Jump to content


  • Posts

  • Joined

  • Last visited

  • Days Won


Everything posted by timgunn

  1. Sounds like Charles has it covered then. If you do go for the Firestorm, please will you post some pics and your impression of its performance?
  2. Just looking at the Atlas website, the burner(s) on the Firestorms are very similar to the burner on the Atlas Mini. If you find the Atlas Mini has all the capability you want/need except for the size, the Firestorm is probably a good choice. Note that the Atlas Mini and Firestorm burners do not have chokes (they are Naturally Aspirated burners) and run a fixed air:fuel ratio. This means you have no means of adjusting the flame temperature. Your only adjustment is how much mixture you feed: gas pressure adjustment. In case you don't know, the gas flow varies as the square root of the pressure, not linearly with pressure. I often see comments along the lines of "I was running at 20 PSI and now I can run at 10 PSI, using half the gas". In fact, having the pressure will reduce the gas flow by just under 30% (to 70.7%; One divided by the square root of Two is 0.707). To halve the gas flow, you'd need one-quarter the pressure: 5 PSI. Having a choke on a Naturally Aspirated burner allows the operator to alter the air:fuel ratio, changing the flame temperature and the composition of the forge atmosphere. Having independent adjustments for Fuel and air on a blown burner does the same thing. The Graham burner is a blown burner and looks like a much more versatile piece of kit for bladesmithing to me. I don't think I know many bladesmiths who would choose a fixed-air:fuel-ratio burner over one with variable air:fuel ratio after using a variable one. It is worth noting that burners can be changed later, if/when the forge needs to do something it cannot do with the original burner, so long as the new burner will fit in the old burner port. I think I'd be checking whether the Graham burner will fit the Firestorm.
  3. Mainly, they snake out. I have had one migrate lengthwise though: I'd missed with a couple of the staples and probably had about 6-7" between the ones that I'd got right (I blame age, poor close-up vision and a lousy viewing angle, but mostly age). The coils moved and bunched up much closer towards one end (staple) than the other. I was on 16AWG elements and a fairly narrow groove: probably one of the first couple I built with the filed-in grooves. I don't know what the mechanism was that made them bunch up, but I'm guessing that inconsistent groove width played a part, and that the movement occurred during heat/cool cycles at/between uses, rather than during a HT. They didn't close up enough to short out coils, but there was obviously more heating where they were close together and, given that I was trying for absolutely even heating, it was enough to register that I didn't want to do that again. I don't think you'll have any particular problem with an exposed junction: Evenheat et al have been using Exposed junction type K thermocouples for a long time AFAIK. Type K drift doesn't seem to get considered much. I am acutely aware of it because my day-job involves processes with long periods at high temperatures. We used to control 24/7 at 1200 degC (2192 degF) with type S thermocouples back in the 1990s, but reduced to 1000 degC (2012 degF) for reduced NOx emissions by the early 2000s. We did some work to compare Types K, N and S at 1000 degC. Type N were MUCH better than Type K, in that they agreed closely with the type S for months, where the type K were starting to show drift within days. The type S, being Platinum-based, were about ten times the price of the Nickel-based type K or N. We settled on type N. HT ovens are a different process, I have to say, but I can't ignore my experience at work and I use Type N for HT ovens. As I understand things, Type N was developed for use in MI assemblies, along with the Nicrobell sheath material. I was wrong in my earlier post: I get Type N thermocouples with Nicrobell sheaths from my local supplier. I get type K with the 310 Stainless sheath for use in forges (1300 degC seems to be a good welding temperature for bladesmithing beginners and is the upper limit for type N. Type K will go up to 1372 degC and is therefore a bit more useful for tuning forges).
