It’s early February. Those two characters (W2) probably make people think of doing taxes. And the steel is probably very little known.
But I followed up the first pair of chisels for little boys with acromegaly trying W2. Why? 80crv2 is OK. It’s missing something hardness-wise, which results in a chisel that’s maybe sweeter on the stones, but I prefer something with some hardness and bite.
There are all kinds of options to try next, but I do not have a heavy blacksmithing setup, so getting a giant ball bearing or a piece of round stock 3″ wide and 3″ long and drawing it out is no bueno.
You’ve probably heard of O1, which I could get. O1 makes a decent chisel, but I’m looking for more than decent. It also lacks toughness (resistance to breaking -not chipping, but breaking from prying). I don’t think the toughness is a big deal as I’ve never broken anything in O1, but I want a notch up.
Day to day chisel making, for me at least, is 1.25% carbon 26c3. It’s unusual for chisels, but it makes a superb chisel. 80crv2 emphasizes toughness and, and there is probably a little bit more improvement in the hardness department, but if the window is that narrow to get it, I’ll leave it to plane irons. It makes a nice fine grained plane iron.
W2 – By Composition
W2 is, by composition, similar to a 1095 spec (0.9-1% carbon for the only retail source), but with the addition of vanadium, and in the case of what’s available, maybe even less hardenable (needing an even faster and more warp-risky quench).
It is low hardenability (needs a very fast quench to get high hardness), with small amounts of manganese and chromium added. And there are a few other alloying elements (a trace of tungsten, silicon, …). The vanadium is important to me. I can make a good chisel out of 1095 now. I couldn’t early on. It also suffers from toughness problems, but more importantly, I haven’t always seen uniformity in broken cross sections in 1095 – and those are the result of quality problems.
If you read the internet for a while, you’ll probably see a history of steels in the last 200 years that goes like (and this is curated for boutiquers – not a complete professional history).
Cast steel
Followed by W1 steel (not W2, but more on that in a second)
Followed by O1 steel
Followed by A2 (boutiquey), and Chrome Vanadium – a very generic term often derided, but the CV steels go anywhere from paint can opener quality alloys to razor and hard drill rod). And in this, is apparently, 80crv2 used commonly according to Larrin Thomas. We just don’t know who uses it. Probably European makers.
The W-steel groups and mild steel and pure iron (instead of wrought) weld electrically. Presumably, their presence came about due to industrial need.
O-1 is more simple – increasing hardenability makes it so that the steel is more stable and can be cooled more slowly and still get full hardness. This is a big deal to a manufacturer, and it’s important for a machine shop making one-off dies or parts. Oh – and also importantly, if you need a reasonably good die, it can harden in much thicker cross sections than cast steel and W1.
O1 is a good steel, and wonderfully easy to execute – and this discussion is keeping me from talking about W2 -but there’s a little something missing from it for me, both in irons and chisels. We are talking about the narrowest of things. I could make all of my tools out of O1 and work wood and never lack for something to use and make nice things efficiently. I think the same isn’t true for everything out there. For example, if I were actually going to do a large volume of work by hand only, I’d have no tolerance for Lee Valley’s V11 chisels. They work, of course, but their abrasion resistance is out of place on chisels and the edge doesn’t hold up as well as cheaper steels at same hardness.
That itself may sound odd to folks who have had O1 and V11 from Lee Valley – because LV specs O1 pretty soft and it doesn’t hold up well in apex critical things – like chiseling. That’s a choice on LV’s part – I could only speculate as to why – whether that’s a product of manufacturing ease dealing with O1 or if it’s a preference to have something that sharpens really easy at the cost of performance.
Oh – and that dimensional stability thing. O1 was seen as very stable compared to water hardening steels. But A2 and other air hardening steels as a follow-on are more stable yet, and O1 is fast becoming panned by commercial heat treatment services. Rob Lee mentioned the same thing to me (publicly on a forum) – that he likes O1 – but with his business hat on, he likes V11. When I quench XHP (which is V11 by Xray analysis results), it just stays straight. I get it.
