Camber Curves of Front vs. Rear Suspension 5

[mark512]

I haven't managed to find a lot about this online - there seems to be plenty of information regarding "setting up" a particular vehicle for a specific purpose (street, track, drifting, etc.) but these discussions are mostly limited to adjusting the static cambers of a particular car.

What I'm looking for is the more general case: in an arbitrary vehicle where the static cambers and camber gains of the front and rear suspensions can be arbitrarily chosen, what are the factors affecting their selection, but in particular, what affects the choice of variation between front and rear?

For example, it seems to me that if there is a vehicle with 50/50 static weight distribution, double wishbones and identical tires on all four corners, why not have the camber curves of the front and rear suspensions be identical? I'm aware that camber angle with respect to the road affects tire grip and camber thrust and therefore affects understeer/oversteer, but there also appears to be many other ways of adjusting this balance other than by having different wheel cambers.

Thanks in advance!

 

[GregLocock]

Are you talking about production carx or track cars?fwd or rwd or awd? Roughly speaking a lot of received wisdom is based on crossplies which are far more sensitive to camber than radials.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[mark512]

In this case more of a production car, and RWD, but not excessively powerful (in the order of 0.5 G maximum acceleration.)

And yes, I've noticed that with a lot of the older literature they're assuming you're using bias-ply tires, drum brakes, etc.

 

[GregLocock]

In the ideal case forr a production car the static camber setting is optimised for minimum tire wear. Camber gsins are often an outcome of suspension architecture and desired RCH. The front suspension also has a limitation that camber gain affects bump steer, so if your rack is in the wrong place you may have to modify the camber gain to get the desired bump steer. There is also an upper limit to camber gain as gyroscopic precession causes feedback in the steering.

Since the most common front suspension has a fairly awful camber curve I think it is safe to assume that camber gain is not an especially crucial curve, at least for linear range. For the driven axle the ideal camber gain is sufficient to keep the wheels square to the road, for traction.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[mark512]

That's the feeling that I was getting - that camber gain is more of a "design output" than a "design input" as in the better approach might be to consider RCH, RC migration, bump steer, steering geometry (KPI, scrub), packaging constraints, etc. first and given these, if the resulting solution's camber gain seems reasonable, then that's it.

"For the driven axle the ideal camber gain is sufficient to keep the wheels square to the road, for traction" - that's also what I've been looking at, and with regards to static camber in the alignment specs for performance RWD cars, it seems like there always is more negative static camber called for in the rear than the front.

 

[cibachrome]

In God We Trust. All others bring data...

 

[LionelHutz]

The C5+ Corvettes are double wishbone all around and they don't spec more negative camber in the rear than the front. Just the opposite.

 

[mark512]

Cibachrome, thanks for the data! So if I understand this correctly, this is degrees of camber change with respect to the road, of the outside wheel per degree of roll?

It makes sense to me that the trucks have a rear value of near 0 deg/deg because they probably have live axles, and the front values are nearly all positive as the shallow camber curves of typical front suspensions (as I understand, and as Greg pointed out) make the front wheel lose negative camber with roll.

What I'm having trouble understanding is the fact that practically all of the rear suspensions (that aren't live axles) seem to have a net negative camber gain with roll. With a track of 1500mm and a roll angle of 5 degrees, this corresponds to about 65mm of wheel travel on each side. That seems to imply a camber gain of in excess of 2 degrees per inch of travel, or 75 degrees per meter, whereas when I've looked at camber gain per unit of linear travel, I haven't yet seen anything in excess of 1 deg/in or 40 deg/m...

 

[BrianPetersen]

... most of that 2 degrees per inch being accounted for by the body roll. Suspension with small camber change relative to the bodyshell means the camber will be close to that of the angle change of the bodyshell itself.

Picture what happens with pure trailing arms ... the camber equals the body roll.

 

[GregLocock]

I think the main reason why irs rwd cars do not have 100% camber/roll compensation is that it implies a low rch, so you need a big rear sta bar to control roll. Heavy, hard to package,

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[cibachrome]

IRS RWD cars with high HP motors need to maintain REAR scrub radius as a result of tire Mx changes under Fx and Fy application. The best way to confirm this is to find tires of same size and compare Mx at low Fy-slip conditions. Run them on the car and they will talk to you. If you loose the static negative scrub because of the tire Mx shift to positive, its a new dance. Try wheel offset, too, if they package/fit, but this affects toe stiffness and that you won't like, either.

