Thread: Converting spent files into chisels

If my Dad was alive today he would be 108. When he was 14 he went to work with his uncle as a Butcher/Slaughterman. First thing his uncle did was give him an old rat tail file, a brick batt and a handle off an old crank phone. He spent every spare hour rubbing down the file with the brick until smooth , then fitted the old phone handle as a grip. He now had a sharpening steel. His pride and joy. I still have and use the steel today. I also have a Dick flat steel but like to finish on Dads old file jobby and haven't found a smooth steel I like better. He also made his knives from files or old leaf springs depending on their use. A good rub down a few times with mutton fat and they never seem to rust. Amazing.

I'll stay mostly out of the discussion because I've gone off of the deep end as far as heat treatment goes.

But, as Ian mentions a good working chisel without tempering - partial conversion, so part martensite and probably some pearlite or some other structure. if it ain't broke, don't fix it. In practice when pushing things, full martensite conversion and tempering back is better, but if there is no problem in use, what good is "better than just fine".

Files are water hardening steel - the older files typically something like Carbon 130, which is still on spec sheets. Very similar to japanese white. (carbon 130 also C125U and 1.1563, but good luck finding it).

Small amount of manganese, small amount of chromium - both for hardenability. Quench potential is probably as high as 70, and 65 with double 400F temper.

Water hardening is just a term, though - thin cross sections transitioning to high hardness in oil are water hardening steel with enough hardenability to do that.

Heating a file to just past nonmagnetic and then quenching in brine is a legitimate option, though.

If concerned about brittleness, heat the section to be heated five times just to nonmagnetic and then let cool for about a minute to magnetic again. Takes a couple of minutes, and can drastically reduce grain size and improve stability of an edge.

What takes bobbing up and down in fast oil to be hardened can just be pushed into brine, though, and the outright speed of the brine (warping) will be offset by the lack of having to move a tool around. I always get a straight result on the second quench, and so just quench once before the cycles I just mentioned, and then the final quench (of chisels) in brine doesn't cause a problem. Brine being 10-15% salt by weight - faster than plain water but seemingly much more organized and less of a crack threat.

Like you say, David, "if it ain't broke...."

I have almost certainly reached the peak of my chisel-making career - I had an immediate need for the first one I made (I needed a very thin chisel about 3/4" wide for a specific job). Because it turned out so well I made it a couple of smaller companions, just for fun & to see if I could repeat what I expected was a mere fluke. If I were starting out in my woodworking career all over, with the sort of miniscule tool budget I began with back then, I would seriously consider making myself a set of paring chisels from files (I loved the set you made a year or two back), but I lucked into a 'set' of socket New-Havens many years ago & they make superb parers, so that itch is already well-scratched.

I'm not sure my primitive metallurgical skills would hold up making a larger chisel anyway, it seems to me, after making quite a few plane blades that the smaller pieces are easier to heat-treat - they heat more quickly & evenly in the relevant area & quench quickly & evenly. Up to a point, that is, with very small pieces I have trouble keeping them at the right temperature between forge & quench, even with the tin of quenching fluid right beside the mouth of the 'forge'. My shed opens to the prevailing wind direction, which is usually a good thing but if there's a breeze, things can become very awkward for heat-treating.

I can see how easily I could get sucked-in to the allure & mystery of more advanced metal working & experimental with different alloys, but wood is far & away my preferred medium, my metal-working is definitely just a means to further that end, so I have resisted that temptation successfully (so far).

Cheers,
Ian

Originally Posted by IanW

Like you say, David, "if it ain't broke...."

I have almost certainly reached the peak of my chisel-making career - I had an immediate need for the first one I made (I needed a very thin chisel about 3/4" wide for a specific job). Because it turned out so well I made it a couple of smaller companions, just for fun & to see if I could repeat what I expected was a mere fluke. If I were starting out in my woodworking career all over, with the sort of miniscule tool budget I began with back then, I would seriously consider making myself a set of paring chisels from files (I loved the set you made a year or two back), but I lucked into a 'set' of socket New-Havens many years ago & they make superb parers, so that itch is already well-scratched.

