Jump to content

David Weaver

  • Posts

  • Joined

  • Last visited

  • Days Won


David Weaver last won the day on September 26 2023

David Weaver had the most liked content!


16 Good

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Thanks! There are machine made chisels in the boutique market for about $100-$150 per in that style. I'm curious if there is much of a market for handmade, as it starts to be worth thinking about as side money. But I don't have the time to do it for side money.
  2. Not today, but last week. Two mortise chisels (old oval bolstered for scale), both forged out of round stock and then ground. Longer of the two is O1, shorter is 52100 - both driven to as high as I can get hardness without bloating grain. 63 for O1 with two 400F tempers, and 64 for 52100. both work well in hardwood for their intended purpose - mortising. Wasn't initially sure if the 1/2: square-ish cross section of the bigger one would get high hardness through and through in brine. But 52100, even at high hardness, is probably better for a mortise chisel than O1. A professional woodworker wouldn't break either, but if I provide something like this to amateurs, it has to be made with someone over-assuming what "levering" means in mortising. They are differentially hardened to some extent, though. they're not unhardened near the bolster but they are probably closer to 55/60, which with at least the 52100 chisel means a break will end at the business end. Still would ruin the experience, though. I did hammer one out of round rod in about 15 minutes at lunch yesterday, and another one today. The brown handled one I like enough to make myself a set of 5. if you feel like aesthetics could be improved near the bolster, you're right - it could be tidied up, but there's no real precedent in tools for something like that. Mid to late 1800s would be a little less neat in most cases. Fair to say that I would neither measure up to a bladesmith or a real blacksmith, either. And probably never will have the skill to do either of those with the focus being on tools.
  3. precision's spec - https://www.precisionbrand.com/wp-content/uploads/2021/07/certs-drill-rod-water-hardened.pdf This doesn't say W1 that I could see, but it's what mcmaster carr's spec for their W1 is on their site, and a couple of the cut offs still had a "watercrat" (the brand name they use for this steel) was listed on the tag. EDIT - I see W1 in the footer at the bottom. My broken samples show carbides just bigger than 1095 and with slightly more disparity in size.
  4. so, for both W1 and 52100, I use an induction forge for heat treatment. There's no hold time, but rather normalization by cycling. * Forge * Two strong normalization cycles (like temperature around where you'd get scaling) and cool to nonmagnetic plus some for all but the quench, but not a complete cool and not down close to tempering temperature. Two minimum, but sometimes three or four. * quench in brine just as the last of the normalization cycle group is going to go to nonmagnetic. this is underheated for 52100 and will result in perhaps something like 64 out of a brine quench, but the point is that I so far see less distortion on the final quench if doing this. there can be some distortion here - I heat the steel and remove it in the next steps. * heat to the point of transitioning to magnetic multiple times. As low as I can get that point to be. It can be anywhere from three to six, and cools are to about black heat or at least when sensing there is no lack of magnetism. it's nice to have several pieces going at once for this so that you aren't standing staring at one piece. - the first one of these is where any significant distortion is removed by light tapping on the anvil - as straight as I can get things by eye * heat to nonmagnetic, then heat past. This is not well defined, I guess. You have to get a feel for how much a steel allows. 1084 will bloat almost instantly with higher heat. 26c3 is slower to do it, same with 80crv2, O1 is slower than 26c3, and 52100 seems to be at least as slow as O1 or slower. It also benefits the most from a push. I have not bought etch yet to polish and etch a sample and measure grain size, but 52100 is a fraction of the grains you'd find in a commercial nicholson file, at least by look - which is what one would expect, and it's still finer than just getting a sample out and heating it moderately and quenching. With W1, the brine is almost mandatory unless something is very thin. if something is very thin, I use parks 50 instead. 52100 in chisels has tolerated it fine - here is a mortise chisel I made yesterday. It's 52100, surprisingly 69 out of the quench (67 an hour out of the quench, 69 the next morning average of multiple strikes). 