More pics here, but the paper is behind a paywall (not free):
https://journals.aps.org/prapplied/a...lied.10.014009
An earlier paper was published in Proc Nat Acad Sci (free, as in speech and beer):
http://www.pnas.org/content/112/32/9828
From my limited understanding:
1. Gummy metals "bunch" (= large amplitude folding) at the cutter/material interface. This increases the cutting force needed to create the chip.
2. Coating the surface changes the fluid dynamics at the metal surface to induce segmentation of the "bunching" (way beyond my understanding). (Maybe the coating acts like a "chip break", or maybe it's a non-linear effect involving complex fluids and boundary layer effects that is beyond my comprehension.)
Where this might be useful immediately for us hobbyists is when tapping holes. After drilling, coat the holes with dykem/sharpie/glue stick before tapping. You could probably measure the torque needed with and without the coating using one of those "indicator"-style torque wrenches.
Looking back, I think I might have seen this long ago when machining OFHC copper. I could have sworn that the copper cut without a burr out at the first pass (after layout with Dykem) but then got worse after that first pass. But I just wanted to get the the part finished, so I did not with follow up additional tests. Also, I might be remembering it incorrectly -- that was long ago and far away.