I'm not sure my primitive metallurgical skills would hold up making a larger chisel anyway, it seems to me, after making quite a few plane blades that the smaller pieces are easier to heat-treat - they heat more quickly & evenly in the relevant area & quench quickly & evenly. Up to a point, that is, with very small pieces I have trouble keeping them at the right temperature between forge & quench, even with the tin of quenching fluid right beside the mouth of the 'forge'. My shed opens to the prevailing wind direction, which is usually a good thing but if there's a breeze, things can become very awkward for heat-treating.

I can see how easily I could get sucked-in to the allure & mystery of more advanced metal working & experimental with different alloys, but wood is far & away my preferred medium, my metal-working is definitely just a means to further that end, so I have resisted that temptation successfully (so far).

Cheers,
Ian

That may be the case (for you) - but there are probably others who will pick up a file or other steel and not get results that they want. I guess the default here in the states would be O1, but that's not free like an old used file.

Using a file without any annealing and just grinding to shape with a strong sanding belt (ceramic being a good idea) and a bucket or a light spray on the belt is also perfectly fine.

I can't really comment much about various alloys - 52100 is about the most complex thing I've mastered - and if you guys have euro suppliers of steel down there, you're ahead of me as there's a whole world of file-like steel sold in europe that I can't get here in round rods.

Chisels - yes, full length hardening definitely some difference vs. plane irons.

The usefulness of brine in hard to transition steels - and some files just won't cooperate with medium and medium fast oils, or even fast oils - was a recent surprise to me. It should allow close to full hardness with less of a heat. Probably not suggested for O1 - haven't tried it. I thought it would crack everything, but with files, 52100, W1 so far (there's a tight spec W1 sold in rods here - basically a 1% carbon steel rather than the spec's wide range) - i haven't actually cracked a single thing in brine. It's a potential curiosity for me later with tapered irons as oil and a 3/16" taper plane iron are a no go unless using O1.

Too, the brine conversion with the right heat is complete enough that it is starting to be reminiscent of older english tools, they way they sharpen easy with a very dry feel and they cast aside their wire edge faster than underpants in a nudist colony.

Brine has been a revelation for 52100, which gives me a good option for forged mortise chisels. 69 out of the quench, 64 double tempered at 400F, and the chewy bendy nature of 52100 is "cooked out of it" by converting more of the final result to martensite. Experimentation to this point mostly necessity out of the much narrower range of round rod here (to make one piece chisels without folding on a bolster) vs. flat stock. We can even get hitachi white 1 flat stock now. At any rate, the results for 52100 are off of the commercial charts.

W1 and W2 also benefit a lot from brine. I asked George here about it as W1-like steels and some 1095 are iffy through hardening in oil, and said "I recall reading over and over that brine is a bad idea, but I haven't had any trouble with W1 yet in it" and he quickly said "I never had any trouble with brine and W1, either. Not once".

if just a limited number is needed, though, I'd do exactly what you're doing. The more typical suggestion in the past of trying to find a reliable heat treater for a few small chisels or a couple of plane irons is insane and limiting. the bolster issue is a pretty big hump to get over for a few chisels and you can, of course, choose a heavy ferrule material and good hard handle wood and sidestep it. I have an aversion to chisels without a bolster, but it's not on the shop made side - it's on things like Blue spruce selling for high price with no bolster because the chisels are more or less just a bunch of sourced lego pieces at mediocre spec. I would make hand made hand forged chisels for the price they charge and be happy with the margin. Or in the case of LV, for what LV charges for some very awkward mortise chisels made out of a steel that probably should never be used for mortising and limits the ability to make anything but really fat mortise chisels.

I cracked a couple in brine, but that was rehardening commercial chisels using an oxy-propane torch and home made "Super-quench."

I did have other problems, such as excessively coarse grain, and I never did anything about that either.