64 after a strong 400F double temper and I did knock the tip off to look at it -the grain is as fine as I've had with any other method, but the behavior is definitely changed by what is what.....lack of retained austenite or reduction in it? I don't know. I like the change for chisels. I would not know how to use an electric furnace without a bunch of experimenting as I've heat treated several hundred items and gotten to this over time testing hardness and snapping samples to look at grain size ,and then evaluating chisel performance (stronger/harder and tough enough is better in a woodworking chisel than toughest possible and then hardness secondary - well, at least for performance. Making consumer chisels that are hard to break without bending for proof of abuse might be different). I see on larrin's article that he was able to DET anneal 52100 and get over 67 hardness out of a sample. I think if he pushes heat up a little and uses a brine quench, he'll be at 69, so I don't think there is any rule breaking doing this other than that the last heat takes experience to learn, and this obviously isn't going to work for anything that really needs a long high temp normalization soak. I have samples that larrin tested when I was let's say 85% as good at this as I am now and they are good - O1 matches larrin's graphs, 26c3 is about twice as tough on average as larrin's charts - but i don't know if the high tail of those charts is real testing or software estimated. But I think 26c3 is not a great candidate for a furnace due to the potential for plate martensite, and the ability to shoot temp high with an induction forge or a high heat forge for a very short period of time may have some merit. By the way, the skill part of this is a little harder with 52100 because of how much range it has. W1 is easy because it doesn't really gain anything by pushing more than just a little past nonmagnetic for a short period of time. hardness is about the same either way, so no reason to risk bloating. it's important doing this to actually count time and note color and bloat grain, which I typically do. but I haven't for W1 because I already kind of handled that with 1095 and I have a hardness tester now. I'll do as little overshot as possible without sacrificing hardness. In a short heat overshot (like 15 seconds), there is more room than most people think....except 1084, which bloats almost instantly. ..... Which brings me to my last comment here... ....W1 is a really wide spec on internet charts. At least one supplier here in the US (precision industries) offers it in a much tighter spec - it's min 0.95% and max 1.05% carbon, so as long as you're getting the same thing from the same supplier and it's branded like their W1 is, you're safe. I have not bought any other W1 because it's too uncommon to find a melt sheet or something specifying what it is. if it's .75% carbon, I have no use for it. If it's 1.3%, I may not either, but I doubt much ever sells with much over 1%.
  5. I have to admit that when reading about intentional graphitization, it was for much higher carbon steels if I can recall - plain steels that are 1.5-2% carbon. I think. But I could be wrong. I did see that silver steel and those really higher carbon steels can be used with precipitation hardening. I'm just a guy in a garage but I've got relatively good skill hardening and tempering and am not afraid to push quench speed. Are you in europe? I think my personal issue mentioned in this thread is batch related, and would bet that if I were in europe and could just go to another supplier, the next 115crv3 rod would've been just what I wanted. In the states, i'm in limbo using W1 and 52100, but I have experimented enough with 52100 to push the hardness to high levels (68/69 out of the quench and 63/64 after a long double temper at 400F, or I guess that's about 205C if you're most other places in the world. ) That's quenching in brine, though - brine could not get my rods of 115crv3 to harden to spec or survive even tempering.
  6. going back to the picture of the plane iron, even though I think the composition of the steel outside of the carbides is different there, the carbides showing up as tadpoles (a round carbide leaves sort of a comet pattern as it protects the matrix behind it) don't look much different. In the 1095 in that case, the blotches show up in large bold spots like that seemingly at random. I could hammer 52100 here into a plane iron or hammer out some of the 125cr1 into a plane iron and do this, but I'm not sure if I want to go to that much effort yet. It's not that hard, but it would waste about an hour just to see if those spots would show up in a wear bevel. And 1.25% carbon isn't a steel that would have widespread appeal in a plane iron because this and 26c3 will have pretty much no abrasion resistance beyond any other plain steel. Thanks for the responses, by the way.
  7. >>the elements are still dissolved in that they are part of the matrix. They just aren't evenly dispersed. << This is what I'm trying to convey. I am working by observation and not education, so I didn't know how to describe it. A slither of chromium going through the steel in a "droplet" does seem unlikely, which is why I used the word "assume". I'm also assuming that if there was chromium like that, it would find a buddy either chemically or through dispersion and even if not uniformly distributed, mate with that buddy. Or to put it differently, I think the spot in the plane iron was poorly dispersed (undissolved isn't the right word, but that's what I used above I think - the stuff dissolves, but doesn't diffuse evenly into the matrix). So even if that were the case and visually something like this would occur, I have no reason at this point to believe it's going to be as drastic as the much more blatant smudge on the plane iron. Even that one doesn't result in voids, it just spoils a plane iron's ability to do even work. The iron is usable, but not for a demanding user wanting smooth with the iron and not sand or something later. What I'm leaning toward other than for 125cr1 is using steel that looks the best in the snapped samples, though. Not because I have a reason to do it, but because I don't have a reason not to. Testing chisels in actual use is easy, though, and I'll know the answer to this soon enough. And hopefully in the not too distant future, have some snapped samples of white #1. I have seen these smudges only in steels with chromium added, but not in all with chromium, obviously.