So in the history of what we see used in boutique tools, stability wins. And 1095, 26c3 and now confirmed -W2 – are far from being stable in heat treat. They will warp if you don’t do everything right, and the faster you chase the quench and lower the temperature tail at speed, the more the warp. We can learn to deal with that – both in improving technique and in follow-up grinding.
So that brings us to W2 (vs. the W1 you’ll see mentioned everywhere). The original specs of these water hardening steels are very wide. That probably has to do with patents. 0.7% carbon to 1.5% carbon with ranges for other alloying elements. While the classification is wide, you may love a 0.9% carbon version and not 0.7% or especially 1.5%. I don’t order W1 steel because it’s not often found with a mill origin or certificate of actual composition.
But W2 is sold by New Jersey Steel Baron with batch certificates and a much tighter spec. So without being able to get my favorite (26c3) in 3/8″ bar stock, it’s just the thing to try. Carbon appears to be about 0.91-0.97%, manganese is half of what you might expect, and there’s a small amount of chromium to help hardenability and probably to keep some of the carbon in carbides and not in the matrix of the steel – too much carbon in solution and not in carbides leads to toughness problems. This is what is occurring in 1095 and O1.
And the steel is from Buderus and not “mill not named”. Good.
It sounds like…..
…..a plain steel that will require focusing on a simple but well executed heat treatment, an eye toward limiting warp, and rewarding chasing the steel from hot to cold as cold as possible and as fast as possible. And it’s not expensive, which is nice, but not that big of a deal for a hobbyist.
Could it be the steel that makes tools that feel like “old tools”? O1 doesn’t feel like old tools, nor does 80crv2, and I’ll admit, if 26c3 is landing at 64 hardness with a full double temper, it’s a bit hard compared to older tools. And the potential to use it up to 66 hardness after tempering – it will stifle sharpening stones.
Reading about W2 finds me landing at blade forums. That’s the site that I got banned from for talking about forge heat treating and being insistent that for simple steels, there’s no drawback. Interestingly, what I asked initially was if there is a “1095 with vanadium”. The answer there was no and my answer as to why (having the vanadium to pin grain size small and drive temperature past furnace soak temps just prior to quench to chase hardness), that’s what started trouble. “you can’t do that!!”. 26c3s results bettering furnace results (by a lot) and O1 matching wasn’t enough proof and nobody could seem to mention that W2 is available. So finding discussion of it there after the fact is humorous.
The discussion is littered with comments of not getting it hard enough, which isn’t a surprise – live by the furance, die by the furnace. Chase it slightly hotter than needed for a matter of only a few seconds and then quench as fast as possible and guess what -that concern went away. It’s bonkers hard. Right on the heels of 26c3 before tempering.
A brand new file will not touch it, not even the slightest anything on the sharpest corner. After a double temper at 375F, it still has a bit of a hard tempered attitude – just what we want in chisels. it’s a bit stifling for an india stone and skates on oilstones. That sounds like it’s too hard, but it allows use of the india stone to do minor work (grind for anything else), and an oilstone will polish and leave a blinding edge.
How the Chisels Turned Out
A review of what I want. A chisel that will not roll, but will not chip easily. 26c3 does this. A chisel with high hardness as that’s needed for holding a reasonably fine apex. 26c3 does this, of course. 80crv2 fell short in both of these a little. And excellent burr performance if possible – as in, a burr is raised on the middle stone and disappears on the fine stone without creating a nuisance after finish honing as softer steel might.
The hardness ended up higher than I anticipated so at this point, other than experimenting with some samples later and snapping to examine grain (more for longer-term consideration to use in both chisels and plane irons), the one thing that will expose large grain is chiseling something really hard. Like near water density wood across the grain.
26c3 handles this. 80crv2 rolls quickly.
W2 handled it just fine.

