Caster on rear independent suspension

[rcx194]

I spent ages asking questions, reading and learning how to design independent suspension geometry (thanks again to Greg Locock from this forum who helped a lot), I started with the front and modelled it all in Solidworks and suspension analysis software. Then I just duplicated the same for the rear..... and then moved on to other things. It's just dawned on me though, I can't see why the rear would need any caster? Or would it need some, but less caster, just to reduce load on suspension links? From what I understand the caster is mainly/purely for self centering the steering?

 

[SwinnyGG]

Caster is controlled to control steering dynamics, sure- but caster in the rear can have multiple effects, none of them bad **IF** they are accounted for in the design phase. For example, rear caster has the potential to effect rear bump steer, and rear toe compliance can have positive or detrimental effects based on rear caster. It's not something that should be set without consideration.

 

[benta]

Caster on rear wheels make little sense, unless they are also part of the steering (like BMW 850CSi, for instance).
Otherwise, offsetting the wheel to the back or the front could be to make place for drive shafts etc.

 

[BUGGAR]

Probably not relevant but if you have a lot of play in your suspension with positive caster (or is it negative?), your wheels will fight to turn around and wobble like a VW with loose torsion bar link pins.

 

[BrianPetersen]

Plenty of rear suspension designs don't have anything resembling a "steering axis" defined between upper and lower ball joints. Lots of multi-link rear suspensions have arms pointing in various directions but not really defining what you would normally consider to be a "steering axis". With no upper-and-lower-ball-joint "steering axis", there is no defined "caster".

 

[MikeHalloran]

Second generation Corvairs had a 'toe link' on the rear suspension's trailing arm. Alignment techs thought it was for their convenience in setting rear toe.
It was normally level-ish, and clearly intended to provide rear roll understeer.

Mike Halloran
Pembroke Pines, FL, USA

 

[GregLocock]

Brain - kinematically you can define a thing that looks like a steer axis for any suspension in the real world (apply a torque at the contact patch about the vertical axis, what axis does the wheel rotate about) , and that axis will have a side view angle to the vertical, even if it is zero. But since the primary effect of castor, kinematically, is to camber the wheel in response to steer, and the steer angles at the rear are (hopefully) tiny, it rapidly becomes an Oozlum bird. I don't even bother reporting on it when we prepare a new set of hardpoints, there again, I've been working on twist beams and live rear axles for the last few years, which certainly fit your description.



Cheers

Greg Locock


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

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

 

[cibachrome]

Rear caster (transcribed from the front axle) would likely add a considerable amount of lateral force compliance oversteer to your net rear cornering compliance recipe (increases the rear axle sideslip gradient magnitude). That's NOT considered good by any means. Adding NEGATIVE caster to the rear on the other hand would likely add some lateral force compliance understeer. You will like the transient response better, but it adds understeer to the whole vehicle, too, if the front remains unchanged. That means the steering gain goes down as well as some other things. A good crutch for those stuck using the same tires front and rear and also a close to 50% front weight distribution. And you should be pulling on a chain with the outside tire, as they say, instead of pushing on one, if you get my meaning).

 

[140Airpower]

One of the effects of caster is an effect called "trail". The steering axis does not have to be angled to get this effect. It occurs when the vertical rotation axis is in front of or behind the spindle to contact patch axis. This kind of geometry can introduce compliance steer and that can be used to tune the chassis response to steering inputs.
In particular, sudden steering inputs can result in small oscillations at the rear due to tire compliance. This is most apparent with a stiff suspension. If two of the lateral links in a 5 link system attach to the upright at vertically aligned points this establishes the equivalent of a king pin axis. Then the third lateral link can control rear toe and can be compliance mounted and tuned to eliminate the oscillation, leaving a smooth step response.

 

[BrianPetersen]

True and many designs are like that - but the location of where those lateral links join the knuckle is often nowhere near where one would put ball joints in an upper-and-lower-A-arm-type front suspension, and they need not be!

 

[rcx194]

God these projects take time.... I started this topic back in August and I'm only just sitting down to finalise the rear suspension design.

So I don't get a lot of the complex stuff in this thread, but basically I'm understanding "no need for caster at the rear". My plan is to make the rear geometry same as the front, so that the front and rear aren't doing different things. Except at the rear there won't be any caster.

 

[Tmoose]

OP recently said "My plan is to make the rear geometry same as the front, so that the front and rear aren't doing different things. "

I'm nowhere near an expert, but I'm thinking different front and rear camber curves, toe in curves etc may have been used to advantage on vehicles with any weight bias, whether going straight or cornering.

 

[BrianPetersen]

I'm nowhere near that expert, either, but the desire to preserve some measure of understeer in the interest of keeping the vehicle stable usually results in different geometry front to rear.

Common arrangement: MacPherson or double-wishbone front with little or no camber gain with body roll, and beam-axle or some variant thereof in the rear that keeps the rear tires more square to the road with body roll.

You can still do this with double-wishbone front and rear (I presume this is what the original poster is doing); there's nothing saying the attachment points to the chassis have to be the same front to rear, and there's nothing saying you can't design in some roll understeer and/or compliance understeer in how the rear "steering arms" are attached.

 

[cibachrome]

If you feel obligated to use the same suspension pieces front and rear, your better off installing the rear backwards so the K&C effects produce an understeering vehicle. BTW: any rear K&C steer effects are not going to save your ifs, ands, or butts in a severe limit maneuver because tires don't hear you any more when they are approching peak slip angle. Maybe camber (depending on the specific tire construction, wheel and pressures utilized). Only Fz works to hold the back end in place. That would be a working PAIR of tires, too, however you want to configure them, including a dog at a hydrant scenario..

 

[GregLocock]

One way to think about it is that the driver controls and senses the front suspension directly, whereas the rear suspension is sensed via the time lags introduced by compliances and inertias. So, one might imagine that a very neutral rear suspension would be better than one that replicates the passive control seen in the front suspension. In addition a rear suspension in a RWD car needs to be compromised heavily towards traction.

Cheers

Greg Locock


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

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

 

[rcx194]

Thanks everyone. Greg the car is a 4x4.

If you feel obligated to use the same suspension pieces front and rear, your better off installing the rear backwards so the K&C effects produce an understeering vehicle

I'm not obligated to use the same parts at the rear, they have to be different as the differing castor will result in different sized wishbones. Plus being at the rear, the rear wishbones would be a mirror of the front ones.

On the same topic. I'm using 40mm scrub on the front in order to get some feedback from the road. This is probably also not necessary at the front so should I switch to centre point at the rear so that the kingpin goes through the centre of the tyre contact patch?

 

[GregLocock]

Scrub radius at the rear is another thing that really doesn't matter much. You could use it to give some toe in on throttle I suppose.

Cheers

Greg Locock


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

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

 

[BUGGAR]

The car is a 4x4 - what mechanicals are you basing it on?

 

[rcx194]

Hi Buggar,

You asked once before in another thread of mine ツ
I am using a lot of parts such as wheel spindle and driveshafts from a Mercedes commercial vehicle because they are low cost and strong. The uprights are going to be simple custom ones similar to these: