William Beaty wrote:
> ... I stacked up sixteen layers, squeezed it with two different > micrometers to exclude air, and repeatdly saw values of 6-7 microns per > layer. Also, the spool was stamped with "6" (but could have been "9.") > Also, Kapton is commonly sold in thicknesses of 25u, 12.5u, and 6u. So it > has to be 6u, right? [...] > > I sent some to Ludwik, and he obtained a thickness of 13u by accurate > milligram weighing. Hmmm. I quickly checked again and got 11u. DOH. > > So I tried again, stacking and measuring layers 2,4,8,16 thick. The > thicker stacks always gave 6-7 microns [...] What's going on? > > Now that I was guestioning my own measurements, it was clear that > excessive strain on the micrometer was reliably reducing my measurements > by a factor of 2. I could "feel" the jaws touch down on the stack at a > particular measurement. But if I squeezed the jaw tips a little, I always > obtained another number which was 1/2 the first one. I'd been ignoring > this initial "feel," and trying to squeeze with the same force in order to > "exclude air." Amazingly, this reduced the readings by exactly the needed > amount, even for a variety of trials with 4,8,16 -thick stacks. > > [ ... ] even managed to subvert my > cross-checks by compressing the differing stacks with differing force over > many trials. No, the force should have been the same in each trial. Think about what you've got. If each layer is behaving elastically, then the question you need to ask is, what force is needed to compress 1 layer to 1/2 its thickness? Call that force "F". And then you need to ask, if you stack N layers, what force will be needed to compress each layer in the stack ... and, consequently, the whole stack ... to 1/2 its thickness? Why, "F", of course! Just because a layer is in the middle of a stack doesn't change its elastic properties, and all layers in the stack are going to feel exactly the same stress. Just like a string, only in compression, not stretch. So, your "cross check" didn't cross check anything: If you squeezed to the same "finger feel" in each case, it wouldn't matter how many layers you stacked up, you'd always get the same off-by-2x answer. And if your two different mikes had the same ratio of turns/cm then you'd end up exerting the same force with each at a given "finger feel", and once again, there's no real cross check there.
