Bending rear axle for negative camber

[david-]

hello I've been watching this forum alittle while but didn't register before now
I'm from Norway and building myself a corolla RWD special saloon type racecar, very low weight and handling is my big priority, engine power isn't that important 1600ccm with approx 230bhp

however I understand bending a rear axle to get negative camber might be useful especially when using slicks, how much should I go for? 1-1.5degree?
I did find these pictures on the web



you can see they cut the axle, bend it correct way and weld it together, is this the best idea?

I need to know, if I use a straight edge like they used for alignment, how to calculate mm into degree's? I'm no math expert so I need help!!

is toe-in something I would want? or should I just keep it at zero?
wheelbase is 2400mm

I was also thinking of lowering the panhard rod bracket on the axle itself to get it level at ride height, the car is approx 20cm lower than stock so some modifications has to be done so the geometry will be correct again

I've also been searching for the mumford linkage the last few days, is there any kits or bits for sale anywhere? or do you have to construct everything yourself?

Thanks for any help!

[Jeff 8]

Wouldn't you need a CV joint to prevent the axle rubbing inside the housing and extremely short bearing life?

I think V8 Supercars run something like this or at least they used to but I think they had a CV.

Jeff

[JamesH]

I think you can get away with a small bend in the English axle without problems. The end bearing will always align correctly, and the fit in to the diff at other end is a pretty 'lose' spline anyway. The angles involved are pretty small.

I was accidentally running toe out on my Locost after an accident - we attempted to get the axle straight and thought we had but was still a bit out. Used a hydraulic ram and two axle stands at ends of axle rather than cut and weld. Ran for a couple of years I think with no problems. Apart from bad handling.

[terence]

I dont think you should be looking for any more than 1deg Neg,it will only wreck the hub bearings in no time at all.A very small amount of toe-in will prove beneficial but you should combine both at the outer end of the axle with one cut/weld. Sonny Howard of SHP supplys Engish axles ,ready modified,off the shelf.

[greenamex2]

I have a Corolla GT I race in the UK with slicks, 1700cc, 9500RPM, 218BHP (soon to be a bit more) etc.

We sorted through a pile of axle casing to find one that was suitably pre-bent and ended up with 1/2 a degree negative on both sides.

The tyre temperatures say we need a 1/4 of a degree more.

Unfortunately bending the axle has two downsides -

- The cars increase in mechanical drag is suprisingly high, it is measurably harder to push and you only have a little engine to push it with.
- The driveshafts in these axles are notoriously weak, this just makes it a lot worse. We snap a now almost unobtainable driveshaft every two years or so. And yes, we do have the later thicker halfshafts.

I have also converted the axle to a watts linkage.

The watts linkage, and the ability to relocate the roll centre had a far greater effect on handling than the extra negative camber did.

Because I am fed up with the unreliability of the standard axle I am having a new axle made by Elite (going to be cheaper in the long run). I am not bothering with any negative camber, spring rates, damper settings and roll centers are much better tuning tools.

[Al Weyman]

I agree with denis, IMHO not worth the hassle and unreliability issues.

[GORDON STREETER]

If you care to check (as Denis said) you will find that the manufacturing tolerances on rear axle casings are not that good (especially on historic stuff) and you can use that to your advantage. Most rear axle casings will bend to neg camber anyway after years of abuse on the road. I do run a neg camber axle myself but to get even 1 deg you would be surprised how much you have to bend it !
I remember David Yates bent his Mustang rear axle at Spa a few years back in a crash, and Norman Ricketts and I hack sawed the casing and assembled it with its wheels/checked it with a tracking gauge and welded it back together. Surprising what you have to do in the pits sometimes !

[esper]

I see a mig welder on the pics just some welds on top of the axle . shrink it on top . it works on axles of some English cars (I realy forgot which but half a degree it did get) and indeed the tolerances were crap by some standard ones I measured (a BMC car with a big six in front)

[david-]

thanks for all the answers, as I now understand I will not bend the axle, but if it's not straight I will correct it

another question I have since I'm no math expert, when measuring straightness into degree's, example: when using a 1metre straight edge from center, how many mm is 1 degree? how to calculate this?

have anyone ever used mumford linkage? if I'm going to build something better than a panhard rod I want the best solution, only thing I have against the panhard rod is that it's hard to get a low rollcenter without ruining ground clearance

Dennis: what brand/dimension tires do you use?
how big is big enough? i was thinking of using something like 225-255 575mm on 15" inch rims ? is 255 to wide when thinking about rolling resistance? approx 230bhp and 750-800kg raceweight?

