Phil;

I don't believe there is any materials compatibility issue either, but the fact remains that silicon fluid in these systems extend the service life of the components tremendously, so the fluid is clearly a factor. I believe part of the explanation is that silicon fluid has better lubricating properites.

As far as your suggestion about corrosion susceptibility of the aluminum housing...I don't know if I totally agree...I have always seen longitudinal scratching in the bores of fail slaves which used DOT3 or 4 fluids (to go along with some pitting... but not really enough to allow seal bypassing)...so I asked myself: What can cause this?...and since I NEVER dumped any grit into the system (and never even found any which might be responsible), I had to figure out the scoring mechanism...I finally decided and am convinced that it had to be the (hard) piston cocking in the bore, making contact with the wall, and being dragged along it during movement (and this would be consistant with the lubricating properites of the DOT3 and 4 fluids not being as good as those of silicon, and allowing this to occur).

To keep the piston from cocking in the bore, I now O'ring the "dry end" of the pistons...I have yet had a clutch slave fail after making this mod (along with going to silicon fluid)...and it must be going on ten years...without repeating the DOT3/4 vs. DOT5 discussions, I would bet that an o'ringed piston, in a DOT3 or 4 filed system, which had yearly fluid changes, would benefit also.

Cheers