I looked at the photo initially and saw the quench lines and the coloring and thought "Too hot,too thin, partial quench".
Sadly, all too often an improper quench is blamed on the oil, not the procedures used. Going to a slower oil will just make a lesser quality edge by allowing more pearlite to form. 1095 needs the speed.
I know the metallurgy thing is not your style ,Tai, but what happens in cases like this is a factor of large grain size, caused by too much heat, and the partial quench causing different structures to form along pathways that were started in the overheating.
A simplified time-line is approx. like this:
The blade edge is overheated, leading to large grain growth . This particularly affects the tip and edge. A particularly hot spot may rise into the bevel in a rounded hot spot. Along these grains, the alloy ingredients in the steel collect in ribbons, instead of evenly distributing themselves.
( In 1095 - Carbon, Manganese,Phosporus, Silicon,, sometimes Vanadium ......all assuming the steel is actually 1095????? )
The blade is partially quenched into parks #50 that is too cool for optimal quenching. The quench is already somewhat compromised by the partial submersion, and the cooler than optimal quenchant temp also retards even transformation. This results in cooler and hotter places on the blade.
The blade is pulled out of the quench after a few seconds. This is one of those things that lots of people do, that leads to cracks and breaks, post HT. The spine is pearlite bound if it misses the nose, around 1000F, so why subject the edge to the severe stress of a sudden uneven cooling and very irregularly applied auto-tempering by allowing the pearlitic spine to bleed heat into the not yet martensite edge? The edge is going to start to turn from very soft austenite into glass hard and brittle martensite in ten to thirty seconds......why not let it have the smoothest ride, by leaving the blade submerged in the quench oil for two or three minutes.
OK, so the blade was partial quenched, and then taken out of the oil and set down on the bench too cool. The pearlite along the spine was already formed,and although it is considered a soft structure, it is actually pretty hard. It can skate a file sometimes. The middle of the blade is where the blade is still a mix of pearlite and austenite. This will change into pearlite and martensite as the blade cools, with the heat from the spine leaching into the mix and causing a line or wavy pattern. The pearlite in this area will be a softer type than the pearlite in the spine. A little bit later the martensite along the edge forms suddenly as the blade crosses about 400F. This hard and brittle stuff can be literally pulled away from the pearlite along those giant boundaries formed where an area was overheated. Why?...because the grain boundaries block the growth of the pearlite "fingers" growing into the martensite edge and thus holding on to it. The edge pulls away or splits as a crack. Cracks perpendicular to the edge are usually caused by the stress of the entire edge giving way as the martensite expands.This is sort of like a board cracking when bent too much. This is the dreaded ping in yaki-ire. Cracks parallel to the edge ( or curved and separating cracks like this one) are caused by the hard martensite forming along a dedicated line, usually the product of grain growth and alloys accumulating along that boundary ( alloy banding). This gives a path for the stress to follow, and separated easily. The example would be a board with a knot or big grain swirl breaking at that spot when bent, You know exactly where it will break, and it usually separates in a clean line. The knot sometimes falls free of the board completely. The problem is worsened by the fact that the alloy ingredients are not distributed, so the steel has places with too much alloy, and places where it is starved of alloy....and thus the steel in those areas is just C and Fe, making it extra brittle.
The solutions are not to use a slower oil and get a lesser quality blade, but to use the Parks #50 at 120F and HT the blade at a lower temperature and with better temperature control. Once the partial quench has had about 5-7 seconds, completely submerge the blade and let it sit in the oil for two or three minutes. Once cooled to a temperature you can hold in your hand, wipe it off and immediately place in the tempering oven.
This won't guarantee that a blade will never crack, but will give you the best chances.
The second solution is to use a steel more matched to the type of quench done. 1084 may have be a better choice for such a quench. Canola would have worked for 1084....and your shop would smell like baking cookies !!!
I did not go into the HT before the quench, as that is another long subject . But, on a blade like this, cycling it down through a series of descending normalization temps will reduce the grain size and help re-distribute any alloy segregation.