  4. If you've got a way to deal with the dust, and it sounds like you have, don't faff about with pullsaws and stuff: use a router. I've built several HT ovens, mostly using IFB. The first one or two I used a saw to cut each side of the grooves, broke out the middle bit and filed in the bottom of the grooves with a length of bent studding (allthread). That got really old, really quickly. For number three, I used a router and have done so on every one since. Where I've used CF board, it's been for doors and roofs, with no grooves in the board, but I don't have good enough dust control to feel comfortable routing the fibre board. I have to say that my experience has been that you really do need the staples. I keep the staple spacing to under 3", using a 10mm or 3/8" OD coil of Kanthal A1, initially in 16AWG and later 1.6mm (about 14 AWG). When I've tried to get away with wider spacing, things have not gone well. I am fairly certain that 12 kW is going to be waaay too much for realistic control. I'd suggest 2 x 6 kW coils in parallel to give the option of running them in series at 3 kW total if you find it's the case (I don't expect my opinion to make any difference to your using the 12 kW initial setup). I'd keep the cycle time as short as possible and use SSD(s) for switching. I use a 2-second output cycle time. Your thermocouple looks like a 6mm/1/4" Mineral Insulated assembly. I'm guessing it has an insulated junction? With 12 kW of heating, I'm inclined to think the thermocouple response will be too slow. The control thermocouple needs to be the fastest-responding thing in there. If it responds slowly and damps out the temperature fluctuations, you'll get the temperature at the edge of the blade (lots of surface area, very thin, so minimal thermal mass) going up and down with the power on/power off cycles of the elements, with the damping effect of the slow-responding thermocouple only allowing the controller to see the average temperature over the last ten or twenty seconds. I usually use a 3mm or 1/8" Mineral Insulated type N thermocouple (developed to be an "improved type K", with much less tendency to drift) with grounded junction to get a fast enough response (and I "only" use 3 kW of elements for a 23" or 28" HT oven). Sheath material is either the proprietary "Super Omegaclad XL", when the budget allows, or type 310 stainless when I can piggyback onto an order through work from a local supplier. The first run to high temperature gets a black Oxide layer on the sheath (to get it to normal operating condition) and the second run (from cold) is the tuning run. I should note that I build my HT ovens to handle both Austenitizing and tempering temperatures and I am very careful to build/tune for minimum overshoot at tempering temperatures. If you are only looking at using yours for Austenitizing temperatures, this will be much less of an issue.
  5. For "us", it's not so easy. A workpiece might go into a forge clean and ground with a low emissivity. It'll get hot in the forge. If the forge atmosphere is sufficiently reducing, it might keep the low emissivity. As soon as the workpiece comes out into the air, an Oxide layer will start to form and the emissivity will increase as it does so. We'll decide the temperature is good and hit the workpiece with a hammer, causing the Oxide layer to detach and the emissivity to drop. The hot workpiece will immediately start to build a fresh Oxide layer and the emissivity will increase again. IR temperature measurement can certainly be useful, but it's probably not for most smiths unless they are able to significantly modify their process to accommodate IR temperature measurement.
  6. The biggest problem with them is getting an accurate emissivity value. In an industrial process where the surface condition at each point in the process is consistent, the emissivity can be set by measuring the workpiece temperature using a thermocouple (this may mean drilling a hole in a sacrificial sample workpiece and fitting a thermocouple in the hole, pointing the IR pyrometer at the workpiece and adjusting the emissivity until the IR and thermocouple readings are the same.
  7. A heat-treat kiln will usually do a better job of firing ceramics than a pottery kiln will do of Heat-Treating blades. The size thing is fairly obvious, but there are fairly fundamental differences between the tasks. For ceramics, the aim is to do heat-work: there is quite a large permissible band around the nominal temperature over which the temperature can fluctuate and, as long as the average temperature is held for the correct time, everything is fine. For Heat-Treating blades, we are much more concerned with precise temperatures and our acceptable band is much narrower. Typically, pottery kilns use Electromechanical relays/contactors and switch on an output cycle of 30 seconds or more. HT kilns tend to use SSRs (Solid State Relays), which can switch much faster, and can use an output cycle of a couple of seconds or so. At 50% power, a pottery kiln would be "ON" for 15 seconds and "Off" for 15 seconds, where a HT kiln would be "ON" for one second and "OFF" for one second. You'll get a saw-tooth effect: rising temperature during the "ON" and falling temperature during the "OFF". However, the pottery kiln will give big saw-teeth, where the HT kiln will give little saw-teeth.