They did get very hard, though. I still had problems with excessive chipping after temper, which I now attribute to my quick heat and dunk without doing any sort of multiple normalizations or multiple tempers. I keep thinking I ought to have another go at it, if for no other reason than to see what could be if these fellows sorted out their process... But that would require me to sort out my process.

Originally Posted by truckjohn

I cracked a couple in brine, but that was rehardening commercial chisels using an oxy-propane torch and home made "Super-quench."

I did have other problems, such as excessively coarse grain, and I never did anything about that either.

They did get very hard, though. I still had problems with excessive chipping after temper, which I now attribute to my quick heat and dunk without doing any sort of multiple normalizations or multiple tempers. I keep thinking I ought to have another go at it, if for no other reason than to see what could be if these fellows sorted out their process... But that would require me to sort out my process.

yeah, not much survives coarse grain. if you break a commercial file and keep the grain as a sample, it'll give you a target. If anything is discernibly more coarse, you will find it out in use - it just depends on how much. I typically see about half that size in terms of coarseness under the microscope after thermal cycling, and haven't hardened anything in a while without thermal cycling - it's at this part part of the straightening and stabilizing process, as well austenitizing.

Even heat is mandatory for even time, but that is doable by hand. I guess that's really one of the keys - controlling things and then whatever you can see as evenly as possible by eye is going to be better quality than a lot of commercial heat treatment.

Too, when you get large grain, snap a sample large.

then heat the steel until it's barely nonmagnetic - check the color creeping up slowly so that you're not tempted to use an oxy combo torch and overshoot it too much - then do five cycles of that with air cooling back to black (no need to fully cool) and complete magnetism. You'll be shocked. the induction forge really speeds this process up - it's not part of most commercial cycles, but it's also the part of being able to match or better commercial cycles on simple steels.

I bring up the color because when I first moved to the induction forge, I ran it full on high no matter what. I'm always trying to go faster as most people will do and the several seconds or so past a temp where magnetism will be lost (but it will take several seconds) allows for overheating.

I guess I should ask - how thin is the file - if it's a thin one, then maybe there is a too thin and it will crack. I haven't done anything but chisels, which are about 0.3" at one end and .15 at the other, and wedge shaped as far as thickness goes. there's no long flat .10 thickness area, and I definitely have some smaller metal files that are super hard and that thin.

one last comment regarding files and brine. This isn't necessarily beginner's fodder, but it goes along with brine. and I guess I have an advantage now a portable diamond cone tester in house.

I have long talked about pushing heat as far as you can get away with in a quick heat - before grain bloats, and you have to count time and color and make mental notes or keep notes to find out when grain bloats.

the reason for that push is a lot of water hardening stuff is sucking wind to hit the upper end of the quench hardness schedule in fast oil, and you can actually increase the time needed to get past the upper end of the quench (below the point where pearlite will form and ruin the result) if you increase temperature at quench. So if you have a fast steel that needs a second and it has to decline 600 degrees, you may get a two second open time with another 200 degrees. it sounds like a small difference, but it can be a yes or no kind of thing, or pass fail.

With brine and steels that show a low temperature soak before quench, like Carbon130, 1084, W1, etc, you have the quench speed and you only need to get truly nonmagnetic and then quench and you'll be within a point.

so point here being just get the nonmagnetism, evenly, and then get the thing being hardened into brine. W1 will hit 68/69 with a very relaxed heat if it's into brine with a little up and down agitation. In parks 50, it wouldn't even harden more than superficially without the push (precision industries W1 is almost identical to the W2 sold by NJSB, just slightly higher carbon, but it's hardenability is very low...on purpose).

some of the Carbon 130 samples can be just as low hardenability as that and then some others have .1 or .15% more manganese and they harden in oil - it's a crapshoot.

