How to measure drilling thrust force (feed force) 

Thank you for the feedback, have a great day! :)

I hope "fun to watch" is a good thing :)

Very good point! All deflections were tilting back; there was no measurable lateral deflection. 0.5" above the spindle nose .001", 4" above the spindle nose (middle of the head's face plate) .005", and 19" above the spindle nose (at the back and top of the column) .018" (or about 0.05 degrees.) 1 thou sounds good, but the head is low when measuring against the scale. Put a long drill bit in, and column-head-tool total may be worse than the .018". Yours is a great suggestion and deserves more investigation.

@@kentvandervelden If? I understand the engineering calculations for beam deflection. Doubling the height for the head position on the column and using the exact same load would result in the total deflection numbers to be squared with all else being equal. I've read it takes approximately 150 ft. lbs of down force to drill a 1/2" hole in mild steel with a brand new correctly sharpened drill. So adding a standard length drill chuck plus a 1/2" dia. jobbers length drill and the column deflection numbers are going to quickly skyrocket. Your test numbers are more than worth while checks, and I'm sure you understand there still only static numbers. Larger hole drilling is also going to exert high twisting forces on the head / column. "For every action there's an equal and opposite reaction". I have the same head / column bolted to the rear of my manual lathe as a light weight secondary mill so your numbers were pretty interesting. I've long thought that epoxy bonding & bolting a heavy wall steel I beam to the rear of the column with a heavy welded bottom plate on the I beam with that bolted down to a rigid bench top would do a massive amount to stiffen these little mills. Adding epoxy / granite inside the column and machine base would get you about the most the machine is ever going to be capable of. As they are it mostly doesn't matter how much money you throw at adding ball screws and nuts. The machine is incapable of utilizing that higher level of accuracy without doing a lot more to it. Machine mass, rigidity and vibration dampening is EVERYTHING with machine tools.

That's awesome, thank you! The beam should be fairly level when the scale is compressed, but the beam bending in the middle shouldn't be a problem. Ideally, the scale would be checked with a few weights too.

Hi David! I'm curious about the practicalities of building an instrumenting dynamometer. They are very expensive even used on eBay. The accuracy does not need to be high, relative measurements in multiple axes (XYZC) would be awesome. Maybe something we brainstorm about in April?

@@kentvandervelden sure! I picked up a few strain gauges for a project a year ago and I was surprised at how cheap they are. There is an inexpensive and easy to use single chip amp/ADC which is easy to find. Much easier than college when the load cells in the labs were cumbersome, expensive things.

Could you write code to generate g-code? If you need to respond to inputs quickly, could you modify LinuxCNC or Marlin? If something in the middle of timing requirements is ok, LinuxCNC can be controlled from Python. The inputs can be read and outputs controlled and g-code sent. Hope this helps. Sounds like you have an interesting project.

Thank you Andy, yes, in part, please see the response to The Turning Point's similar suggestion. Deflection while milling has more axes involved, but the forces should less too, but on finishing cuts, I must take things easy. If you have more suggestions, would love to hear them

Btw, Grizzly 8x30, is this Grizzly's smaller knee mill, the G0731? That mill looks awesome. It's a shame it's not living up to it's appearance.

Kent VanderVelden I have the Grizzly G0731, I really really should have just got a Bridgeport or BP clone. Pushing up on the spindle just be hand deflects it a few thousands.