Semi-trailing arm mods - roll center moving downwards

[DB Boatman]

I have some question about a BMW 2002 that is likely to get a 2-2.5 fold increase in hp. It's a pure toy car with occasional trips through the mountains and road course/track days. The issue is a question about roll center height from fitting E30 rear control arms to a BMW 2002. Fabrication is required as the 2002 arms have a 20° sweep angle c/w the E30 is 15° and the E30 has a wider rear track.

I was wondering about the impact on the RC from moving the E30 arm pivot mounts in-board to retain the same wheel track as the 2002. I did the not-to-scale drawing to try to understand what may happen. It seems the instant center will move outwards as a joint result of the sweep angle change and moving pivots points towards the centre of the car. Coupled with IC moving and narrower track, the RC will move downwards.

Doing some reading and to be honest I'm baffled. I have read Mitchell's Roll Center myths and reality and understand very little. I noticed that Brian Petersen said in another thread "Roll center too high at nominal ride height or in bump = BAD", which I get.

What's the impact of moving a roll center down on the rear of a semi-training arm suspension? Is this an issue that warrants any further thinking? Need I look further with actual measurements?

 

[BrianPetersen]

I betcha you won't notice the difference.

 

[gruntguru]

Agreed. It certainly won't be worse in a performance application. Lower RC, less toe change (in bump), less camber change.

je suis charlie

 

[BUGGAR]

Moving the "roll centre" down will slightly decrease the roll resistance of the chassis/suspension system, likely an undetectable amount as others have said.
Now for my rant: Forget "roll center" and think "suspension geometry induced roll stiffness". If the "roll centre" migrates, the chassis roll stiffness changes. The "roll centre" is merely a means to measure this stiffness, so just use roll stiffness directly.

 

[DB Boatman]

Thank you fellas

 

[sierra4000]

Keep rear RCH down on this BMW, this improve traction during corner exit, factory cars E30 had about 10° sweep angle.
Racing 2002 cars go extremely low rear ride height for RCH reasons.
Try find on Google.

Radek

 

[DB Boatman]

Thanks. The standard E30 arms are 15° and race cars had 10° to 12°. On the 2002 suspension, I'm mindful of the bum down nose up attitude of the 2002. Mine has that as it is as the ride height set up by a specialist race car suspension bloke. The main problem I have been getting occasional sudden oversteer. I think it's binding of the flimsy 2002 rear STAs with Kmac toe/camber adjusters which have a large metal eccentric bush with little compliance. E30 rear suspension offers stronger diff and arms, disc and a stronger diff.

 

[BrianPetersen]

Lateral cornering loads cause toe-out compliance-steer of all trailing-arm-type suspensions - whether it's bending of the arm itself, or compliance of the bushings, or bending of the subframe that the arms are attached to.

 

[sierra4000]

I recently performed several measurements on the E30, most deflection is in 1)subframe 2)wheel bearing 3)arm

https://www.youtube.com/watch?v=3LPVXFKCMkk

Some easy reinforcement had significant effect on toe and camber stiffness

 

[cibachrome]

A Track (Panhard) bar can be used to produce lateral force deflection understeer in a trailing arm suspensions. The link or Watts element(s) must be behind the axle's for application point and the arms or bushings must be compliant enough to allow movement/rotation.

So, who needs a gas tank or a spare tire ?

 

[sierra4000]

Cibachrome,
You have a real car example?
and what happen with RCH?

 

[BrianPetersen]

No example, but I know what he means. I don't have an easy way of drawing anything and posting it ...

Look at the top-view diagram in the first post. Imagine extending the trailing arm rearward (down on the drawing) past the wheel centerline and building in a place to connect a ball joint. That's one end of the "panhard rod". Now identify the pivot axis (in 3D space) of that trailing arm. The other end of the "panhard rod" attaches to a ball joint somewhere along that pivot axis towards the other side of the car. Essentially it preserves the semitrailing-arm geometry (the roll center is not affected) but it braces (triangulates) the trailing arm against bending. If the attachment point to the trailing arm is aft of the wheel centerline, bending loads on the trailing arm (and the associated bushing deflection loads!) from cornering actually go in the other direction. As a bonus, if the attachment point on the trailing arm is as low to the ground as possible, it relieves the responsibility of the trailing arm to take up the twisting loads imposed by cornering. You need one for each side!

Practical issues abound. If the vehicle in question is front drive, a popular spot to put the petrol tank is in between the trailing arms and aft of the twist-beam or subframe, thus slicing right through where you want to put these radius rods. On small cars (like mine!) it's also a popular spot to put the spare tire; the petrol tank on mine is in front of the twist-beam. If the vehicle in question is rear drive, the diff and halfshafts are in this area. There's a limit to how far across to the other side of the vehicle that you can put the fixed ends of these radius rods, because if it's on the other side of the centerline, the left and right have to cross each other somehow.

I've seen a picture of something like this on the rear (twist-beam) of an older VW Golf. Basically they ran those braces from the ends of the trailing arms to the middle of the twist-beam. It's still a twist-beam axle, but the ends of the trailing arms are now triangulated to the middle of the twist-beam.

 

[GregLocock]

I've never put a watts on an IRS. Obviously it would have to be centre to body rather than arms to body. The solution to compliance oversteer on an IRS is to use a sensible IRS layout. Reacting lateral loads by squirming around in mechanisms with bushes in coning is doomed to failure.

I have seen watts prototyped on a twist beam suspension, it was a bit nicer but not worth the dollars in production, and I'm sure the packaging guys would have thrown it out.




Cheers

Greg Locock


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

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

 

[sierra4000]

Greg,
Watt was used on Opel Astra (Chevrolet Cruze) twist beam.

 

[moon161]

 

[GregLocock]

Wow. OK I see how they can get the exhaust through, but they've got room for a 5 gallon tank in there! I guess they must have had a high floor.

Cheers

Greg Locock


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

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

 

[cibachrome]

GM's X cars had flexible blade controls arms and a Panhard bar behind the axle. Deflection understeer. Lots of games and gimmicks in NASCAR using a spongy truck arm transmission mount and a panhard bar. Same deal. AS I recall, there was a twist axle with a Watt's link in that same type of suspension. (Audi ?) Costly, but effectively eliminated the lateral force deflection oversteer that plagues these 1 piece stamped steel and welded suspensions. Of course a properly executed arm to arm anti-roll bar also helps a bunch with your semi trailing arm issues.

RCH is defined by the shear center location in an open section twist beam suspension. Not the section's centroid, but the shear center. So, the shape of the beam and it's orientation in a side view is all you need to know. The shear center property also defines a few other important suspension K&C properties.