Do we know what the OE ring material is? Of the failed OE rings I've removed they were a mixed bag of different discolorations and deformations so I'm not confident that there's a reliable supply made from the same material. As I've experienced with waterpumps recently, OE doesn't necessarily mean quality unfortunately, and there is more than one source.
This is the reason I looked to O-rings suppliers and I did quite a bit of reading, and there's some conflicting information out there. Thinking back, I did think I was taking a punt close to the temperature limit with the Viton, but
Simply Bearings actually list Viton with a similar temperature tolerance to EPDM and having read about Simons results with EPDM I believed it should be good enough. I think I'd read that Viton was ok with water up to 120deg, so with coolant (not water) at 90deg it should have been ok, at least better than the suspect 'rubber' OE parts I've used previously.
If I can find a reliable source of OE spec O-rings I'll happily try this, and hopefully with the supposedly more rigid pipe work it'll give good long term results.
I have made enquiries about Aflas/FEPM O-rings, which should be even more tolerant of high temp OAT than EPDM or Viton, so I'll let you know how that goes.
With respect to oversize O-rings, what do you think the potential of a 1-1/2" x 1/8" (31.34mm x 3.53mm) would be? It's a little thick and a bit narrow on the ID, but with a little stretch to the nominal 32.5mm would thin down a little closer to the nominal 3mm. I know this introduces another stress into the equation, but could it work?
Simon, what spec are you currently using, and what do you think caused the failure on your epdm rings?