Epoxy granite

I have a lot of natural granite scrap pieces and a machine can breaks the granite scrap to dust.
So what size of grit is good to make the granite ?
Also what epoxy must use to do epoxy granite mix ?

This sort of project takes a lot of research and it's unlikely somebody is going to pop up with the answer. The best place to start is the spreadsheet that some German fellow made that calculates ideal ratios of aggregate and epoxy depending on size and density. Do enough research and to ll run into it! (I don't have it right now so I can't)

Sent from my SM-G960U using Tapatalk

The basic idea is that you have a mix of sizes of aggregate, so that the medium-sized chunks fill the gaps between the larger ones, and smaller ones fill between the medium-sized ones, and all the epoxy has to do is stick them all together, without filling any large voids.

Yes, what awerby said. Except by that logic you could just use sand and no bigger pieces which would be easier to mix. But apparently it's much stronger if you use bigger pieces, with the limit being something like 1/3rd the width of the smallest cross-section of the casting you're making.
Epoxy with the lowest possible viscosity is desirable.

Sent from my SM-G960U using Tapatalk

There was a reference in a german language book on epoxy granite (the book is listed in the mega E/G thread), which stated that the ideal largest agregate size is 1/5 the smallest feature thickness.

Thomas Zietz was the guy who made the excel spread sheet (attached) to optimise the aggregate mix, for a closest approximation to a Fuller curve of maximum packing density.

Crushed granite is not ideal as it is irregular and sharp edged. The highest packing density occurs with rounded pebble shaped alluvial gravel. Ideally you should have as many different aggregate sizes as possible from your maximum size, down to the finest powder you can get. At least five different aggregate sizes. Sand is still rather coarse. A super fine powder, like quarz flour, makes a significant difference to the resulting Youngs modulus.

Ideally you are aiming for the lowest epoxy ratio possible (8% is a good target). Anymore than the minimum, reduces rigidity and and can cause separation/settling, with the associated pooling of epoxy on top. An 8% mix is a real b**** to work with. You will need a concrete vibrator, ideally using an inverter to ensure that you can tune the vibration to find the resonant frequencies of the mold.

Depending on what you are trying to use it for, you can deviate pretty far from those ideals, and still get pretty usable results. There are of examples of 15% epoxy and sand used for smaller machine bases working acceptably for some.
Mark