Same for FWD, high power motor front suspensions. Broken wrists from "Torque Induced Steer".

 

[cibachrome]

Here's some data to contemplate:

Pneumatic scrub is a term used to represent tire Mx normalized by Fz. Hence it's a dynamic scrub radius modifier. It relates actual contact patch line of action with the geometric one. If you want to maintain your negative scrub radius for longitudinal criteria, you engage Mr. Purple or Mr. Red. If you want the car to be a handful, get get Mr. Green or Mr. Purple-Dashes, assuming your car already has a nominal but not gigantic amount of negative scrub at the spindle. Another easy science project can be reversing the tires on the rims to swap the signs of P-Scrub. But, be watchful of conicity and ply steer details before you grab that "ahaah" moment. Uni-directional tires may be players in this phenom.

 

[mark512]

Thanks again for the data! The conversation started with camber curves but has moved to the amount of negative scrub radius that an IRS on an RWD car would need - would you therefore consider keeping scrub radius negative under all circumstances to be more essential than having ideal camber curves?

 

[Isdtbower]

Good example of both conditions. This is pretty common in the offroad community where fabrication may be more revered than geometry. 39"tire, 25 psi, 17" rim. This allows lots of tire to soak up rocks, etc at speed. Yes he went for a ride.

 

[mark512]

I was doing some more rummaging around these forums and found this from '04:

http://www.eng-tips.com/viewthread.cfm?qid=98416

Is there a method for setting camber compensation? I suppose one it would be a strong function of the vehicle roll gradient (and the time distribution the car spends in corners or on straight aheads, although this would be quite area dependent). I'm not sure how to set a target value though. Any advice?

Well, that is a damn fine question. In practice it is pretty much set by tyre wear considerations, but if we didn't have to worry about that it would be nice to keep just a little negative camber on the maximum corner.

So, with a vehicle roll gain R deg/g, it would seem reasonable to set the camber gain C=R+0.5 deg/g

In practice this does not happen. Typically C=~0.5*R

That last equation seems relatively consistent with the RWD independent suspension production car camber gains that I could find on the internet, approximately 1 degree of camber gain per inch of travel, and with typical track widths, an inch of travel corresponds to approximately 2 degrees of roll.

 

[GregLocock]

To give you some idea on this from a design perspective for the front axle no alarm bells would ring if the camber gain fell anywhere in the range 20-30deg/m, whereas say bump steer would normally be maintained within 0.5 deg/m of target. I can't remember the equivalent bands for an IRS, as I've been working on live rear axles or twist beams for whst seems like a decade!

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[cibachrome]

Put on your seatbelt and pull it tight, because a topic to entertain the curious has risen that's always been a favorite of mine: "What should my 'Camber Curve' be "? If you have a clean sheet of paper and access to tire data that you believe, you can take a stab at this problem in an unusual way. When you get to play with a tire on a tire test machine, it becomes very clear that the Mx data channel gives you insight to the tread surface conditions, belt stability, tread life and Fy output. In manual control, you can hover the tester at inclination angles necessary to give the best heat distribution, the best sound ("frying bacon" in a race slick) and a high Fy at an array of pressures and Fz loads. Rim width gets into this, too.

You create a 3D function of Mx as a function of slip and inclination angle at a nominal Fz load. We want to take slices of this plot to determine where the Mx is zero for a slip angle or two in the range of Fz’s your vehicle is likely to generate at high Fy levels. The slices across slip angle are meaningful.
Next, we fit these Mx slices to another tire friendly (I'm a spline fan) function and calculate the zero crossings. Naturally, Matlab provides all the necessary service function, so this whole deal takes more statements to make the pretty plots than it does to do the math, once the data is fitted appropriately. Then we contemplate the graph of these inclination (camber) angles vs. the Fz loads that produced them at a few slip angles. A decent vehicle dynamics model of your car should already have given you approximate 4 wheel loads and a roll angle at your limit of control (whatever that means to you). In the case of my limited knowledge of the car you are dreaming about, all the wheel loads in the Fz range (inside, outside, front and rear) would be set in the middle of the slip zone contemplated.

Bingo, this ought to produce "Happy Tires" as I call them: tread hot and like chewing gum, temps uniform, and the tread band pealing of slowly instead of chunking, delaminating, marbling or smoking. (Sounds like a steak cooking, eh ?). (for track tires that is). For dirt tires, you need trough data, Camber Ridge data, or load wheel data (if anyone is crazy or rich enough to spend money running load wheels in muddy/dirty conditions).

But, words of caution: Your camber curve is important in roll for cornering maneuvers. That means some roll angle from suspension travel is necessary (to make use of your camber by roll curve). Roll implies suspension travel and that's the x variable in your camber curve. (I prefer it to be the Y axis, but there is convention, you see.) AND, (and this is a BIG and), the lateral camber compliance (from bending and soggy wheel bearings, etc.) plays into this construct as well as caster angle (caster is camber by steer, eh ?)

If you only optimize the front, your extra grip will make the car very loose. Your driver will complain. If you have the freedom to do the front and rear, you will get some extra Ayg's but stability may not be improved because there are other features of your chassis which are not conducive to this (Like weight distribution or TLLTD or a solid rear axle).

But, it's all better than the usual hand waving, smoke, mirrors, packaging constraints, lack of time excuses, blank stares and legacy explanations I've heard so much about over the years. This process makes for some awesome plots, too. Very impressive in a design review with Management who know nothing at all about the subject !

 

[Isdtbower]

Thanks. Great post for me.

The practical Camber/Caster bandwagon. Any ideas to make us think and measure is helpful. In my world with no testing facilities, it is real world of pictures and video.

This last weekend I showed a picture similar to the one above... to about 20 offroad racers, industry leaders, manufacturers, tuners, fabricators, suspension designers, media, etc. Only two picked up on Camber and secondly "pneumatic scrub." It made for great conversation, and I learned a lot of other tricks that you didn't know that you didn't know or see.

From your discussion, I see that the "car" above is being supported well into the sidewall...and not the tread. So I guess that helps the driver from rolling so easily.

Interesting: I recently saw on a prototype tire a longitudinal strip of full height "rubber" on the line between the sidewall and tread. I guess that is to keep the lugs together on roll or against something. The sidewall lugs were all placed to easily side slip....and maybe not make the "pneumatic scrub" worse. Someone is not talking to others. Basically the shock and tire guys are just trying to fix what the suspension designer missed....or ignored.

 

[gruntguru]

So I guess that helps the driver from rolling so easily.
Yes - by "un-gripping" the tyre - on a race car.

Surely it would be preferable to keep the tyre upright and somehow reduce the massive migration of the outside wheel contact patch towards the front view CG? Can't help thinking a lightweight beam axle would do a better job.

je suis charlie

 

[Isdtbower]

"Surely it would be preferable to keep the tyre upright and somehow reduce the massive migration of the outside wheel contact patch towards the front view CG? Can't help thinking a lightweight beam axle would do a better job."

Not in the OP's question, but a beam axle does do a better job offroad until the speeds increase. The straight axle always/usually places the tread flat with the ground...no matter the body roll or terrain. There is also the benefit of turning tighter with a straight axle. (45*) And is generally stronger. Lightweight is not in the vocabulary as axles can be up to 500 lbs to prevent breaking. 40 spline axles (1.7") are becoming standard in rigs even weighing less than 3000lbs. Air pressure can be dropped to below 8 lbs and more with no detrimental effects on "offroading." Competition tires are filled with lead or water and weigh in at 300-500lbs ea. From competition you can go all the way up to OEM which do perfectly well for the intended environment.

At some point this changes to the advantage of a IFS suspension. (This is also debated in the offroad community) The faster you go the more critical steering becomes. Articulation of 15-25" in the front usually means full hydraulic steering which is not "perfect" for high speed, and has no centering. But there have been significant improvements in this lately. IFS has mechanical steering so it is precise....or as good as you can design it considering bumpsteer. There are other positives at speed (3' whoops at 80-100mph) but also the negatives such as this thread about Camber curves. Driven IFS or 4x4 IFS is relatively new to high end racers needing lots of articulation. Areas of design that might have been ignored are now as important as the traditional. Amazing how most say that 4x4 IFS is a "compromise" as a "legacy explanation". (I love OEM techi wording sometimes)