A bit of help please with rear camber gain

[DMSentra]

I'm designing my rear suspension and having trouble achieving the amount of camber gain I thought I should have. Using an unequal length dual A arm design with the front legs forward and the rear leg close to 90 deg. to chassis centerline. Upper arms horizontal length are about 5", lowers are close to 13".

Moving the arm pickup points around I can get to about .25 deg. of - gain until I really don't like the angles of the upper arms.
Is it a tough thing to gain .75 deg of negative with rear arms, or am I missing something simple?

[phoenix]

To do some calculations I need to know the vertical distance between the two outer pickups of the wishbones where they attach to the upright. What is this dimension?

Also, over what range of vertical wheel movement do you hope to achieve your camber gain? The less wheel movement you have the less opportunity there is to generate camber gain.

[DMSentra]

Vertical distance has been tried from 7" up to 11". Vertical movement should be held to 2".
Now that I state that, it seems pretty obvious where my problem is. It bothers me that I have never realized this rear gain limitation before. Happily going about the build thinking there will be no stumbling blocks I CAN'T work around has been my SOP. A mathmatic impossibility isn't what I expected. I'll be interested to see what you come up with...

[phoenix]

I used the following parameters in my calcs:

Vertical distance between the upper and lower outer wishbone pivots set at 7", the static lower wishbone set horizontal (inner pivot height = outer pivot height) and the top wishbone inclined at 8.6 degrees (inner pivot 0.75" lower than outer pivot).

Results:

I calculate the following camber changes over a 2" range:

0.5" 0.76 degrees
1.0" 1.05 degrees
1.5" 1.39 degrees
2.0" 1.81 degrees

After 2" of bump wheel travel, the upper wishbone will be inclined at just under 30 degrees. The lower wishbone will be inclined 8.8 degrees (upwards at the outer end).

Changing to 11" vertical between the upper and lower outer pivots, the maximum camber change at 2" of bump I get is 1.15 degrees. The smaller 7" vertical distance between the outer pivot points will always generate more camber compensation than a larger value.

By my calculations, if you can tolerate a higher rear roll centre by raising the inner pivot of the lower wishbone, you could get up to 2.45 degrees of camber compensation. If you can also tolerate a maximum upper wishbone angle at full bump of 41 degrees, you could have 4.76 degrees of camber compensation.

If you get different results, then we need to discuss why this might be!

[phoenix]

Glad I bothered.....

[jedrinck]

I care!

[dtype38]

Interesting stuff pheonix, but can I ask how you would relate that camber compensation to that actual roll of the car?

I ask because for my car I built my suspension to maximise steady state cornering by camber compensating the outer loaded wheels by "car roll angle + 1 extra deg" and the inner unloaded wheels by "-car roll angle only" The idea was to help the walls of the loaded tyres, but keep the max tread possible on the ground for the unloaded tyres. I'm just fiddling with them again this winter, so your thoughts would be welcome.

[DMSentra]

Originally Posted by phoenix

Glad I bothered.....

I'm very sorry. My susp. software magically became unusable, and some personal issues upset me enough I didn't care to continue. I apologize for seemingly wasting your time. I do appreciate your trying to help. I'll just have to put this on hold for a while. Sorry.

[phoenix]

Originally Posted by dtype38

Interesting stuff pheonix, but can I ask how you would relate that camber compensation to that actual roll of the car?

I ask because for my car I built my suspension to maximise steady state cornering by camber compensating the outer loaded wheels by "car roll angle + 1 extra deg" and the inner unloaded wheels by "-car roll angle only" The idea was to help the walls of the loaded tyres, but keep the max tread possible on the ground for the unloaded tyres. I'm just fiddling with them again this winter, so your thoughts would be welcome.

I will try to explain the relationship as clearly as I can:

If in a 1G turn your car develops two degrees of roll and you had no camber compensation, then 2 degrees of positive camber would be added to the outside wheel and two degrees of negative camber added to the inside wheel.

If in a 1G turn the suspension on the outside is compressed by 1 inch, then to have a neutralising effect - i.e. to keep the same amount of camber on the ouside wheel as you have when the car is static - you would want the suspension to add 2 degrees of negative camber on the outside wheel at 1" deflection and to add 2 degrees of positive camber to the inside wheel with 1" of suspension droop. The camber curve that DMSentra has would obviouly not satisfy this requirement, adding only 1.05 degrees at 1" deflection.

This is a very simple view and in the real world things will be more complex.
For example:

1) On a very stiff car there may not be enough suspension movement in roll to produce sufficient change in camber, so more static negative camber will be required to ensure that the angle of the tyre to the road is correct in a maximum G turn. DMSentra would need an extra 0.95 degrees of static negative camber to get the 2 degrees required at 1" bump.