  8. yes, I'm being lazy and still haven't purchased nital. Let's assume this is not a carbide but undissolved elements, presuming that's possible, would that show up on a polished sample? I'm aware that you can highlight carbides and show grain boundaries. I'm sort of at a halt with this one because visually, there's not enough there to really know what this is for sure. Other than to mention that the bars are like vegas. As in, if these things appear in one bar, they stay in that bar, so to speak, and I don't see it seemingly at random from one bar to the next. But if it's as you say, it's possible that one of the bars breaks in a different way than the others. Because chisels fail with very small chipping (if they're good enough to gradually show only small defects after heavy use), it's not really possible either to say "aha, there's a defect that occurs at a bright smudge". More, do I see any higher defect rate at same hardness. Regardless of what it is, the chisel test is the critical test. If nothing is affected at the edge in heavy use, then this is just a curiosity. The bright spot on the 1095 plane iron is something I didn't know was there. I saw a strange dull line on wood where I expected just an uninterrupted bright surface. the dull line sent me looking for nicking in the iron and instead, I found that long bright spot. Actually, I should say I expected a deflection. Nicking leaves a very characteristic little ridge that you can typically feel with a fingernail.
  9. so more a geometry/orientation of breaking thing? here is a picture of 1095 at the edge of a plane iron, though the splotches in this steel were more solid looking. In order to get a picture like this, though, I have to make a plane iron, and then use it in a plane with a chipbreaker, which rubs the wood in a "scoop" shape and wears the edge the way it's shown here. I would estimate the wear shown here is only several thousandths of an inch. In this case, the shiny part also wore at a different rate than the rest of the steel. It's slightly further out at the edge and marked work. This bar stock was from AKS, and thus (unless something changed), the company making it isn't known. I never got a similar picture of this on the better scope - it doesn't just look like the others where it looks like "painted" grains, but rather like filled. I'm not sure it's just a geometric break pattern on the rods, though - anything is possible, but it's not variable. As in, if I snap more stock from the jantz rod, it will show the same thing every time. If I snap starrett O1, it will never show that. here's an example where I think the reflection is shearing or something at edges - apologies for the wordy picture. I put these three together just to send to a few people who were involved with brainstorming when I first started doing low temperature cycles. None of them heat treat, but one of them had read verhoeven for leisure just because he was interested in steel in general in tools and knives. The broken buderus sample in the middle shows those marks only at the edges of breaks, though. I'ts otherwise uniform. We were curious about the hock O1 irons (these are common in woodworking) because they look like they're either laser cut or blanked or something, and they are dirt cheap for anything made in the western world given their quality (wholesale for something like $25 per and retail for about $40). I suppose I could test my theory by viewing the bright splotch samples at an angle, but can say from the metallurgical pictures, the reflections do not change position when going from one scope to the next. The cheap scope pictures are never actually straight on, so the metallurgical scope and the cheap scopes are viewing at different angles.
