suspension geometry, wider rear track and wider rear tires.

On my car, front camber is -2.6 and rear -2.7. The car understeers some in slow (1st gear) corners, but is quite neutral in medium speed (50-60 mph, hey, it's auto-cross) corners. Of course, there are many other factors involved so this is hardly definitive.

Anyway, if you run a bigger than normal tire stagger and/or wheel spacers on the rear of a "normal" car, you might want to experiment with a smaller delta in camber front to rear.

Add: Re Peter's graph and your comment thereon - My car has raised spindles and my front tire wear does indicate I could run a little less negative camber. I'll be playing with the pyrometer when the auto-cross season starts in earnest.

I am not an expert but I think getting the camber right is more important with a wide tire.

Yes, the TL has the rear turbo anti squat and shorter arm for better camber curve. I think it is supposed to gain camber at about the same rate as the car leans. You will still see camber loss due to the bushings. I think he dose suggested less for the TL/930 and notes it is more sensitive to setting.

I think the normal and TL front ends are about the same except for a wider wheel base and they tilt the front end to dial in more anti dive. Redoing the front struts by raising the spindles might a great match the the rear on a TL.

Also if I recall, BA's camber recommendations were for a street car that did 'some' track work. I think he notes a race car may need more neg camber dialed in.

Tire stager should not effect ideal camber. It would effect ideal ft to rr spring rates.

Again, I am not an expert.

The only thing that gains negative camber at the same rate the car leans is a 356 ...

A turbo has around 1.5* the camber gain that a 'regular' 911 has (I think)

So instead of about .2 deg per deg of lean it is about .3?

My interpretation of the Turbo rear suspension

Yellow = NA 911 rear
Green = Turbo Rear
Red = Front

http://forums.pelicanparts.com/uploa...1233710181.jpg

I am thinking that we can get the most neg camber up front we need the A arms level to the track during a full on turn. This pushes the bottom of the strut out the farthest from the car center as is possable.

I seems that a 10.75" A arm is off parrellel to the track it is shorter as follows:

3 degrees --> .015" shorter
4 degrees --> .026" shorter
5 degrees --> .041" shorter

We lean abouit 4 deg in a full on turn with a stock suspention. Becaue of the relation ship of the a arm to the car it rotates about 20% more that the car its self.

If I calulated it right one deg of camber is equal to about .3" movement of the top or bottom of the strut in or out.

I think we need to set the A arm inner pivot about .75" higher at rest than the center point of the ball joint.

This would put the a arm very close to square to the track under a full g turn and deliver the most "effective neg camber" we can get.

If the A arm is set to get the most neg camber at rest w the a arm level to the ground we are can potentally lose over a full deg of effective camer in a full on corner.

Dose this make sense?

Boy did I misread the numbers. I thought .041 was .41!

The camber curve is very flat near parallel.

Camber should only reduce about 15/100's of a deg if the a arm is set parallel to the ground for a stock suspension at rest when making a full on corner. ;)