[Al Weyman]

Sounds wide to me, I remember a guy I used to race against had an RS2000 Mk2 and went from 15 inch wheels to wider 17" diameter ones and went slower!

[terence]

[quote=david-;2424851]thanks for all the answers, as I now understand I will not bend the axle, but if it's not straight I will correct it

another question I have since I'm no math expert, when measuring straightness into degree's, example: when using a 1metre straight edge from center, how many mm is 1 degree? how to calculate this?[quote]



As a rough guide its about 2mm per degree.

[greenamex2]

Originally Posted by david-

have anyone ever used mumford linkage? if I'm going to build something better than a panhard rod I want the best solution, only thing I have against the panhard rod is that it's hard to get a low rollcenter without ruining ground clearance

Dennis: what brand/dimension tires do you use?
how big is big enough? i was thinking of using something like 225-255 575mm on 15" inch rims ? is 255 to wide when thinking about rolling resistance? approx 230bhp and 750-800kg raceweight?

I looked at a Mumford linkage and decided it was too complicated for the extra benefit. On the standard Corolla Panhard rod setup, if you have lowered the car a lot (about 4 inches in my case) it causes significant issues with lateral axle movement. You have to do something with it, evey if it is just relocating the body end. This is why we went watts linkage (we even re-used the chassis panhard rod mount).

Regs permitting you also need to convert to equal length rear links, especially if you go to rose joints. Otherwise you end up with mounting points ripped out of the chassis and all sorts of 'stiction' issues.

I am on 205/580 15" Avon slicks on 8" rims. I currently used hard compound (one set = one season) but would probably drop a compound or two rather than go up in width. You haven't got much power/torque to drag around wide tyres, not to mention the air resistance.

My car weighs around 800Kg. It is massively lightened but does then include a large fuel tank, FEA designed cage (AE86 bodies are pretty wobbly, plus I have dependants to worry about) etc.

[JohnD]

david,
I fear that terence's 2mm/degree may be out by a factor of ten. And you can't say one degree is 2mm so two degrees are 4mm; the relationship is not linear, though at this extreme end of the curve it is, more or less.

The answer is to take the bull by his trigonometrical horns. The ratio between the bend (distance from the straight) and the length from the midline is "Tan A" (tangent A) where A is the camber angle.
Divide the bend by the length from the mid point of the axle, and look up the result in a table of trig ratios.

For instance 2mm/1000mm= 0.002 The angle for which the Tan is 0.002 is 0 degrees 7 minutes (0.12 of a degree)


Using the Tan ratio, at a distance from the midline of one metre, the camber will be;

one degree with 17.5mm of bend (deviation from a straight line)
1.5 degree with 26.2mm of bend
Two degrees with 34.9mm of bend

Hope that helps.

John

[greenamex2]

In order to get 1 degree on a 800mm length (which is what I guesstimate the axle tube length is) you would need to bend the tube up about 14mm.

Calculation = 800mm * SIN (1 degree)

Which is a lot!

[GORDON STREETER]

Originally Posted by Denis Bassom

In order to get 1 degree on a 800mm length (which is what I guesstimate the axle tube length is) you would need to bend the tube up about 14mm.

Calculation = 800mm * SIN (1 degree)

Which is a lot!

I second that Denis !

[esper]

math gurus you wont bend the axcle in the center will you?

[jedrinck]

Born after the bronze age and well east of Detroit I have never owned a car with a live axle, but when I read this thread a passage out of a book called 'chassis engineering' by Herb Adams came to mind where he describes camber and toe adjustments on a Ford 9 inch by means of heat shrinking. So to gain negative camber one would heat the top of the tube which will initially expand and then contract on cooling more than it expanded on heating causing permanent deformation. There is even a number (just found the book). Mr. Adams says that if you heat a 6 inch section of the tube until it is red hot, it will shrink for about a 1/2 degree bend.

[JohnD]

Originally Posted by Denis Bassom

In order to get 1 degree on a 800mm length (which is what I guesstimate the axle tube length is) you would need to bend the tube up about 14mm.

Calculation = 800mm * SIN (1 degree)

Which is a lot!

