Hold-Down and Cooling Problems Cutting Metal

Cooling solutions tend to make the table slippery and interfere with vacuum holding ability. Here are some tips for working with metal on the CNC. April 16, 2009

Question
Most of our machining is wood and plastic related, but lately we have been trying some non-ferrous metals. The double sided masking tape we use in addition to the vacuum holds parts nicely, but when we are working with metals we have to use coolant. We are only using soapy water, and working on a sheet of plastic to keep our MDF spoilboard dry, but as soon as the water gets to the tape, the parts move. We have tried spray-glue and contact cement, with very poor results. Any ideas on holding down small parts cut out of sheets of aluminum or copper using water coolant?

Forum Responses
(CNC Forum)
From contributor B:
Seems to me your best (and perhaps only) bet would be to onion skin the parts.



From contributor R:
Depending on the software you are programming with, you may be able to leave tabs to hold the parts to the web. I know that Alphacam and Microvellum can do this; I'm sure others can as well. If your software can't do it within the program, it should not be too hard to add then in a thin sheet application.


From contributor M:
Speeds and feeds, cutter geometry, cutter diameter, machining strategies and cutter fluid application become considerably more critical when cutting aluminum and copper. There are many reasons, but essentially the "bandwidth" of cutting parameters (constant surface footage, material removal rates, chip loads, tool edge life) is much smaller and finite for metals than for wood and plastic. It becomes much more sensitive to the conditions of the particular setup and application.

You mentioned a soapy water solution. I am unsure of the method of delivery, but I would recommend a soluble oil solution (1 part oil - 20 parts water) delivered via pneumatic mister.

Also, typically you should be running in an rpm band of less than half that of normal wood and plastics machining. Chip load would be cutter specific. Tabbing would also be recommended on smaller parts.



From contributor A:
It may not be the answer you are looking for, but one more option would be to subcontract the job to a laser or plasma shop. I have seen thousands of very tiny parts nested on a sheet, like special washers smaller than a dime with intricate detail cut on a laser. No side force, no movement.


From contributor O:
Have you tried air for cooling? A nozzle directed at the point of cutting with one of those articulated stiff air lines.


From contributor A:
Here's an idea - may be practical or not depending on how small is small for your parts. Perhaps you could use a plastic or metal tray - could be the size of your spoilboard, half that size, or just slightly larger than your parts to contain the water for coolant. Then use a double sided pod or a custom shop made pod stuck to the bottom of the tray or to a half inch block glued to the bottom of the tray. The gaskets may provide a good water seal and keep the parts stuck. You would only need to make the tray once and it would keep the water contained for you as well.


From contributor K:
Rather than using any liquid, I would purchase a cold air gun. The brand I use is Vortec. It uses vortex tube technology to lower the compressed air temp around 75 deg lower than ambient.


From the original questioner:
Thanks for all of the ideas. I have kind of hybridized my approach, using a combination of several methods. First I had a new aperture machined so I'm only putting out mist. Second, I have switched to an acrylic film tape that is double sided. Third I have built a wet-dry shop vac attachment to suck up any extra liquid, as the pooling was compromising the tape more than anything. This last suggestion from contributor K, however, is really interesting. I checked out the vortex site and sub-zero spot cooling looks like about the coolest thing I've seen since we started running the CNC. Thanks for the help. I'm running a few more tests tomorrow!


From contributor S:
Try tabbing the parts and using a downcut bit.