  8. Not wishing to be that guy, but... Are you likely to be able to get a straight, burr-free cut with whatever you can put together? I'd hate to see anyone spend the time and effort of making a guillotine, only to find that it takes as long with the grinder to deburr as it does to cut the pieces with a slitting disk in the first place, and you then have to stack twisted pieces instead of flat ones. A 14" Carbide-toothed chop-saw would be my choice, but I am somewhat biased because I already have one. It's noisy, throws nasty sharp chips and is nobody's idea of a fun way to spend time, but it's quick, doesn't twist the metal and needs effectively zero cleanup on the cut ends (to be fair, the last mm or so usually breaks off one side and gets left on the other. Cleanup is usually one stroke with a file). An abrasive chop-saw would probably be ok, as said above. Also noisy, messy and the mess can't simply be cleaned up with a magnet like the carbide one. I've not used mine since getting the Carbide one, except for cutting fork-lift tines. A mitre saw with a Carbide-toothed-metal-cutting blade (or just a metal-cutting blade if you already have a mitre saw) might be a viable solution.
  9. I use the Amal injectors. There is nothing better that I've found, and I've been looking pretty hard for many years. Availability has been rather intermittent for a while now. When available, shipping to the US should be no problem, since Burlen, who took over the Amal IP, primarily supply carburettors and parts for old British cars and ship all over the world. It is worth noting that the Amal injectors use 55-degree BSP threads based on the Whitworth threadform, rather than the 60-degree NP threads used in North America: not a big problem, but something to be aware of. The gas connections are the ones you'll definitely need to adapt. BSP to NPT adaptors are available over here. Pneumatics specialist suppliers are where I tend to source them when I need them. The thread pitches on 1/2" and 3/4" are the same for BSP and NPT, so the burner tube should screw into the injector body ok with plenty of PTFE tape to cushion things and prevent damage to the thread in the injector body. Pressure isn't significant there and sealing isn't really an issue, but the thread continues on into the injector body to become the choke adjustment and you really don't want to damage it. I have built HT forges using a 20-22" length of 10" thinwall pipe, lined with a single layer of 1" blanket and with the ends being pressed-in disks of 1" board. They use a 1/2" Amal injector, which is fitted with a flame retention cup (made by welding a short section of 3/4" tube on the outside of the 1/2" pipe, then a 3" length of 1" pipe onto the 3/4" pipe), *very* approximately in accordance with the drawing on the Amal leaflet. The aim was to make a smaller version of the Don Fogg 55-gallon drum forge, more suited to making knives than swords. I spent some time trying to get it to work well and the breakthrough was effectively turning it over: the Don Fogg design uses the burner at the bottom one end and the exhaust/work port high up the other end. The Don Fogg burner port ID is bigger than the burner OD and the temperature control is achieved by adjusting the burner throughput relative to the airflow, due to convection, getting in around the outside of the burner. The forge atmosphere seems to be Oxidizing, though obviously much less so than air in an electric HT oven. For the smaller one, I made the burner port ID a pretty good fit for the burner OD and this allows it to run a rich/reducing atmosphere, adjusting the choke to get the desired temperature with excess fuel gas. It took a while to get it working well and the breakthrough came when I effectively turned it upside-down, with the burner at the top and the exhaust/work port at the bottom. Effectively the hot gases fill the top of the chamber and the exhaust comes out at the bottom exhaust/work port. The Australian Gameco burners look like they are a viable alternative to the Amal injectors. There is also a cast-and-tapped body available from China on Aliexpress, ebay, Amazon, etc, that looks like the Gameco one. It "should" allow you to make a good burner but I have no direct experience. The key to controllability for HT from a burner is a screw-adjustable choke IME. Compared to a sliding choke, the difference is difficult to comprehend until experienced. Holding Austenitizing temperature indefinitely to within a degree is simply not a problem.
  10. I think I'd file down the key to fit the slot depth, rather than file down the slot to fit the key depth. The key will come out and allow you to do the filing in a vice (vise) where it's easy. I've done enough filing to know I'm not good at it. Using a safe-edged file to increase the slot width should keep you from chewing up the bottom of the slot by running the safe-edge on it. If you need to deepen the slot, the safe-edge moves out of the equation and your opportunities for lousing things up increase considerably because you are working in 3 dimensions instead of 2: think flying under a bridge vs driving under a bridge. Once you've got the slot width right, using the loose key to check. it becomes a simple job to match the key depth.