  10. Jantz round 52100 (after forging - picture looks the same with a sliver sawn off of the round rod without forging and just heat treated quickly) same steel after forging, but lower magnification (don't have one at the same magnification). And a round bar of who-knows-what-52100 from ebay. I'm sure it's just someone cutting rods into short lengths. I went searching for others - didn't get to mcmaster yet - when seeing the splotches in the jantz round bar. This is clean stuff, though. Buderus flat stock from NJSB looks like this second clean sample, too. So far, round stock has a lower "clean snapped sample" rate than flat stock. 125cr1 is above is only the third sample of flat stock that I've seen (out of about 25? or 30?) that shows something other than uniformity. ..edit..I found a picture of the buderus. I have the last picture because I was excited at the time to have finally gotten 52100 to high hardness out of the quench (68 here, but I've seen 69 out of brine quenched 52100 from the same bar)....without bloating the grain that is. I have some work to do to see if there are any spotted difference in use. I like the jantz-stock finished chisels just fine. they're high hardness and I see very little indication of fault at the edge so far. So, I think this lack of dispersion, if that's what it is, doesn't really have much effect on woodworking tools and probably wouldn't on knives. and a previously shown picture from the indian metallurgical scope at high magnification. This, too, is from the jantz material with the bright marks. I'm fairly sure I posted this elsewhere on here, but it does show a good clearer view of the bright spot.
  11. Sorry I don't have better pictures. i have a few scopes (one legitimate, and then a couple of the hand held types that are $15 or so). None is both high clarity at high magnification plus significant depth of field. The better (indian metallurgical scope, the kind you get off of ebay for $450) can take very good pictures, but the field or depth of focal length is very shallow. I'm guessing it's intended to view flat samples, and it's very good at that, just not taking an expanse of rough torn surface. I can get pictures of the individual bright areas under the better scope, though, if it's helpful. As a rank amateur, I'm assuming these are just alloying elements that weren't well dissolved. There's not much in 125cr1, and I'm guessing it's not the manganese as this hardened just fine through and through and snapped (untempered) like glass. Or put differently, if I had to bet at lloyds, I'd say chromium. I'll show to of 52100 in a separate post.
  12. This is sort of an offshoot of the thread about the toughness of W1, and then some following things I've seen (just how uniform some 52100 rod is and how not some others is - does it make much difference in a chisel? I'll have to figure that out when I get a chance to really use one with unsightly snapped samples and one with really clean and uniform stuff. Unsightly as in bright spots that look like undispersed alloying, not large grain). So, I've had great luck with 26c3 chisels. I can't beat them with anything else, but I can't get 26c3 thicker than 0.25" either, can't get it in rounds, and that limits the tang thickness on chisels to less than I'd like. Even if using flat stock and forging on a bolster, I have to decide if I'm going to get better at smithing so that I can bunch up the material at the tang to something more like 0.35" square from 0.25" bar, or just use thicker bar. 125cr1 comes in a thicker bar. it's cheaper, too, but the cost difference isn't meaningful to me. I was hoping it would look as good in snapped samples of 26c3, but it doesn't. I have no evidence yet that it matters, but have observed better performance in something else (O1) when the sample is clean and neat. there is less small nicking in samples that look better - it's correlation at this point, not a conclusive statement that visually better samples make practically better tools. ---------------- this is 125cr1 at about 40-50x optical. The actual material thickness is about 0.18" For comparison, here is what 26c3 looks like at around 40-50x magnification (the little hand scope that I use is dishonestly rated, so their claims of magnification don't pan out - thus guessing). This is just a single quench - at one point, I would always single quench material without pushing the temperature very high to make sure that I could better it or at least not make it work with cycles. Now, I just cycle everything. the black stuff is just dirt. on the bottom. in the upper part of the sample, it may be in the steel itself. The same piece thermally cycled then and broken again. if I had access to unlimited charpy testing, I'd see if it makes a difference. I need to cycle the grain now after post forging normalization,anyway. Back to the 125cr1 - the splotches. what are they? at higher magnification, they look like this. Some are bigger than others, but they look more like "painted" sand rather than solid bits in the steel. This is very similar to the difference I've seen in one 52100 sample vs. another. hardness of both is about the same 68-69 in parks 50 in thinner cross sections and 69 or 70 in brine with less variance if cross sections are as thick as these. if the edge of chisels made of both show the same amount of defects after some given uses, then I will care less. if 125cr1 roundly spanks W1 in harder woods like 26c3 does, then same thing.
