Blokeinamoke,,, you`d be pretty right if the wheels were turned left or right so the diff would come into play,,, or if one wheel was weighted more than the other,,, & if he actually had a plate type or adjustable type diff,,,
but with both wheels evenly on the ground, both wheels pointing straight ahead etc,,, it`s shear sudden impact value thats causing the distruction, the diff may as well not even be there as it`s just the same as having a mini spool or solid straight through axle... the diff will only come into play if one wheel starts to move faster than the other.
I was going to suggest bigger axles & bigger CVs like the morris 1500 ones, we did this mod to a speedway car years ago & i am doing it again now with my Turbo V8 mini, i believe that Peter Thompson from Towoomba also did this conversion on his 1510cc twin cam injected KAD headed sports sedan back in the day.
But still,,, the theory of a thinner axle allowing it to twist is a good one & is quite popular among many engineers & certaily works in most situations so i won`t argue with that at all, it`s taking the load away (or reducing the impact from) from the thin splined section & obviously the cirlicp groove is where they commonly break,,,
I just prefer to beef up the areas where they break & then keep chasing the breakages as they transfer elsewhere until you`ve nailed all the weak points
I know it adds a little more weight here & there but also allows for even more grunt to be used at a later date... lets say (for example) If you used Austin 1800 CVs with the very much bigger splinned/circlip groove section, then you can use the morris 1500 axles & even thin them down to produce some "give/twist".... i`m not saying it`s an easy job tho
Quite often it`s best to have some clutch slip & rubber uni joints
(kidding)
), but Morris 1500 CVs would make sure they last. I should also say that both my driveshafts will be equal length and as such, very short. 286mm end to end.