  11. A file will do it. As said, you will want a hand file (less than about 4mm thick) with a safe edge and you'll need to work on both sides evenly to keep things aligned. Are there any grub screws in the pulley? If so, they'll make things a bit less critical in terms of slot accuracy. The bore provides the 90-degree alignment to the shaft, the key/keyway provides the driving force and the grub screw(s) prevent axial movement. In essence, the likelihood is that you'll end up with a slot with convex sides (unless you are good enough with a file that you'd have just done it and not bothered asking the question) and the critical part will be that the narrowest section of the slot is a good enough fit on the key to drive it without play. If you get it pretty close, but not perfect, you can usually use Loctite (or one of the cheaper alternatives) for additional hold. I'd probably go with Loctite 290 or an equivalent "wicking" product, as it will allow you to get everything correctly aligned dry and then apply a few drops that will wick into the gaps and set up. There are other products, like Loctite 620 or 638, that are stronger retainers, but you'll need to apply these, assemble and get things aligned correctly pretty quickly if you use them. Don't expect to be able to easily remove the wheel if you Loctite it.
  12. I promise that it's a single piece ball valve, aka one-piece ball valve. I have no idea whether the ones in the video are any good, but the video seems to show the difference between the different numbers of pieces. I think a 1000 PSI rating is pretty standard in stainless and the bodies usually seem to be cast. In Carbon steel, they are usually machined from Hex bar stock IME and may have higher pressure ratings. They are almost never automated and any slop is usually between the handle and the spindle, though I only have experience of manual valves that probably don't see a thousand open/close cycles per year. If yours is moved automatically, I'd expect loads of slop. https://www.new-line.com/valves-gauges-filters/needle-control-and-ball-valves-hydraulic/hydraulic-ball-valves-gate-valves-and-butterfly-valves/npt-1-pc-carbon-steel-ball-valve Seems to be the sort of thing you need, though it doesn't have the welded-on mounting bracket. I suspect that was done by the guy who built the hammer and you may need to do the same sort of thing yourself. I don't know how much distortion welding it might cause
  13. The industrial controllers generally allow the operator to adjust the setpoint in "vanilla" PID control (no ramp/soak), but anything beyond that often requires a God-level passcode that allows access to absolutely everything. I have had a knifemaker, with an oven I'd built, call me because he'd hit a wrong button whilst working through a menu and changed the thermocouple type from N to R or S. He was smart enough to recognize that he was getting temperatures in the tempering range when set for the Austenitizing range and called me. I'd made a similar mistake in the past, so quickly identified the problem. It was fixed during a 5-minute phone call. You will need to consider the level of technical expertise that can be expected from the end-user: on his own, I don't know whether he would ever have sorted the problem.
  14. I don't know how familiar you are with PID temperature control generally. I set up PID controllers in my day job: I'm certainly no expert, but I'm not completely clueless. There were a couple of things I didn't fully understand about the process when I built my first HT oven. With the benefit of hindsight, I would have appreciated being told. Pottery kilns tend to provide "Heat-Work" and the important parameter is the average temperature maintained over several hours. The workpieces usually have considerable thermal mass which helps damp out temperature fluctuations. HT ovens need to provide quite precise temperature control, with minimal deviation from setpoint. This is best provided by switching the element power through SSRs with a 2-5 second output cycle time. Most pottery kilns switch through relays/contactors on a 30 second-plus cycle time and this tends to give a saw-tooth temperature profile with relatively large teeth. With the reduced thermal mass of the work, and reduced damping, when Heat-Treating blades, this can give sizeable temperature swings, particularly near the cutting edge. Look for a DC-pulse output PID controller that will work with SSRs. If you are running on 220V in the US, I gather 2 SSRs are needed, so you'll want at least 40 mA of DC output current to switch 2 together. I assume you'll be using an industrial PID controller, primarily because they are much cheaper than dedicated kiln controllers. Maybe even a controller with ramp/soak programming capability? Part of the reason they are cheap is because they don't have the same level of segregation between "technical" and "non-technical" settings. The dedicated kiln controllers usually allow the operator to set up the ramp/soak segments of the program, but deny access to any of the things that could result in a failure of the kiln itself: things like thermocouple type, the actual P. I & D settings, etc. When you look for PID controllers, DOWNLOAD and thoroughly read the manual for each one you are considering. If you can't download the manual without jumping through hoops, look for another controller. If you can't understand at least most of the manual, look for another controller. The reason for this is as follows: you will need to set the controller up to do what it needs to do. You will therefore need to understand what the settings mean. If you find yourself out of your depth, you can either contact the suppliers tech support (in which case you will need to understand your process well enough to explain it clearly), or you can call for help, here or on another forum, with a link to the manual, hoping that someone familiar with PID controllers in general can look at the manual and offer advice.
  15. Hate to be that pedant, but it makes a difference when searching for info on something you are not already familiar with. It's "Curie point."
  16. Just a thought: what shape were the links? Some WI chain has an oval link in WI, with a crossbar bit in the middle that makes the oval into a pair of back-to-back "D"s. The crossbar is intended to take a compressive load and stop the link stretching to a longer oval under tension. The crossbar can be a completely different material because it only needs to resist compressive loading and sees no tension. I think the crossbars were probably Cast Iron in many cases. I know it's pretty unlikely, but if you've got a links like that and are trying to treat both materials the same way, it might explain things.
  17. I'd strongly suggest reading John Nicholson's thread on the jizzer in the stickys. It's simple and it works. I have built a few forges, including some intended specifically for HT, mainly because actually making knives seems like bloody hard work and forges for other folk to use are well within my comfort zone. It does mean that I probably spend more time on the forge than a typical smith would Bottom entry works very well full-size (55-gallon drum), but I found that top-entry, bottom exhaust worked best for me, with a chamber about 8" diameter and 18" long. To be honest, I'd faffed about with the Don Fogg design to try to reduce it to a sensible size for a hobby knifemaker in the UK. I was using a burner based on an Amal Atmospheric Injector. There's a discussion at: Unfortunately(?), the video of the forge in use (by my test-pilot, who made the video) is no longer available. It was pretty bloody tedious TBH: several minutes of video showing a forge running at 816 degC, with a few seconds where it hit 817 degC and a small adjustment was made to get it back to 816 degC. The forge was made from a length of 10" thinwall stainless pipe (20-24" long? it was scrap, picked up off a site I was working on), lined with 1" of Kaowool and rigidized/coated with a homebrew mix. The ends were lightly pressed-in disks of Ceramic Fibre Board, coated with the same homebrew mix. The thermocouple was a Mineral-Insulated Type K, 6mm diameter and 600mm long below a handle, Type 310 Stainless Steel sheath, curly cable with miniature plug. The pyrometer in this case was a TM902C bought off ebay for under 5 bucks, delvered. I had a number of these over several years and they all performed as well as any of the big-named brands at 20+ times the cost when checked against the calibrator at work. Then I bought ten of them and they were all identically inaccurate: excellent to 800 degC, then there was a progressive error from 800 degC upwards. I can't remember which way it went, but either they read 1290 degC at 1372 degC or they read 1372 degC at 1290 degC. Either way, I could no longer recommend them to anyone without a calibrator. There are several others available on ebay, etc, that also read in degF. Cost is a little higher though. Pics below show it during build and on test. The last pic is of the 1/2" Amal LV injector, which is the clever bit. Gas pressure is high at 40 PSI in the pic, but I would normally use 15-20 PSI (it's a very small jet). The flame temperature is adjusted by moving the knurled part to adjust the air gap between the knurled part and the 45-degree (-ish) taper on the main casting. Temperature control is very good indeed. I was aiming for 800 degC and 801 seemed close enough to me, though another couple of tiny adjustments could have got it to 800. Once adjusted, it'll hold to within a degreeC (2 degF) pretty much indefinitely. If you have access to pipe and fittings with BSP threads, I would definitely recommend the Amal Injector (with the BUTANE jet, even though you'll be running on Propane: the slightly smaller jet will give a slightly higher maximum flame temperature when that's what you want). The BSP threads probably make things difficult for the USA, where NPT threads are the norm. Burlen supply carburettors and parts for old British cars and will ship pretty much anywhere. http://amalcarb.co.uk/downloadfiles/amal/amal_gas_injectors.pdf
  18. I think it's (probably) a one-piece ball valve and I'm guessing at 1" for the 1/2" bore. These are always reduced-bore valves (AFAIK). It's worth doing a quick web search for the differences between 1,2 &3-piece valves so you've got an idea of what you are looking at. One-piece are pretty similar to 2-piece except that the bit that screws in and holds the seat in, has the same (parallel) pipe thread as the nominal size of the valve. I'd probably have a look for another one-piece that looks similar enough that you can pilfer the moving parts and transfer them into the body you have. There may be readily-available one-piece valves with mounting lugs on the bottom, but I don't recall seeing any. Most of the ones I've seen with mounting lugs have them on the top to take an actuator (though TBH, usually it would be a 3-piece valve for use with an actuator) or to allow mounting to the rear of a panel with the lever accessible from the front of the panel. Welding on a mounting lug, like on the original, is not something I'd particularly want to try. I suspect you are considerably better with a welder than I am, so YMMV.
  19. Sorry: fat-fingered. Please let us know what controller you use and how well it works for you? I for one would appreciate some insight into the alternatives to a VFD-and3-phase-motor approach.
  20. I'm not sure what the speed:torque characteristic curve of a shaded-pole motor looks like. I have only ever encountered shaded-pole motors in fixed-speed or variable-speed-fan applications and assumed they had a variable-torque characteristic that made them suitable for use with variable-speed fans, but probably not with constant-torque applications. I could, of course, be completely wrong. If yours does have a variable-torque characteristic, it is unlikely to work well at reduced speed on a grinder.
  21. I know it's probably not necessary, but I'll say it anyway: dry it out properly before you fire it up. Depending on the "refractory paste" you have used, this can take from "a bit longer than you'd expect" to near-geological time. The worst stuff I used was a readily-available "refractory cement", which I assume was actually clay-based. After a couple of days I thought it was dry enough, but it wasn't. On first firing, the surface dried out pretty much instantly, the moisture behind it flashed to steam and lifted the surface as bubbles. These broke and flakes fell off, then the process repeated. For the little forges (made from a 12" length of 8" pipe, 300 x 200mm, lined with a double layer of 1", 25mm, blanket), it was maybe ten-bucks-worth of useful learning experience: pretty good value in my book, and better value still if it helps anyone else. After that, I dried my little forges in the oven (on the lowest setting, like drying beef for jerky), risking the wrath of She-Who-Must-Be-Tolerated. Yours seems too big for that, so the incandescent light bulb inside, or similar, might be prudent. Hydraulic-setting refractories shouldn't be as bad, but can still crack quite badly if they are wet enough to flash off steam.
  22. Looks good Jaro. The 3-port diverter should work too, but I think the porting on these is usually set up to keep the combined area of the port openings fairly constant, which might limit the ability to throttle the flow and increase the backpressure. I don't really have much experience of those valves though.
  23. Sorry: fat fingers. The bleed-off-excess-air method will work fine though. It just "feels wrong" bleeding off the majority of the air that you are compressing without doing anything useful with it (which is why I mentioned the air curtain). I don't think I've seen a real gas forge actually using more than about 18 CFM of air, which, if my calculations are somewhere near the mark, corresponds to between 5.5 lb/hr and 8 lb/hr of Propane use, though I'm sure some of the guys with big power hammers exceed this.
  24. If it's 3-phase, and particularly if it can run on 230V 3-phase (where is Hobbit country by the way?), I'd be inclined to run it on a VFD. I tend to build my VFDs into IP65 enclosures with sockets on for things to plug into, so one VFD can run several different machines simply by plugging the appropriate one in. The controls are on a long cable, so I can move it to the appropriate machine. It doesn't make the VFD any cheaper, but it's a lot easier to justify the expense if it's something you can use for multiple purposes. The
  25. It's a side-channel blower (aka regenerative blower). I love them. I feel they are an order of magnitude "better" for forges than most conventional centrifugal blowers. However, the Pressure:flow curve is unusual. Without seeing the curve for the actual blower you have, my guess would be that the blower can do 85 CFM at zero pressure (about 144 M3/hr), 150 mbar (60" WC, about 2.2 PSI?) at zero flow and that it will follow a curve between these 2 endpoints. If you try to throttle it to get down to the flow you need, you will very probably (perhaps almost certainly, depending on the precise model/characteristics) cause it to overheat. The unusual characteristic of side-channel blowers is that the more they are throttled, the harder they work, the more power they use and the more heat they put into the air that they move. Rather than simply throttling the line to the burner, you are MUCH better off fitting a tee, such that you throttle the burner feed downstream of the tee AND bleed off surplus flow through a second valve off the tee. If you do this intelligently, you can arrange to bleed off the excess air to an air curtain across the mouth of the forge.
  • Create New...