one 80crv2 chisel on the left – two in W2 at the right

The back sides. The left and right chisels are done and working. the one in the middle, I’m keeping along with the left. I’ll mail out the other one in the next two days. Chisel 2 is probably not prepared yet in this picture, but back flattening and setup was done after. Both w2 chisels are better than either 80crv2 chisel by a lot – which is a good thing. That was the objective.
These chisels are good enough that I’m not going to make two more in 125cr1, which NJSB listed just recently as an alternative to 26c3.
When I mention all of these alloys, I know it’s dizzying – without describing them and the characteristics, the discussion lacks resolution. But the details probably make for the need to ….make notes. I can’t really help that. Anyone who has worked through these discussions of steel will be long past “1095 is for saws, O1, A2 and V11 are for chisels”.
We’re not really looking for light and airy at this point – we’re looking for results, differentiation and learning. As one of my college professors said (in a challenging class, where it always seemed like he kept us confused and thinking hard) “learning hurts”. I have one goal when making chisels that stands out to me with everything else secondary. Can the chisel that I just made match anything vintage that I’ve seen and better anything current outside of japan. The answer for 80crv2 was no. So far for W2, the answer is yes. Finding that is what I want.
How do I Test?
I test chisels right off of the first grind. That’s two-pronged. First, it should be the worst part of the first several inches of chisel length, and second, if I can’t make that part workable, then from the making standpoint, I need to revise what I’m doing. It may be true that commercial chisels or irons can be lacking for some length, but that’s preventable.
If they are better a little further in, that’s fine – but I want the first grind and sharpen to have ideal characteristics.
I test in order:
- By feel of the grind – if a chisel is soft, I will be able to tell finish grinding. This is unfortunate because it lets the air out of the balloon a little early. By soft, I don’t mean it’s 53 hardness instead of 63, I mean if it’s 60 and I’m hoping for 63, you can feel a pretty significant difference in how the chisel feels while grinding, and of course, some difference in the speed. Softer leads to more of a bite from ceramic belts, and harder more of a skate.
- By feel of honing. For plain steels, a little bit of skating and not much steel removal on a fine india, hardness is high (62+, perhaps 63+). Skating on oilstones mostly aside from honing peaky scratches, also same hardness. Anything less, and it’s a matter of how much. Once the burr is established, in this high hardness range, it will generally come off of a plain steel on mid-fine oilstone and what is raised further will not be large. Softer can make a decent chisel, but it’s not my objective.
- Buffing off the burr – the amount of “stripe” made at the very apex is highly dependent on hardness. Subjectively 59/60 hardness will buff twice as fast as 63/64.
And then tests in wood:
- Paring hardwood (picture below). i try to pick something that is hard enough to be differentiable. Not pine or poplar.
- Malleting a volume of harder cherry or hard maple. Cherry isn’t as hard as maple, but what damages an edge on one malleting seems to work on the other. Put differently, if I mallet a volume of wood and see any sizable defects, I’ll find it in either one. Zero defects in an edge is acceptable. It takes a fine microscope (not a 40x loupe, but something more like 150x optical to really differentiate and fine if there is nicking and how much. Sharpening removes a thousandth – anything more is a hassle if it’s avoidable. That is, a long interval of heavy use should be addressed by routine sharpening. A mediocre chisel won’t meet that standard unless it’s made pretty blunt).
- Paring wood again after malleting is also fine – the surface should remain line free.
- If that is passed, I have some older rosewood that is more than 90% of the density of water. It will destroy the edge of a mediocre chisel malleting across the grain. If a chisel is really good (a properly made japanese chisel, or one made out of very good quality files -properly – or 26c3, or the best of vintage chisels), it will tolerate some amount of malleting with a regular (not steepened) edge and take no more damage than the depth of regular sharpening.
How important are these tests? For you, maybe no big deal. For me as a maker, or some kind of critical comparisons of what you’re doing – furniture, fly rods or whatever – they are what will make you better. If you don’t have the fire for that, you’re destined to end up talking about making things on forums and not making much.


















