Page,
It would be such a wonderful world if all it was, was a simple thermal expansion thing. The problems that arise in a blade that has different structures in it (differential quench, such as a clay coated blade) comes from the different sizes of the cubic arrangement of the atoms. In the FCC form, things are nice and tight, taking up less volume. The BCC form happens when the carbon leisurely shifts into the center of the cube, but only slightly increases the size of the cubes. When the FCC changes into BCT, the whole structure rearranges into nested tetragons,not stacked blocks. This happens all at once (Near the speed of sound IIRC), and there is an expansion of the volume.
A momentary digression:
The way the structures are arranged is what makes the steel harder, tougher, etc.
The cubes with carbon in the faces move very easily, and thus Austenite is rubbery ( think of stacked blocks with the faces greased). The cubes where the carbon is in the center have a very even arrangement, and can be moved left and right-up and down with little effort, thus Pearlite is soft ( stacked blocks with smooth sanded faces).
The arrangement where the tetragons (cubes sitting on the points) form is really locked together strongly, and does not move easily (Think of stacked egg crates). It will shear before disrupting. Thus Martensite is very hard and breaks easily under stress.
OK, Back to the blade:
When an uncoated 1095 (or W1-W2) blade is quenched ,the time needed for the Austenite to start its journey toward Martensite is very short...less than a second.It has to drop from 1350F to below 900F in that step. Once below 900F, the blade is still rubbery, and can be straightened or bent easily.Then ,as it crosses the Ms ,around 400F it rapidly converts to extremely brittle Martensite. The good thing is that all the blade is undergoing this drastic conversion at the same time. But,give it any place for all that sudden stress to concentrate - A deep scratch, sharp angle, thin section,etc. - and the structures will shear, creating a crack. Some of these cracks are microscopic, and some sever the blade in half. Uneven changes of thickness can cause warpage, too.
Now, coat part of the blade with clay to slow down the thicker spine's cooling, and the blade will form both BCC and BCT structures from the FCC austenite. The first thing that happens is the edge falls to below 900F and that part stays as FCC for the time being. While this is happening the spine changes to BCC. This slightly expands the spine, pushing the tip down a little. When the FCC edge reaches about 400F, the structure instantly converts to BCT Martensite. This happens fast and violently, while increasing in volume, too. If there is the softer pearlite already formed in the spine, it moves under the pressure, and the tip curves up - creating the sori. If the spine is too hard/thick it will resist the pressure, and the edge may shear itself in several places to release some of the stress.
If any parameter is altered in this process, bad things can happen.
A crack in the clay can create a tiny hard spot upon quench (which will break under the pressure).
Uneven or improper blade geometry can cause warp/twist/cracks.
Too high or too low austenitizing temperature ,or too short a soak....
Improper quench.
Gremlins ( This need much further study, but is the most common uncontrollable parameter. Kevin has tried for years, but has never got a good micrograph of one. I believe they will etch only with a pyridine-butyric acid mix.That is why he has not pursued this farther.)
Kevin will clean this up as needed if any of the process is not explained quite right.The above is a simplified technical explanation.
Stacy