Some nice fellow on bladesmith forum did a check of four files and just as you (and I) would expect, none of them were lower carbon - they were all 1.2% carbon or more. And even the cheapest chinese files were very clean steel - they just had more chromium added, which I don't think is positive, but it must assist modern heat treatment. Instead of being like .15% chromium, they were more like 0.7%, which is a huge difference, but they were also >1.3% carbon by just a little - some of that increase probably to match up with the extra chromium.

Yeah, well, you see why I don't want to get serious about making chisels and delving into the endless mysteries of steel alloys & their arcane heat-treatment requirements!

For a mug like me, whose only requirement is a couple of odd-sized plane blades now & then, or a non-commercial tool like the chisels I just made, there are cheap & pretty idiot-proof choices (anything that works for me has got to be idiot-proof!). The 1084 I've been using for plane blades seems to fit the near-idiot-proof requirement and costs just a few $$s per blade. The blades I've made with it are surprisingly good, much better than I expected. I'm getting a very good bang for my buck there.

Spent files are a cost-free & readily-available raw material, so you can muck about & destroy a few until you get a tolerable result. The former scientist in me would like to know what hardness I'm hitting & what the grain structure looks like,etc, but the lazy woodworker in me is quite satisfied when a blade works well & the numbers become irrelevant. I guess I have enough experience now to get a crude estimate of harness on the sharpening stones by the speed of steel removal & the size of the burr formed on different grits. By sheer luck, and the tolerant properties of the steel, I got a fine enough grain to take a fine edge off an 8,000 water stone, and one of the chisels I made is as good as any chisel I own when it comes to maintaining it.

If I tried to make several sets without refining my methods I know I'd hit trouble. I'm all too-aware that repeatability and best results come from meticulous technique, but for an occasional blade or two I can afford to muck about a little - if it don't come right the first time, just do it over. You certainly would not wish to adopt that approach if you were trying to make chisels for a living!

Cheers,

I suppose that's probably how things come across (that it's complicated), though a couple of basic things make it less so. I guess it falls into worth learning those things if you're going to make a dozen or so tools and they're not just O1, down to perhaps worth knowing on the first tool if it's the case of an uncommon tool and poor heat treatment. Not that frequent from professionally made very old tools, but can be the consequence of finagling between new and when one gets a tool.

It's kind of like planemaking in general. Planemaking sounds like a great thing to try to "make yourself a set of planes". You then learn that you can make yourself a set of planes you can use, or you can make five sets and get to making fantastic planes if you refine each iteration.

It's more tolerable to use a good-enough chisel than a good-enough smoothing plane or jointer, though.

David, I reckon there are two aspects one discovers when you decide to get serious about making any type of tool.

With familiarity, what seemed mysterious & complex to begin with, often becomes far less so as you learn about a process, particularly if you approach it rationally & try to understand the basic principles involved. I'm guessing that is what you are alluding to in your opening sentence.

And of course, "practice makes perfect" so the more you practice, the better you get at anything. But it's become a bit of an "arms race" for me - as I get better I also become more critical, so what I thought was "good" or even "very good" 5 years ago is a bit ordinary today. Once you do get reasonably competent at something, the rate of improvement slows & becomes finely incremental, so at my present rate of progress, the "perfect plane" is still about 50 years away.......

Cheers,

one thing that stands out to me, Ian. I am pushing the limits of what can be gotten out of alloys without bloating grain - are other people doing it, too? Probably. you can't really do it with simple steels and an electric furnace because you can't chase more carbon into solution quickly -it's more what soak and how long.

so, I am getting things that I wouldn't have expected before. 69 and occasionally maxing out my hardness tester at 70-ish with 52100, and I haven't tried brine on the 26c3 (1.25% carbon) stuff, maybe it won't get any better than 52100. What's odd is being able to get 64 hardness out of that 52100 after tempering it really as far as I'd practically want to go. Two long tempers at 400F.

I'm sure the first time I quenched 52100, I didn't get 64, and probably not close out of the quench.