  13. I so much appreciate your point of view after mostly working in a vacuum. I started this in a vacuum of sorts, then got exposure to the "public" too late. it's not a religion to me, at least - just a matter of what is and what isn't, and what can be done and what can't be. if it can be done, then how, and everyone can choose what they want to do or try without pet methods needing to be the "one true and only" way. My view isn't discussion of this so much for knife makers or people smithing - everyone can make that decision on their own. It's more toward the odd woodworker who needs to make a couple of specialty tools and gets advice that the only thing they can do is have a machine shop make something from a pattern and then ship whatever they're making to peters or someone else like that. And thus nothing gets made, when the better quality bar stock O1, for example, can be cut and filed and does reasonably well if it's heated enough (but not too much) quenched and then tempered in an oven. We can all move on from there and start manipulating grain and so on (and forging, of course) but there's a lot of people who don't make things they could make because they're told they'd be junk. I'm still using my first iron. Not everything turned out to be great since then, but the move to making things got me away from constantly trolling old woodworking sites and pulling my hair out about something that's $30 but cost $60 to ship from England.
  14. It's definitely not BS (bad forge heat treat)- i've seen plenty of bad forge heat treat, and didn't get into forging things until long after getting fascinated with heat treatment. I'm throwing a theory out here, but I think if someone really romanticizes using an anvil and forging things, it'd be easy to be uninterested in really nailing heat treat because you're in love with being at the anvil. Larrin showed me pictures of the other samples he'd seen or evaluated and I more or less said "nothing of mine looks like that". Definitely not anti-furnace, either. I'm waiting for the reason to buy or make one and in the back of my head is pencils - I desperately want to make pencils entirely by hand including the leads. The pencil making part wasn't hard to figure out. These are just made with a little scraping plane that I made that grooves incense cedar halves, you glue them together with a lead in them (glue the lead, too!! or it'll slip out later if the wood is compressed and then expands), and then cut the bits free from their original boards and drag the square mess across the bottom of a plane until it's six sided. But what really makes a sublime pencil is a silky waxed lead that doesn't break, but feels like it's riding on a cushion. Like the lead in a blackwing. I could be a nut for drilling down on that, or wanting to, but leads have to be fired so that will be the thing that finally trips me over the edge. making a furnace at this point is starting to look more likely than buying one, though. Something that would do simple stainless steels in a quick soak like their datasheets say to would also be nice, but I just make so little in stainless, and AEB-L is amenable to a very high fast heat and quench. I will make knives for people from time to time, and have never gotten anything back but shock (because they're being compared to something like Victorinox). I was a little bit loaded for bear after success with the first samples and wanted to work up a process for AEB-L and XHP and get them tested by Larrin until I nailed it, but I quickly learned that I was effectively a portrait photographer marketing to the amish doing that. and if I were going to make knives and sell them rather than just filling the odd request from friends and then friends of friends, I'd just buy the furnace. I avoided forging for a while out of some fear that I would ruin the microstructure of the steel - which as mentioned, shows up in a tool you don't prefer to an older tool. And that whole thing drives my desire to make something people will prefer in general, and thus the antics about W1 here and wanting to really figure out what is the best that I can do with round rod.
  15. I got a reply from Larrin, who is the only public information source I know of using the same machine and methods to publish a bunch of stuff. Despite the original project showing someone saying they'd send W1 and W2, Larrin confirmed he didn't recall ever testing them. He didn't answer my second question of whether or not he was interested in testing some samples for hardness and toughness, but if he remembers our exchange about 26c3, he'd know I'm not using a furnace and probably not be interested, anyway. I have to take credit for something, though - and I may not be right in assuming credit for this, but I think I am. When I sent the first O1 and 26c3 samples, Larrin had mentioned he hadn't seen good samples heat treated in a forge. The next article he wrote, he said something like "only 20%" of the forge samples were suitable. My next batch was substandard because I thought I could probably just use the same method for O1 and 26c3 on 1084 and 1095, and that wasn't the case (what makes the first two excellent will bloat grain on 1084 especially - the 1095 sample wasn't that bad, it just didn't beat anything). And I think he was happy about that. Larrin's not the only source of information, but he is probably - with the long data sets and no cost to access information - one of the sources people will find early and maybe read the most. I was a little surprised when stumbling onto this forum to find that other people are still heat treating in a forge sometimes and taking it seriously. I got the sense that it was forbidden fruit or something.
  • Create New...