Trailing Link Flexure

[bhart]

Does anybody know of current designs that utilize a/multiple flexure joints as a substitute for a spherical type joint?

I have been considering ways to accomplish this for the purposes of a racing suspension, where it is generally required that a trailing link joint be very stiff along the axis of the link (minimize compliance) but very flexible(or pivoting in the case of the spherical). For these reasons, it did not seem that a rubber-based joint was feasible, but I don't really know much about what is out there in the way of rubber-based joints.

Looking for resources and open to any suggestions.

 

[NormPeterson]

Is this intended for a vehicle that sees much, or any, street use?

Most dedicated race cars use all-metal spherical rod ends, which tend to be harsh and may not be street-legal where you live.

There's something called a "jonny joint" that I think is essentially a steel ball in nonmetallic race halves. Good axial location, without being as harsh as rod ends.

Alternatively, you can get creative with existing polyurethane bushings and mill/drill/sand away some of the "bad" stiffnesses. There's likely some long-term loss in durability associated with mods of this nature, but you should be able to get at least a couple of years of mixed normal driving and competitive use out of a set.

Norm

 

[bhart]

Norm,

No street use, strictly a race vehicle. Also, not looking for something that is "less harsh" or anything like that.

I just mainly want to eliminate friction surfaces from the suspension system by eliminating sliding pivots. A similar idea to what open wheel cars (F1, Indy, etc.) often do w/ their inboard wishbone pivots, but in the trailing link case, the link actually needs 2DOF @ the joint, whereas the open wheel arrangement replaces two 2DOF joints (two sphericals which actually create a 1DOF pivot for the wishbone) into a 1DOF pivot axis (flexures) & that works nicely b/c they always know the axis that they would like to pivot on. In order to replace the spherical in the trailing link application, however, you still need 2DOF --- so a flexible mount analogy would be to mount the trailing link to a torsion bar & build the link attachment so that it is not very stiff in one direction @ the attachment point, thereby getting you another DOF through the flexing of the structure, but in a controlled sense.

The problem w/ this proposal is that you would ideally like the torsion bar to be very soft in torsion, but stiff in bending so that you can mount it transverse to the trailing link & it can support the trailing link's axial loads, but in order to make a torsion bar softer torsionally, you are probably going to reduce the bending stiffness as well which is why I found it appropriate to throw this out on Eng-Tips to try and gather some ideas & examples on similar applications.

 

[NormPeterson]

I think you need all three rotational DOF's free so that suspension movement won't be loading the link in anything but tension (and compression if this is for the drive axle). X-axis is torsion, Y-axis is covered by the pin axis, and there's also a small Z-axis bending effect due to roll (and bump, if this is an independent arrangement). You can minimize the X and Z effects with a tall thin blade-style of link if this is not the drive axle (Greg Locock's Focus) or if it's the drive axle on a low power class (you've essentially described the early VW beetle trailing link design). Regardless how stiff or soft the torsion bar is, at least one support on each end will still have to be either sliding or compliant.

I don't yet see any requirement that the torsion bar be of round cross-section, so maybe its shape can be tuned to the needs. IIRC, the early VW's used a collection of flats at one end or the other (I think early dune buggies removed one or two to better suit the much lower weight). Perhaps your T-bar could be round where it has to pass through each support tapering vertically and longitudinally into a wide, thin flat over most of its length (maintaining some minimum I without much R, hence not much torsion or Z shear).

On the other hand, voided bushings have been used by the OE mfrs to offer relationships between the various stiffnesses different from what homogeneous pieces provide. They've also used bushings that are oval in cross-section. And one aftermarket supplier that I know of offers poly bushings of a 3-piece design (a relatively rigid central slice with softer outer pieces) that's of generally similar intent.

FWIW, lubed poly bushings don't offer all that much resistance to motion, nor would I expect the JJ's to do so either. Even twenty ft-lbs of torque over a link as short as 10" corresponds to only two pounds of vertical force at the ends. Bonded rubber bushings actually have a spring rate of their own (plus a bit of hystersis damping) that has a greater effect.

Norm

 

[bhart]

Norm,

I see what you are saying about the 3DOF requirement, but the way that I have analyzed a trailing link in the past is to say... OK we have two points with a fixed distance in between them, one is grounded (chassis pivot) so what do we have to specify to determine where the other point (axle pivot), and in my experience you only need two pieces of information to locate the other point, therefore it only has 2DOF. A U-joint (2DOF) is essentially a real life spherical coordinate system, and it can always point where you tell it to, right? Couldn't you actually use a U-joint as a trailing link pivot w/o introducing alternate stress modes to the link? If yes, then I will stand by the 2DOF thing, but I can definitely envision in my head the situation that you are describing that appears to require 3DOF... to me, the 3rd degree of freedom has just meant that you could walk up to the link at anytime and twist it freely on its axis, but the suspension doesn't really care if its trailing links are twisting on their axes.

That is an excellent idea about the variable section T-bar, I will have to run some calcs to see what potential there is... so far, it's more clever than anything I've looked at up to this point.

Thanks again.

 

[GregLocock]

We use rubber bushes at the lower end of our front shock absorbers. They are functionally equivalent to a spherical joint, but made of rubber.

The stiffness in the radial direction is of the order of 20000 N/mm, which is at the lower limit of what you'll see from a rod-end.

They can cope with 20 degrees of articulation in torsion, but probably only 5 degrees or so in coning.

I don't know what the axial stiffness is, probably at least 5000 N/mm

I'll grab a part number on Monday, you'll be able to get them via P&A.


Cheers

Greg Locock

 

[bhart]

Greg,

The part # would be greatly appreciated.

 

[modesign]

The latest Racecar Engineering magazine details the Vauxhall (The Seat is very nearly the same) BTCC race car rear suspensions. These use a subframe with multiple flexure joints. The article has an excellent drawing of the layout. This is done because of the regulations on modifications and is said to provide almost the efect of a true multi-link rear suspension The car won the championship so it works!

Sandy Cormack

 

[GregLocock]

bhart - P&A part number is AU 18174 B, for Ford of Australia Falcon AU or BA model, ie since about 2000

It'll take a couple of tons at least. It's a rubber bush that will need to run in a smooth steel housing, dia 32.0, length 47.3, with a 45 degree chamfer outwards from that to press it in.



Cheers

Greg Locock

 

[Tmoose]

"It's a rubber bush that will need to run in a smooth steel housing........"

The bushing is intended to slide or spin in service ?

 

[GregLocock]

It is a stick slip bush, in concept. I don't think it moves much in practice, but you need a smooth finish just to avoid damaging the rubber as you press it in.



Cheers

Greg Locock