The first time I did a large O1 iron becomes the point here - I could get a lot of BTUs but not enough temperature and I eventually got it nonmagnetic with a giant weed torch after feeling like my skin was starting to crisp. It hardened - partial transformation, and is probably about 60 hardness. I put it in a plane and just used it out of concern it wasn't hard enough even for that and it worked well enough for most medium hardwoods. I sort of forgot about it other than remembering it wasn't tempered.

Then as I got deeper in, I was able to plane wood with a more recent iron just slightly harder, but that held up better.

is it important? it must not have been because most of the time, I just sharpened the other iron and planed with it.

when you mentioned you got the same result with a file, but a step better (it works great), it made me recall that - sometimes coming up short isn't really coming up short.

i could see the partial transformation working very well on file steel. It doesn't work well on 52100 unless refusing to break is better than quality of the edge - sometimes it is if it's not woodworking. Would be good on a machete. What's worse in a thicket - a knife broken off at the handles or one with a little bit of edge rolling.

Hmm, that was a question I had some time ago when I started heat-treating steels like 1084, but it got lost in the conversation at the time - what does 'partial transformation' actually mean in terms of edge performance? I suppose that would depend very much on the actual alloy concerned?

I was worried that if my blades are coming out of the quench at a hardness that doesn't seem to require tempering to reduce brittleness, is the resulting metal structure less durable and/or less fine than from a 'full' hardening & tempering? It doesn't seem to be, the 'file chisels' are taking an excellent edge & standing up very well (as parers, I haven't tried whacking them into a lump of ironbark & never will!) so they could be just a bit more brittle than optimum & I haven't discovered that yet.

So is this what they mean by steels like 1084 & O1 being more "forgiving" in that they will tolerate less stringent conditions and still produce tolerable edges?

Cheers,

I could abuse the concepts verbally here, but there are two ways to end up with something that's about right hardness without tempering:
1) underheat the steel before quench
2) quench is too slow to get full transformation

We're generally chasing carbides and tempered martensite. The two cases above may result in some tempered martensite and carbides, but not the same everything else. Could be pearlite or retained austenite, I don't know- but something that doesn't harden.

So, we have two theoretical cases:
* tempered martensite with nothing else (needs to tempered) maybe aside from carbides. This is actually near attainable with something like W1 and probably 1084 and 1095, etc.
* a mix of untempered martensite and other stuff

If both are the same hardness, the first bullet point results in steel that is stable and per level of hardness by a diamond cone test, actually has more strength and stability in an edge. In an extreme example, a super thin chisel or a straight razor would seem hard enough, but the edge would deflect more easily.

Stability has more to do with whether the steel will stay the same in use. I don't know if we really challenge this - we're constantly sharpening off the steel we're using and the rest of the tool isn't stressed much. If we have a tool that we're stressing more than the apex or just behind it, then probably we're using the wrong tool.

I think for woodworking tools, we're kind of looking for something that behaves more like a good razor steel.

My underhardened and untempered iron is fine in most things. It's less nice to sharpen (the wire edge is a little more persistent), but that's tolerable. In hardwoods with anything challenging, it shows defects sooner, though.

As for the two choices of what's easy to harden:
1) 1084 - needs only to get just to magnetic and then be quenched fast. Fast for really thin things means the speed of going hot to cold and may not actually require a fast oil or water. There is no grain growth protection in it, though, so heating a color past nonmagnetic even for 15 seconds will change how good the result is
2) O1 is like a photographic negative - there is so much manganese in, but it also needs a higher heat to attain schedule hardness, which is listed as anywhere from 64-66 depending on the soak temperature. To heat a shade post nonmagnetic in the critical part of a tool for 15 seconds gives a better result. higher hardness, more stable edge, and does not result in grain growth. Full hardness should occur in almost any reasonable oil choice whether a tool is .25" thick or .08" thick at the business end

What makes each easy is different. Heat treating one like the other is a recipe for mediocrity. 1084 can have high toughness with good hardness, but if the grain is bloated just a moderate amount, it can literally limit toughness to a fifth or a tenth of what it could have been. O1 on the other hand doesn't have high toughness potential, it has moderate toughness potential and pretty good hardness potential - not out of the realm of good taste to use it at 63 hardness. When it's underhardened, you get something like a hardness spec of 59-60 and it's a no man's land. This is why Lee Valley's O1 tools are blah. They're either underheated when hardening or overtempered. Both are counterproductive.

