Big Epoxy Granite machine, Which frame is better ?

[m_c]

I can't help feeling that the thermal expansion of epoxy concrete, typically twice as much as steel, is going to tie that machine in knots. I'm happy to be educated otherwise.

This was something I failed to find any kind of definitive answer to.
The conclusion I drew, was it can't be that big a concern, given there are some very big EG machines with linear rails mounted to them, with no apparent allowance for expansion. So if it's not off any concern over a couple metres, I doubt I'd worry about it on your typical home build.

As for John's build, what I took from his internal structure, was it was more about keeping things in place during the pour/cast, rather providing much in the way of structural support. Remember in that photo, the outer steel plates are the actual mold and designed to be removed after curing, so they can be reused.

[Zorbit]

Removed.

[JAZZCNC]

I can't help feeling that the thermal expansion of epoxy concrete, typically twice as much as steel, is going to tie that machine in knots. I'm happy to be educated otherwise.

I don't think so.! Some of the best and most accurate lathes and Milling machine in the world use Epoxy granite because of it's thermal advantages and the fact they can build in extra mechanisms ie: running coolant thru the frame to increase this even more. Kern for instance and the Bigger DMG Mori's etc all use EG for there frames.

[pippin88]

Kern uses aluminium for the linear rail mounting I believe. They cool / thermally control all the mounting surfaces.

[Zorbit]

Removed.

[Voicecoil]

Being intrigued by this I went digging and found the thermal expansion coefficient. It will vary with the mix (more epoxy = more expansion it seems), but typical figures are 12...16 x 10-6/deg C. Tool steel came out as 11 x 10-6, so they're pretty close.

[Zorbit]

Removed.

[Voicecoil]

I found figures ranging from 12.5 - 28 for epoxy concrete. I'm guessing that the lowest figure is for carefully selected and graded aggregate and the best performing resin, although I can't find out which resin that is.

So, careful mixing would get pretty close as you point out. The next question is what difference would it make ? For example, steel/epoxycrete with a difference of 5um per deg C per metre, with a 10 deg temp change and a 1 metre linear guide means a length difference of 50um, not a lot. I don't know what effect that would have, perhaps someone with some engineering expertise can answer that. In my head I can see the bi-metal strips from school science class, they bent dramatically but the materials were carefully selected to do that.

One thing has become clearer to me - careful mixing is vital - the coefficient of expansion of neat epoxy can be 10 times as high as epoxy concrete, the mix needs to be very even.

The papers I read seemed to say that the major deciding factor was the % of mineral filler, and then the particle size distribution - you seem to need a fair % of small stuff to keep the minimum distance of epoxy small and fill in around any larger lumps. Some people were going as high as 90% mineral, albeit with a material which was little more than sand, hence would pack together tightly - the low % of epoxy would eliminate areas of overly-expanding undiluted epoxy I guess. Removing air bubbles is another issue - good luck with your build!