I just found a reference to Sorby using EN45. I need to check this stuff out. You know anything about it?
I've not looked up modern sorby stuff before, but it appears to be a steel with about 0.65% carbon and some additives for hardenability and toughness. That explains why the modern sorby chisels are soft if that's what's in them - it's more of an alloy that you'd expect to see in a tool box site chisel, and when optimally hardened, it'll end up in the high 50s hardness (which creates an edge that will have trouble in hardwoods).
As far as the old sheffield steels go, I have probably a 150 or so irons and have made about 30 or 40 bench planes with different irons (I rarely buy american irons from that time period - I think the auburn/ohio steel stock was not good quality and thus it's difficult to make the irons come close to sheffield or current stuff like Starrett O1). The really old irons are water hardening steel of some sort - I would guess the very early irons are chosen by ore (as in, use the ore that has the right alloying to make a good steel - without necessarily knowing why it does, but once you find it, keep using it). I'm not a reader of steel history, so I can only go by what I feel and what the microscope shows for wear pattern (which is both how evenly an edge wears and also whether or not carbides show up due to surplus carbon or other additives above a percent or so).
Late 1800s, things may have changed and the steel in laminated tools may have been more controlled or alloying adjusted without relying on the ore to have a certain property.
Once irons became solid in sheffield, they feel on the stones to me like they're oil hardening. As time goes on, they got a little softer (the solid steel end to end tapered wooden plane irons all seem to be a bit softer - someone working at the factory would've know why for that decision).
Some of stanley's irons in the early 1900s that are solid feel more like a water hardening steel and they have high hardness potential. I have some block plane irons that don't quite feel like 1095 or sheffield steel, but they're not as slick on the stones as O1 - they're probably one of the W-series of water hardening steels or something similar. I don't see much for carbides other than one here and there in the matrix (which is something that happens with tungsten).
I've never had the chance to play with W-1 steel - the spec is very wide and if you got some that didn't give the carbon content, it can literally vary and still meet spec within a range of something like 0.6 to 1.5. 0.6 and 0.8 are vastly different, let alone the range they give. I looked some up this morning and realized that if we want plane irons made of 1095 (I can send you one sometime - they're cheap and easy for me to make), one of the W series of steels would be a better option - they have only a little bit of alloying but much better toughness than 1095 - more like what you'd get in something like 1084 steel or A2 at high hardness (but without the coarsened grain and without the chromium).
I know there are natural ores with nickel, and natural ores with chromium. Either would've improved the outcome in the 1700s and 1800s by making steel a little tougher and more hardenable. There may have been manganese in them, too (which is in 1095 and others to make them more hardenable). It would've been an interesting time to make things - not knowing what made for a better outcome, but judging based only on the outcome itself. We now push specs which from an engineering standpoint can lead to improvement, but choosing spec first and then accepting outcome leads to stuff like A2 when you compare to something like an older ward or even older yet, a very old butcher laminated iron with extremely fine steel. The latter doesn't plane as long (maybe only 70% as many feet) ,but the edge is pleasant the entire time and sharpens easier at same hardness.
So, summary of sheffield steel:
early - weldable and water hardening - probably success in terms of fineness and toughness determined by ore source
later, like late 1800s - not sure, but still a lot of laminated tools - could still have been done the old way. There's little quality difference between something made around 1800 in sheffield and a 1900 ward iron. The difference I see is in average hardness (sharpening options would've been more starting mid 1800s with introduction of the washita allowing a harder iron, and then of course, synthetics by the late 1800s - starting to appear around that time. I don't know exactly when, and you could get corundum(alumina) before then as it had a lapidary use. When they started making vitrified stones like india stones, though, I'm not sure.
Later yet (before 1950 or so and after the early 1900s) - lots of solid steel irons that feel like water hardening and are often a click or two softer (not sure why - could be sharpenability, etc). As time went on toward the 1950s and marples, etc, were still marketing laminated irons, they were soft. That could've been lack of care, a nod to site workers, or to make grinding after heat treat faster.
After 1950, no-man's land. Lots of use of steels with lower carbon amounts (like EN45) that don't have the potential to have high strength, and thus won't hold a hard fine edge.
(Lie Nielsen's early irons were apparently W1 - the whole W series of steels is interesting as there are some versions that have carbon over 1% but a tighter spec and tungsten to improve toughness. Add a little tungsten, things get better. Add a little chromium, things get better. Add manganese, hardening is vastly more forgiving. Add silicon, and steel gets tougher. Add small amounts of vandium and peak hardness increases slightly and grain grows in heat less easily).
Add them all at one time, though and maybe the fancy sauce might not be that good. I think 1095 with just a bit of chromium added would be much better. 26c3 is basically a drastically cleaner spec of 1095 with a small addition of chromium and carbon at 1.25% instead of 0.9-1%. It's super dandy stuff, but requires quench oil to get full hardness and doesn't wear very long.