Last bit - why I think O1 and files are OK if there is a partial transformation and 52100, for example, is miserable - the properties of what's not martensite in them don't result in super toughness, so they don't have edges that are bendy and create a foil as easily as something like 52100 or 1084.

I think most people bloat 1084 grain if they get it hard enough and never get a sense of its potential toughness. 52100 is a little different in that it sequesters its carbon and doesn't have much floating out in the steel - relatively (like perhaps 0.6% of a 1.05% total carbon content) whereas O1 and files and 1095 definitely aren't short of it at all.

https://i0.wp.com/knifesteelnerds.co...pg?w=752&ssl=1

If you look at this chart, you can see the lower lines - if you move to the left, 1095 and O1 don't gain much toughness. 52100 becomes bonkers tough. I think accidental partial transformation results in less aggravation when a steel will still not hold on to every single little burr or roll that occurs at a microscope level. 52100 instead holds on to whatever starts to deform and won't let go of it. I have pictures of that occurring, but not on this PC. It can be leaning toward being unpleasant to sharpen on a natural stone and still act like that whereas O1 might behave "a little soft".

File steel would be similar to the 1095/O1 line. V11 (XHP) is also somewhere down there, but it suffers from too much carbide volume at the edge, so it's equal toughness in a side impact test, but less good in a fine edge if compared to equal hardness 1% more plain type steels. Or in the case of files, the contest of something like 26c3 (cleaned up file steel) at 64 in a chisel and V11 - not close. Same thing for white 1 when done properly, it's much like the file steel. High hardness, little "not martensite", the carbides are iron carbides and not as easy to crack and in lower volume, and you get excellent edge strength and less that wants to fail at the edge.

Ok David, so I take it that under-hardening & not tempering can result in a decent blade in some steels, not in others, and even though hardness & toughness may not be theoretically optimum, they can be pretty good (again depending on the steel). I have no way of accurately determining temperature, so just rely on a colour I know by pre-checking is above non-magnetic, and in any case, is about the best I can get with a medium-sized blade in my Heath-Robinson 'furnace'.

Unfortunately, I no longer have access to the sort of microscope gear I had in my working days, which included anything up to scanning EM, so can't really study my results as closely as you do. Actually, that's probably fortunate not unfortunate, I can see I could easily get taken up in the minutiae of heat-treating & end up down rabbit holes I have no need to explore! That leaves me with the empirical approach (aka 'suck it & see') which I accept has many limitations, but with at least 60 years of sharpening & using tools behind me, I have a vague idea of what's likely to be usable & what's not as soon as it hits the sharpening stones. I know burr size & persistence are not absolute measures of edge performance, but they've been a pretty good guide, in my experience. If a blade takes a good edge & holds it for a reasonable period in woods like gidgee or she-oak, that's certainly good enough for me.

So the take-home message I get is that the neophyte smith can achieve good to very good results with O1 & 1084, with very basic equipment. And even if not the best possible, 'very good' can be good enough for even demanding woodworkers. If I understood you correctly, O1 is perhaps the more forgiving of the two, but it's 3 times the price of 1084 here, and available in a far more limited range of sizes, so based on the good results I've had with 1084 & how inexpensive it is, I'll stick with it for now.

What I'd like to get across to folks who'd like to try making a few blades for themselves but may be intimidated by the apparent complexity is that it's quite easy to get very acceptable results with a minimalist approach & minimal equipment. Trial & error will soon get you to the point where results become acceptable, so don't give up if your first couple of attempts aren't successful (as I almost did!). Being able to make your own blades can be a liberating & fun thing.....

Cheers,