[ksanalytical]

Hello,

I am outlining a production plan which calls for cutting single crystalline silicon wafer material on an IH benchtop mill. This stuff is extremely hard and must actually be ground with diamond tools like glass. My question is whether anyone has any experience with abnormal machine wear due to abrasive particles like this in the coolant. Depending upon the responses I get, I will probably build some type of barrier around the fixture and use a separate, closed loop coolant system. Or maybe I'm overreacting. The operations will create a fair amount of waste between slicing and pocketing. I fully expect this to adversely affect the coolant pump as well.

Ken

[ImanCarrot]

I've cut and polished Silicon before- grind it, dont cut it. Use a diamond impregnated wheel for parting off- run it fast with stacks of flood coolant. To surface it use either lose abrasive with water or lose diamond abraisive, again water based, not oil based cos it makes a hell of a mess.

Remember Monocrystalline Silicon will machine much better than polycrystalline- the crystal orientation is important.POlycrystalline is a nightmare.

To polish it use Cerox rather than Rouge- Rouge will stain your hands, your clothing your walls and everything (don't ask how I know this lol).

I'm an optical engineer so these tips are based on making Silicon lenses for applications in the Infrared.

I used to diamond turn the stuff using moncocrystalline diamond tools which cost about 400 quid each, but even they didn't last very long on the stuff- it's as hard as a whores heart!

[awerby]

As you correctly fear, those hard sharp particles will get in everywhere they can and cause havoc. There's inevitably some splashing that occurs, and the sliding parts of your mill, if they aren't covered somehow, will pick up that abrasive slurry and use it to commit suicide. I wouldn't worry too much about the pump, though - I've seen sump pumps last for years in a diamond tile-saw, and they're cheap enough to replace when they die. But do put a little care into protecting the vulnerable slides and screws of your mill, if it's expected to last for a long time doing this. The Goretite people might have some stock bellows that will work, otherwise you might talk to them about custom ones.

Andrew Werby
www.computersculpture.com

Hello,

I am outlining a production plan which calls for cutting single crystalline silicon wafer material on an IH benchtop mill. This stuff is extremely hard and must actually be ground with diamond tools like glass. My question is whether anyone has any experience with abnormal machine wear due to abrasive particles like this in the coolant. Depending upon the responses I get, I will probably build some type of barrier around the fixture and use a separate, closed loop coolant system. Or maybe I'm overreacting. The operations will create a fair amount of waste between slicing and pocketing. I fully expect this to adversely affect the coolant pump as well.

Ken

[ksanalytical]

Thanks for the advise and info guys... That's about what I expected. I'll actually be cutting very thin wafers (semi-conductor) with a thin diamond saw at high RPM. The fixture will sit on my 4th axis so I can rotate it 90 degrees easily. I'll probably build a shoe of some sort to direct coolant and minimize splashing as well as building a trough around the edge of the fixture with walls high enough to contain the coolant. My test cuts show that very little liquid is needed as long as it's constantly flowing.

Good notes on polishing. I won't need to do it, but if I can find a way to have the machine do it for me, I might add an extra pass to smooth the pockets out a bit just for looks.

Ken

[ImanCarrot]

It polishes to a mirror finish- use flat felt bobs, but don't let the heat get too high otherwise you'll get an "orange peel" like surface Cerox and water with a medium speed and light pressure- like pressing two fingers on it type of pressure.

If you need it to be perfecly flat use pitch polishers and hand lap it with Cerox and water- takes ages, but I could get them flatter than anything- one twentieth of a wavelength of light at 633 nanometers wavelength. Takes time though and a lot of skill hehe- hand polishing is more of an art than a science though