Use of epoxy for levelling

[Jonathan]

Per Jonathan's data, the suggestion must be that even with careful construction and preparation, the process retains a relatively high error rate, when compared to the required outcome. My addition is to suggest that the impact of the error rate, regardless of cause, could be halved by taking a two pour approach. I think that's relatively sound logic. But that not good enough.

At risk of repeating myself, the error I got with the epoxy was exceedingly small (according to my surface plate) - the large deviation you see at the end of the graph was caused by the bearing seal/lower surface rubbing on the aluminium epoxy boundary. That error went once the rubbing problem was alleviated. The other graphs with large errors were false readings caused by accidentally measuring roll not height. I'm not mentioning values here lest they taken out of context - see the original post for a proper explanation.

I heated the resin in a water bath before pouring. It clearly lowered the viscosity, but I can't comment on what effect this had overall as I have nothing to compare with.