Zlinster wrote:Thanks for that. It's never imagine that rotoflexes might be so difficult. Any idea why moderns aren't as good as back in the day?
I wrote this post in another thread about a year ago, but still holds today.
"Sorry for the long and overly detailed response, but I felt it necessary.
The Metalastik couplings available now are not of the same quality as the original ones from Dunlop. The donuts were designed for industrial cardan shafts in the 1950's or earlier.
The original couplings used a one piece solid steel insert molded with the donut to hold each bolt. The bolt was inserted into the drive flange, into the coupling using hardened steel washers. The assembly was torqued to spec to properly tension the bolt so it would survive the loads in shear and tension. As long as the bolts remained properly tensioned, the couplings worked well. I used to check/torque mine at every oil change. I also marked the nuts with a dab of paint to visually indicate if the bolts started to loosen. Once a bolt loosened, the cyclic loads caused cyclic fatigue and bolt failure. If the bolt failed, the coupling would often fail with it unless it was caught in time.
Unfortunately for all of us, the Dunlop/Metalistik accountants thought they could build a better (read: Cheaper!) coupling. They saw the expense of the bespoke steel bolt inserts as a cost savings/profit opportunity and the original inserts were substituted with some folded mild steel pieces that were spot welded together and formed a square hole for the bolt in the coupling. The new leaflets would barely hold the tension created by the bolted joint. If the bolt did not have enough tension, the joint would start to work back and forth and the bolt would fail from cyclic fatigue. If the bolt was torqued a little too much, the steel inserts would collapse immediately and the bolts would fail from cyclic fatigue. Then the coupling would fail.
It's very difficult to get a torque wrench into the area to perfectly torque the bolts under an Elan, especially at the diff. The last time I replaced my donuts, I torqued one of the bolts too tight and immediately felt the insert collapse. I removed that new coupling and threw it away, ordered another and installed it.
The point of all the above? The donuts are not as originally engineered and supplied. The design has been compromised. The angle at full droop on an Elan is well beyond the 5 degrees specified in the cut sheet from Dunlop. So, no matter if one uses donuts or CV axles or Hooke's joints, one has to decide what compromises one is willing to make for the sake of originality, reliability and safety.
https://www.robush.com/wp-content/uploa ... oflex-.pdfLastly, I race Formula Ford. The pre 1973 cars all used donuts or donuts and u-joints. They are raced that way today. I was witness to a donut failure at Sonoma three years ago. The flailing shaft tore the left rear corner of the car off and the resulting spin put the car into the tire wall at 100mph. A beautiful Titan Mk 6 was written off."