Torsion Beam Rear Suspension Questions.

[XR250]

My new Mazda3 has one. I know the advantages:
-Cheaper to make
-Better Packaging
-Less moving parts = cheaper in the long run

My question is does this setup help during braking to resist dive (as it is essentially a trailing arm suspension)? As the braking force is below the pivot, it will tend to make the rear end squat under braking - true?

Also, when hitting a raised bump, the horizontal component of that force is also below the pivot which would tend to move the wheel over the bump easier than a standard strut setup?

Honestly, Mazda did a nice job with it. The car corners super flat and it not harsh when one wheel hits a bump. The only time I notice its deficiencies are in bumpy corners where it can get a bit out of shape.

Thanks

 

[BrianPetersen]

Yes, trailing-arm-based suspensions (including twist-beam-axles) have strong rear anti-lift during braking. If the geometry is right, there should be little change in rear ride height when braking. The anti-lift from braking balances the unweighting of the rear.

Fun fact, if you stop on a steep uphill and apply the parking brake, and let the car rest against it (so that the rear wheels are stopping the car from moving backwards), that'll lift up the rear pretty significantly.

I'm pretty sure 99% of people won't notice the ride and handling difference between a decent twist-beam and a full multi-link design, which is a lot more expensive ... and they will notice the difference in interior space, if the manufacturer chose to take advantage of how compact a twist-beam is. That's probably why Mazda went from multilink to twist-beam axle with the latest 3.

My current daily driver (Chevy Bolt) uses a twist-beam axle, so did its predecessor, and vehicles I've owned in the past have had multi-link IRS, or MacPherson IRS, or coil-spring axle, or leaf-spring axle. The calibration matters more than the general description. The trucks and vans ride harsh because the spring and damping and bushing rates are off the scale due to the load-carrying ratings.

 

[XR250]

"Fun fact, if you stop on a steep uphill and apply the parking brake, and let the car rest against it (so that the rear wheels are stopping the car from moving backwards), that'll lift up the rear pretty significantly"

Thanks.
Yup, have noticed that when I back up and forget to take off the parking brake, the back jacks up.

" The calibration matters more than the general description. The trucks and vans ride harsh because the spring and damping and bushing rates are off the scale due to the load-carrying ratings."

Makes sense.

 

[GregLocock]

Their main failing is compliance oversteer, which can be moderated by paying a great deal of attention to the design of the main bushing, or using a panhard rod. They also tend to have too little longitudinal compliance leading to impact harshness. However it is obvious these things can be developed out, given the popularity of the design.

Cheers

Greg Locock


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

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

 

[gruntguru]

Greg is the compliance oversteer caused by steer compliance, camber compliance or both?

je suis charlie

 

[GregLocock]

The dominant effect is steer compliance, simply because the lateral force is behind the bushing to the body. Seems to me for an EV the ideal solution would be to put the bushings behind the wheel, giving compliance understeer, some protection for the battery, and more room for the battery. It also gives wheel recession in jounce which tends to be better for impact harshness. I don't remember camber compliance as being much of an issue, not that the solutions would be palatable.

Cheers

Greg Locock


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

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

 

[BrianPetersen]

Lateral force at the rear wheels turns into a torque that twists the entire axle assembly against its bushings. Draw it in top view and draw a free-body diagram of the twist-beam assembly. It's compressing the two bushings in such a way that the deflection points the entire axle towards oversteer. Also, the bending forces applied to both the trailing arm and the entire axle, are in the direction of toe-out (i.e. oversteer) of the outside wheel.

You will see that most newer implementations have the bushing axes in top view at a very interesting angle. This is so that the lateral force applied to the bushings, and the resulting deflection sideways, "steers" the axle towards understeer, countering the above tendency. Those bushings also usually have voids or cutouts in them to carefully tailor how they deflect.

VW Golf Mk1 made no attempt to do anything about this:

https://www.heritagepartscentre.com...it=bounds&height=750&width=600&canvas=600:750

Neither did the Mk2:

https://www.heritagepartscentre.com...jpg?optimize=medium&fit=bounds&height=&width=

Mk3 the bushings started looking funny (ebay link allows you to zoom in):

https://www.ebay.ca/itm/285547483378?hash=item427bf4d0f2:g:PQkAAOSwgiZlSTov

Mk4 got the weird-looking angled bushings:

https://www.idparts.com/images/6340_a4rearaxle.jpg

And that's how most everyone else now does it as well.

 

[GregLocock]

Vehicle dynamics engineers always hope that bushes will move the way they want them to, so angling the bush should encourage the bush to sort of slide axially rather than radially, since the axial compliance is greater. If this works then the centre of rotation in the last photo would be behind the axle, which would be ideal.

However rubber is not compliant with vehicle dynamics engineer's wishes and this effect is quite common fig 15 in

https://www.scitepress.org/Papers/2019/88689/pdf/index.html

that is, yes you change the magnitude of the compliance steer but not the sign.

If you cut a successful (ie understeering) bush apart you will probably find some interesting design features to encourage the deflection to behave the way we want it to. I doubt most ADAMS models have bush models of sufficient complexity to replicate this effect (you'd need to test the bush in a multi axial rig or non linear FEA). Hmm, now I'm puzzled, I got good correlation with a test vehicle despite only having uniaxial bush splines. Perhaps it is easier than I think.



Cheers

Greg Locock


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

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

 

[XR250]

My Mazda2 had a torsion beam and it was def. tail happy. The 3 seems more neutral. This is all very interesting. Thanks.