2) Because of diagonal the weight transfer, the load added to the outside front suspension is likely to be greater than the load removed from the front inside suspension - i.e. the droop on the inside will be less than the upward movement on the outside - so the amount camber change required in droop will be less than that required in bump.

What you have done with your suspension would seem to be inline with this. However, at the rear of the your car you may find that you want the reverse situation. i.e. more camber compensation in droop than in bump, to allow for the diagonal transfer of weight from the inside rear to the outside front, and the lower transfer of weight from the front inside to the rear outside tyre.

If you know from design calculations what the are loads are on each corner in steady state cornering and what roll stiffness you have front and rear, you will be able to calculate the amount of suspension movement at each corner in a 1G turn and feed that information back into your camber compensation design.

I would be interested to hear what you find over the winter and what modifications you decide are required.

[dtype38]

Thank you for that. The bit about diagonal weight transfer hadn't occured to me so I'll have a good think about that before going too far with my changes.

I'm afraid my design calculations are a mixture of maths, educated guesses and empirical data (I work out my suspension displacement by the tyre rub marks on my wheel arches!) I think I've actually got quite compliant suspension for a race car, with a good 3" of wheel travel (1 1/2" up/down) so getting camber change is fairly easy. I also very much doubt I'm getting any where near 1G lateral in the turns.

As for roll stiffness, is the same/related to wheel-rate? I have actually calculated my wheel rates and found I'm using a ratio of 8:5 front:rear. Its a compromise between sufficient support at the front for turn in, and compliance at the back for traction out of corners. But then I've always been happy with a nervous front end, and being short on power I've always wanted to stamp on the loud pedal early in a corner and have stay stuck.

I'll be sure to report back on what I come up with... on previous experience one thing's for sure, it won't be what anyone else is using!! :-)))

[phoenix]

Originally Posted by dtype38

As for roll stiffness, is the same/related to wheel-rate? I have actually calculated my wheel rates and found I'm using a ratio of 8:5 front:rear.

The short answer is no, as roll stiffness (via springs only) is a function of wheel rate and track width.

The amount of weight transferred to or from each wheel is dependent on a number of other factors besides wheel rate - anti roll bar rates (if fitted), the sprung weight of the car, the weight distribution front/rear, the force of the turn, roll centre height, centre of gravity height and the track width.

Only part of the weight will be via the springs, causing suspension movement. If your roll centres are above the track, some of the weight transfer will be via the suspension arms and any weight transferred via anti roll bars will not cause the springs to compress, as the weight is transferred directly to the wheel via the suspension arms. One effect of adding or stiffening bars, therefore, is to reduce suspension travel and so opportunities for camber correction, though there should also be less requirement for camber compensation due to there being less roll!

The unsprung weight of each wheel/hub/brake assembly, it's own centre of gravity and the G force of the turn also add to the total weight transferred to or from each wheel.

Incidentally, anti roll bars also effect the movement of the suspension on the unloaded wheels - the stiffer the bar the less droop you will have on the inside wheel, so this can complicate calculations for camber compensation in droop.

Our car has limited options for control over camber compensation at the rear. As theses are the driven wheels, we set the camber to (zero camber - maximum roll angle) at full droop to ensure we have the best tyre/road contact patch when it is most needed.

[phoenix]

Originally Posted by dtype38

I have actually calculated my wheel rates and found I'm using a ratio of 8:5 front:rear.

Another 'incidentally'.............

Roll stiffness is usually expressed as a percentage, rather than a ratio.

In your case, your wheel rate stiffness is 61.5% front and 38.5% rear - and I guess if you have calculated this from static deflection of the springs it allows for the weight distribution of the car? If you have no roll bars and equal front and rear track widths, then this distribution is probably pretty close to your roll stiffness distribution.

We have a mid-engined car with a slight rearward weight bias and run 59% front and 41% rear roll stiffness distribution. This is pretty close to your figures and in our car this results in understeer occuring at the limit of adhesion before the rear end loses grip and allows for early application of the throttle on corner exit with good traction.

[dtype38]

You're talking at the limit of my understanding here, so thanks for keeping it in laymans terms. Anyway, I calculated the wheel rates using the spring rates and a bit of trigonometry from the position and angle of the spring dampers relative to the wishbones and the tyre contact patches. I verified by calculating the spring compression from my corner weights and they tied to better than 1/16" on the actual springs which is within about 5-8%.

Anyway, as you guess, I'm not running anti-roll bars, front rear track is equal, and my weight distribution is actually F-R 50.2%-49.8% which I measured with corner weight scales (although my front engine is the full height of the car so I think the effective weight at the front is higher than at the rear by a couple of inches). The car weighs about 2300lb in race trim of which I'd say about 350-400lb is unsprung, so hopefully that's not too big a factor. I'll have to do a bit more drawing and measuring to find out where my roll centres are. Are they important?

Oh, btw, sorry for the thread hi-jack DMS.

[phoenix]

Originally Posted by dtype38

I'll have to do a bit more drawing and measuring to find out where my roll centres are. Are they important?

Yes, they are important.

The distance between your roll centre height and your centre of gravity height determines how much of the roll is controlled by your springs.

Extreme examples are:

1) if the roll centre height is the same as the centre of gravity height, the springs will play no part in controlling roll - all resistance to roll will be through the suspension arms and the car will have zero roll, so no vertical load will be added to the tyres - which is bad for grip - and there will be lots of side loading on the tyres, which is also bad for grip.

2) If the roll centre height is at ground level all the resistance to roll will be via the springs and none via the suspension arms. All roll loads will be vertical as far as the tyre is concerened and there will be no side loads.

A rolll centre can also plot to below ground level, but I won't go into that now!

[dtype38]

I better get measuring and plotting then :-))

[dtype38]

Me again

Not sure what units you prefer phoenix, but we've started in old fashioned ones so I may as well continue. If you prefer metric just say. Anyway, far as I can figure I have the following:

Front/rear track, 57" at centre line of tyres.

Front roll centre is approx 6" above road which puts it about 1/2" below the lower inner wishbone pivots.
Rear roll center is approx 3 1/2" above the road and 3" below the lower inner wishbone pivots.

I don't have an accurate figure for the CoG height, but my best estimate is around 15" high, give or take and inch.

Are these good, bad or indifferent?

[phoenix]

It is more common to have the rear roll centre higher than the front - or the front lower than the rear. Does the car feel a little weird in roll, like the rear outside is falling and the front inside rising?

Either way, if you can lower your front roll centre that would be good for front grip as there will be lower side forces on the tyre, leaving more for cornering grip.

Would you let me know your front and rear wheel rates to do some more calcs? - or are they hush-hush? If they are PM me.

[dtype38]

Originally Posted by phoenix

It is more common to have the rear roll centre higher than the front - or the front lower than the rear. Does the car feel a little weird in roll, like the rear outside is falling and the front inside rising?

I'm not sure I have the correct terminology to answer that. When I'm cornering I'm not really aware of any "roll" as such, I concentrate on my fore-aft weight balance and yaw. Not sure if that makes any sense so I'll try to describe what goes on in say a big fast corner...

I don't trail brake into a corner, aiming for a neutral or slightly rear-ward weight balance for initial turn in. I run into the corner quite fast and "encourage" the front to turn. As soon as I get inital rotation I add power and feel for under/oversteer. I use as little steering input as I can from this point, keeping control of the yaw (angle of car to track?) with my right foot. Once into the corner proper I gently coax the front end in toward a slightly earlier than optimum apex aiming to approach that point with slight oversteer. As soon as that's set up I plant my right foot firmly to the metal and push the front end as hard as I can. If all goes to plan, I can straighten up the steering and let the whole kit and kaboodle do a four wheel wotsit to the correct late apex point and keep it there all the way to the exit kerb. I'm not sure its the prettiest way to take a corner, but to be honest, my whole plan is aimed at maximum exit speed. With the competition having more horsies under the bonnet, its the only thing that keeps me in the hunt!

Originally Posted by phoenix

Either way, if you can lower your front roll centre that would be good for front grip as there will be lower side forces on the tyre, leaving more for cornering grip.

Initial turn in on fast entry corners could be a lot better. Sometimes I think she turns in on my will power alone! I'll have a look at the geometry and see if that can be ajusted.

Originally Posted by phoenix

Would you let me know your front and rear wheel rates to do some more calcs? - or are they hush-hush? If they are PM me.

The springing gives me rates at the rims of front-595lb/in, rear-385lb/in but I don't know what the 225/60/15 tyres do to that. If I wobble the car, most of it is tyre walls!

Anyway, all advice gratefully received.

Ps dunno if it makes any difference, but my wheelbase is quite short at 90"

[phoenix]

I had estimated 500lb front and 320lb rear from the 1.5 inch defelction you mentioned - not too far out. I will take another look with the corrected rates.

[phoenix]

I have guessed that your unsprung weight front and rear are the same at 200lbs - this may not be correct but it will do for now.

