10 replies to this topic

[snx843]

Hi.

Just wondering if any engineers can give me a few tips as to the importance or not and effects of rear wheel castor.

Is there any effect in handling from the angle of the suspension upright/kingpin on the rear? I know the wheel doesnt steer but there is scrub/toe angle and bump which might make it mean something?

I know all about front castor but I can't really find any info on the rear.

I am about to knockup a new suspension system for my REAR WHEEL drive car (this question is related to RWD only). This will comprise of dual wishbones, lower H and upper A. I am simply wondering what effect racking the wishbone pickups/upright could have.

Or do I simply put the upright and wishbone mounts flat the floor?

Thanks.

[Greg Locock]

Rear wheel castor gives you a camber change if the wheel steers. But the rear wheels hardly steer (1 degree would be a lot), so you need a lot of rear castor to get any effect. Modern tires give much less camber thrust than crossplies, so the effect is small anyway.

There is one important effect that I know of. If the suspension has zero longitudinal hub offset then the mech trail point is set by the castor. This will govern whether you get lateral compliance understeer or oversteer. On a multilink this is a compliance effect, as such not easy to calculate by hand. I'd rather have understeer, but again the effect is small.

Randomly picking two cars with some pretence at handling, one has 30 degrees of castor, the other has -30. So I suspect that it is not very important, or more accurately that other things get priority.

[cheapracer]

Agree with Greg, I would worry worry more about having a little trail to relieve deflection loads presuming you will use no compliance joints such as heim and ball joints.

Ifyou have compliance then as Greg suggests the thinking cap has to come on.

I think some mistake it as a stability item forgetting you already have the length of the wheelbase as your primary caster.

[RDV]

...as rear wheels (usualy) dont turn, it is a definition only with no effects, as GL pointed out above....in fact if you have two pickups on the top of the upright (for example) one would be the theoretical pivot point of upright, and one would be the theoretical track-rod...as it doesnt turn around pivot they would be interchangeable, giving the +30 or _30 example...depends which you chose.

[Greg Locock]

It's a compliant, rather than kinematic, effect. Vaguely technically it is related to the plan view centre of elastic reaction, or rather more sensibly, the zero steer axis.

[snx843]

Greg Locock, on Jan 4 2010, 00:15, said:

There is one important effect that I know of. If the suspension has zero longitudinal hub offset then the mech trail point is set by the castor. This will govern whether you get lateral compliance understeer or oversteer. On a multilink this is a compliance effect, as such not easy to calculate by hand. I'd rather have understeer, but again the effect is small.

Thanks for the replies, I understand most of the points you make bar this one, car to make the point in easier to digest terms. Thanks.

[cheapracer]

snx843, on Jan 15 2010, 07:44, said:

Thanks for the replies, I understand most of the points you make bar this one, car to make the point in easier to digest terms. Thanks.

Draw a line between the 2 ball joint holes (upper and lower) and see if the center of the wheel bearing is on that line or to either side of it, that is the hub offset.

If you put that offset forward of the ball joint line the side forces under cornering will make the wheel turn inwards (probably understeer) and vice versa (all other things being equal).

This can also be used as a tool to offset other deflections apparent in a suspension layouts - something that they pay Greg a bowl of rice (and a bucket of chips, Australia after all) to resolve.

[Greg Locock]

That's fine for a double wishbone, but in a multilink you have 3 or more ball joints, so you can't draw a line. The way we do it is by finding the longitudunal location at which a lateral force causes no steer.

that defines where the steering axis hits the ground. Then applya torque around the vertical axis, and measure the change in camber and toe. With a bit of maths that gives you the compliant castor angle and kingpin inclination.

So altogether that defines a virtual steer axis.


The same really applies to a double wishbone, since the inboard end compliances do have an effect, and of course if youn have a steering link then you really have a multilink anyway.

[cheapracer]

Give that man an extra bucket of chips and throw in a dim sim.

[mariner]

A question if I may - if you put a lot of anti squat into the rear suspension that may cause a variable castor effect with vertical movment. I think that would be true if the side view longtitudinal lines through the links converge ( or there is a single forward link like the old Jag layout).

If there is a convergence due to anti squat then I think the castor will change with wheel movement even if the geometry causes no steer at the nominal ride height.

Is that sort of variance with vertical wheel movement ever significant in altering lateral steer. If so I could see it ( maybe) causing "squirming" when cornering on very bumpy surfaces.

[Greg Locock]

Yes that is right. I doubt it is signficant. That is, there are far more powerful influences on rough roads.