In house research: flex donut coupling improvements

[Quart Meg Miles]

I reckon the reinforced Guido couplings are the way to go.

You have a link to these?

[Chancer]

A French site but will give you the manufacturers OE part numbers for a further search.

https://www.piecesauto.fr/febi-bilstein-1868069.html

When I said reinforced, I meant they all have a fibre reinforcement moulded into them like a cam belt, that is until they too become collectors car pi?ces and are remade in far flung places by people sat on the ground clamping glued up assemblies with their feet!

[Quart Meg Miles]

A French site but will give you the manufacturers OE part numbers for a further search.

https://www.piecesauto.fr/febi-bilstein-1868069.html

When I said reinforced, I meant they all have a fibre reinforcement moulded into them like a cam belt, that is until they too become collectors car pi?ces and are remade in far flung places by people sat on the ground clamping glued up assemblies with their feet!

I think we've been here before, these are prop-shaft isolators which won't stand the angles required by the Elan family.

[trw99]

I also hate to point - because I have been guilty of this - but we are getting thread drifted a bit here.

Given David's efforts on our behalves in this particular area, may I politely suggest we keep this thread as close to its origins as we can?

Tim

[StressCraxx]

A French site but will give you the manufacturers OE part numbers for a further search.

https://www.piecesauto.fr/febi-bilstein-1868069.html

When I said reinforced, I meant they all have a fibre reinforcement moulded into them like a cam belt, that is until they too become collectors car pi?ces and are remade in far flung places by people sat on the ground clamping glued up assemblies with their feet!

I think we've been here before, these are prop-shaft isolators which won't stand the angles required by the Elan family.

Meg, you are absolutely correct about the angles required by the Elan family. The guibo couplings are not designed for much more than 5 degrees of angular misalignment. The original specs for the Metalastic couplings were also about 3 to 5 degrees. Chunky ignored the spec and used them anyway.

http://www.missionsupplyonline.com/pdf/rotoflex.pdf

I have been racing my Crossle 25F in vintage events. (Lobro CV joints-thank goodness)

We share our run group with Formula B cars, mostly Brabham BT 18, 21, 25s. Great fun when they come howling up behind you in Turn 1. They are making about 200HP with Lotus Twincams. Most of them run at least one metalastic or guibo coupling per half shaft. The other side is a U-joint if a metalastic coupling is not used.

Most of the pro prep shops have been migrating to the guibo couplings because in some instances the metalastic copies have failed with catastrophic results. The pro prep shops also permanently install the centering devices in the stub axles and half shafts. The joints themselves don't fail first. The bolts fail first because the bolted joint tension does not sufficiently exceed the cyclic loads upon the joint. Once a bolt fails, all hell breaks loose.

The metalastic joints used to have a solid steel insert moulded in for attachment. The bolts must be tensioned against the insert well above the cyclic loads. For years the copies have used a pair of folded sheet metal leaves to attach the shafts together. The sheet metal leaves will collapse from the fastener torque needed to keep the bolted joint under proper tension beyond the cyclic loads produced in service. Once the tension on the bolt is near or below the cyclic loads, fatigue failure begins. Loss of fastener tension from the leaflet collapse is the root cause of coupling failure.

The guibo couplings use a solid steel insert. The bolts can be tensioned properly to exceed the expected cyclic loads.

If I didn't already have CV joint axles, I would have converted to the guibo couplings.

[trw99]

David (lotustastic), I was wondering if you have been successful in carrying out any further research on your polyurethane doughnuts? Given other discussions on here, it would be interesting to know.

Tim

[rcfurse]

Having picked up this thread very recently I would offer the following observation. I have a "Guido" type joint in the driveline of one of my other cars. Actually it is between the prop shaft and the diff.
It handles 450BHP. It is VERY stiff. I don't think it is ever going to allow enough wheel travel.

[74Twincam]

I've been watching the donut/CV debate for a bit, trying to make my own decision, and this thread warrants a question, on topic of the original post.

I am a mechanical engineer, and have spent a deal of time with plastic molds, overmolding (or insert molding- combination of plastic bonding to metal in the mold). Polyurethanes come in a variety of properties and offer some very interesting promise for this application, it seems like a worthy pursuit.

Has there been any bench testing performed yet? I'd be more likely to be an early adopter if there was a decent amount of cyclic testing. Cool idea-

[trw99]

Bill, all.

A quick round up of what action David has so far reported, to save trawling through the whole thread:

13 Apr 2014: completed a prototype donut in tough polyurethane. Has improved reinforcement and superior material properties (improved elongation, peel strength, weathering, etc). The material is very resilient, tenacious and tough which should withstand constant flexing and stresses. The design addresses common failure points and should greatly reduce the disadvantage of rubber degradation by atmosphere, chemical attack, temp cycling, etc. The donuts have bonded bolt bosses and reinforcements and will be circular.
Will be using both a car and a test jig, though a testing jig until my Elan gets on the road for further road testing. Making up second prototype in the next week as I want to improve the bonding strength between the metal and polymer.
6 May 2014: made the second prototype with a tougher PU rubber grade and much improved bonding to the metal interleaves.
22 Jan 2015: completed another prototype using a slightly different material. The bonding to metal bosses, interleaves remains excellent. Construct a basic rig for testing full droop angles. The new donut flexes at full droop deflection without issue, needs more flexibility. If the next prototype displays the same bonding strength. I will have to commission some additional bosses/metal bits to be made so several donut sets can be made.
Performing some UV/weathering testing on materials.
One of the improvements is piercing the metal and allowing the "rubber" to flow through coupled with a better bonding agent.
28 Mar 2015: I expect to receive my first samples of improved metal bosses/plates in from the fabricator this week.
17 Nov 2015: A few more couplings have been made along with a second mould. Had some metal parts fabricated for feasibility, they look good, but cost a good deal.
11 Mar 2017: obtained some better machine shop quotes from suppliers for fabrication of plates. Did some manual cycling by flexing the axle incorporating my donuts and they do show somewhat stiffer properties as compared to OEM in terms of force to deflect given same amount of deflection.

Tim