I have used the other figures you have given me for weight, wheel rates, c of g height, track and roll center heights.

In a 1G turn (the car should pull that on decent tyres and it is a good number to work with for this illustration) at total of 590lbs will be transferred from the inside wheels to the outside wheels. 353 lbs is transferred at the front and 237 lbs at the rear - 59.8% front and 40.2% rear.

A total of 90lbs unsprung weight is transferred - which can only be reduced by losing some unsprung weight.

This leaves 500 lbs out of the total 590lbs being transferred via a) the springs and b) the roll centres.

Of this 500lb, a total of 159 lbs is being transferred via the roll centres - 101 lbs at the front and 58 lbs at the rear. The 101 lbs at the front is a side load on the front outside tyre, and so is taking away some of the grip capability of that tyre.

If you lower the front roll centre you will improve the front end grip by reducing or removing this side load, particularly on turn in, as the side load of 101 lbs is transferred immediately you enter the corner via the suspension links, whereas load on the springs (which increases grip) takes a while to transfer, depending on front damper stiffness. If you were to lower the front roll centre to 0 inches, instead of the front begining to slide at 1G it will not slide until you are pulling 1.125G - a 12.5% improvement! If you can only pull .9G on your tyres you will be able to pull 1.01G - still a 12.2% improvement.

By lowering the front roll centre you may find the car a little more over-steery than you like, as the weight transfer would become 7% less at the front (53%) and 7% more at the rear (47%), with only 17lbs being transferred from the rear to the front at 1G. This could be dealt with by softening the rear end slightly, which would improve traction, or adding a front anti roll bar.

One other thing - with the wheel rates you have, your wheel frequencies seem extremely high for a car with no downforce. Do you suffer from any effects like the front tyres pattering under hard braking? Or you fillings falling out?

[dtype38]

I've read that twice and with my sunglasses on I think it all makes sense.

Only slight flaw in your assumptions is that I have outboard front brakes but inboard rears, so front end unsprung weight is about 50lb more than at the rear. as for reducing that, I've got aly alloy rear uprights and the front end the disc bells and callipers are also alloy, so I'm afraid there isn't a lot to be had there.

Regarding the roll centre heights, I only really have ride height to play with to adjust them because the rules of the series I compete in require standard geometry for the wishbones and mounts. As it happens though, lowering the whole car lowers the front roll centre by significantly more than the rear. If I take a grinder to the sump I might be able to get the ride height down far enough to get them at least more equal.

Originally Posted by phoenix

By lowering the front roll centre you may find the car a little more over-steery than you like, as the weight transfer would become 7% less at the front (53%) and 7% more at the rear (47%), with only 17lbs being transferred from the rear to the front at 1G. This could be dealt with by softening the rear end slightly, which would improve traction, or adding a front anti roll bar.

Less under-steery going into a corner is good, more over-steery coming out is definitely a no-no. But may well be worth a go at getting a little more positive at the front end and doing the softening the back a bit (wheel arch clearance permitting).

Originally Posted by phoenix

One other thing - with the wheel rates you have, your wheel frequencies seem extremely high for a car with no downforce. Do you suffer from any effects like the front tyres pattering under hard braking? Or you fillings falling out?

Ha! I spent five years in karts before getting a car. To me this thing feels like a big squishy wobbly thing Just don't let it hear me saying that ok?

More seriously though I used Fred Phun's book to do my basic setup. I stiffened my springing to the point you describe, using minimal damping to allow the tyres best chance to maintain contact, then backed off the spring rates by about 5%. It handled great in the dry, but sadly was a bit of a disaster in the wet. I needed a bit more of a compromise so at the mo I have the springing backed off another 7-8%. That has meant raising the ride height to stop me fouling the wheel arches but it's much better in the wet and not tooo much slower in the dry.

Tell you what though, I think the audience at home are probably getting a bit bored with my suspension, so I'll PM you a bit more info and go from there.

BTW, I'm very happy you're bothering. Thank you

[phoenix]

Originally Posted by dtype38

Less under-steery going into a corner is good, more over-steery coming out is definitely a no-no. But may well be worth a go at getting a little more positive at the front end and doing the softening the back a bit (wheel arch clearance permitting).

With more front end grip - and more certainty on turn in - you might find you can turn in at higher speed for a later apex (the apex you hope to 'drift' to currently) with more precision and then get on the power. Even with less power than 'the rest' this should be a quicker way through most corners....

BTW, the higher un-sprung weight at the front only makes the front roll centre mods more important, sadly.... You will be only able to pull 10% more G......

I take what you mean about 'chatting' here further - I am more than happy to exchange PMs.