Denis,

If your 800mm is measured along the hypoteneuse, then Sine is the right ratio, but you won't will you? At least not until after you have bent it. The original measurement, from centre to wheel will be the side of the triangle adjacent to the angle, so Tangent is the correct ratio. Of course, unless you are a maths pedant like me, you would do very well to measure the two distances as different for such a small angle, here the value of the two ratios is the same, but get beyond 20 degrees and they get very different.
John

[greenamex2]

Technically, neither is correct because you aren't dealing with a right angled triangle!!!

The reason I used sin is because the hypoteneuse length can't be varied significantly. Using tan you get some 'impossible' results much past 25/30 degrees.

Either way, 14mm out of alignment is a lot for a notoriously weak axle!

[greenamex2]

So, being pedantic about things (and because it is more interesting than work!) the correct formula is -

displacement = 2 * centreline length * sin (required angle / 2)

Which I make -

2 * 800mm * sin (1 / 2) = 13.96mm

Just getting the 1/2 a degree I have would require 6.98mm, and with that the car it noticeably harder to push!

[R59]

This is all far too technical.

Put the casing in a ruddy great hydraulic press, supported at the ends, and press down applying a little warmth.

And add 10minutes of toe in while you're at it.

Then buy a large quantity of bearings to replace them regularly before they fail.

Works on English axles.

The Ford (Strange) 9" axle on my Belmont runs 1/2deg neg. No toe-in (yet).It's a fully floating setup, so it works - which is perhaps why your Toyota didn't Denis?

The Winters Quick Change axle on my ASCAR has a number of tubes available for it to allow different camber combinations. For speedway use, we run 1deg neg on the outside, and 1 deg pos on the inside when running radial tyres. The tubes are made with snouts (where the hub bearings fit) that are fitted to facilitate the camber. They look they were made by someone who'd been down the pub on Friday afternoon, but the effect was spot on. Winters (and a number of other manufacturs) make them for up to 2.5deg camber. Any more than that and the axle shaft fouls the tube.

And as for pushing.... anyone who's tried helping me push the Belmont will know that it's a total beaaatcchhH! It weighs 100Kg less than the ASCAR, but feels like it's been left in gear with the handbrake on (which it doesn't have). I put some of it down to the Tru-Trac diff too.
I need a tow truck to get it TO scrutineering, as well as the usual recovery when it's died!

[greenamex2]

Originally Posted by R59

Then buy a large quantity of bearings to replace them regularly before they fail.

It's a fully floating setup, so it works - which is perhaps why your Toyota didn't Denis?

Not at £160+VAT a bearing you don't!!!! Another reason why it is cheaper for for me, in the long run, to go "bespoke".

Funnily enough Toyota homologated a fully floating axle for the Corolla. And also start quoting the most bizarre 'tolerances' for the standard axle. I am sure it was down to notoriously poor Japanese quality control rather than to gain a competitive advantage in Group A racing and rallying!

No wonder it's only competitor in Group A, the Escort RS1600i, never stood a chance!

[R59]

Originally Posted by Denis Bassom

Not at £160+VAT a bearing you don't!!!! Another reason why it is cheaper for for me, in the long run, to go "bespoke".

Funnily enough Toyota homologated a fully floating axle for the Corolla. And also start quoting the most bizarre 'tolerances' for the standard axle. I am sure it was down to notoriously poor Japanese quality control rather than to gain a competitive advantage in Group A racing and rallying!

No wonder it's only competitor in Group A, the Escort RS1600i, never stood a chance!

Going fully floating would be your answer. Don't the rules allow you to mod your axle to do that? You could them make your own bearing carriers to use common or garden SKF/GKN bearings, and save buckets of loot.

Or just graft in a baby Atlas?

[greenamex2]

Originally Posted by R59

Going fully floating would be your answer. Don't the rules allow you to mod your axle to do that? You could them make your own bearing carriers to use common or garden SKF/GKN bearings, and save buckets of loot.

Or just graft in a baby Atlas?

The problem with the Corolla axle is the half shafts snap at the diff end. Going fully floating means you have to replace the whole lot, it wasn't really the 'conversion' that the FIA papers claim!

I looked at the Atlas route but as I needed a custom width, wheel and prop PCD's etc it was just as cheap to get Elite to make one from scratch. Although they do seem to be exceeding their two week quote somewhat, pretty much in keeping with this winter and everyone else!

[zefarelly]

I've heard of people doing this and been offered the servicve for English axles, but personally the potential damage to the car, and potentially me, caused by a broken halfshaft made me think again. Given all the other restrictions etc I run with I doubt it would make much if any difference.