Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Posts posted by timgunn

  1. Side channel blowers, also known as regenerative blowers are nice.


    They look different to most centrifugal blowers, having the inlet and outlet close together on the outside of the casing. It's worth googling them to understand how they work; if I try to explain, people will go all glassy-eyed and nod off.


    Small ones are usually built with the motor integral and the motor is pretty much a standard 2-pole industrial motor except for the drive-end casing which forms part of the blower casing. This means the motor runs at around 3500 RPM and the whole thing is nice and quiet.


    The performance curve is interesting and gives high pressure at low flow. Because of the way they work, the maximum pressure is much higher than a normal centrifugal blower of similar size and RPM can manage.


    Flow control on a fixed-speed side-channel blower is best done by bleeding off surplus flow, rather than by throttling the output. Throttling the output causes the pressure to rise and actually increases the power consumption. If throttling is unavoidable, it can be done on the suction side. I think throttling is probably good for a smelt, as the pressure rises if the tuyere becomes partially obstructed, so the reduction in airflow seems to be less than it would be if a "normal" centrifugal fan was used.


    Owen Bush has a nice little Elmo Reitschle unit which I've seen used on a couple of smelts. It seems to work very well. I think it's one of these:




    On 60 Hz mains, peak pressure at around 150 mbar is about 2.2 PSI and peak flow at 100 M3/hr is about 60 CFM.


    Anyone looking for a blower in the UK might be interested in this supplier:




    Although the maximum flow seems quite low compared to simple centrifugal types, the pressure/flow curve means that the side-channel units tend to have higher flows when used with typical tuyeres, which have smaller areas than the discharge ports of the blowers they are often used with. Larger units are obviously available as well, if more pressure/flow is needed.


    Some of the manufacturers that spring to mind are Nash/Elmo/Reitschle, Rotron, Gast, Becker. I have a UniJet 75, which I think is made by Cattani and marketed as a dental dry suction pump. It was listed on ebay as a vacuum pump. Among the other applications I've seen for small side-channel blowers are air knives for drying and spa blowers.


    New prices are generally high enough to rule them out for our applications. Secondhand prices are quite reasonable, at least in the UK, but they don't seem to come up all that often.

  2. Since we are talking about Forge Temp. I have a question.. Where can I find a good/not to much money Pyrometer ?? I got my forge going but I have no idea what the temp is.. I'm pretty sure I am going to have to redo it because the only way it works well is with the air at full blast and the propane cut way back.. Plus I don't like the pin valve I got because it's not working that well.. Thanks


    The pyrometer itself, or at least the readout, is the cheap and easy part. I use a TM902C bought off ebay; a search will find these at 5 or 6 bucks, delivered from China. They take a type K thermocouple input and have a miniature (flat-pin) socket. Too cheap not to have IMO. Slight downside for those that think in Fahrenheit is that they only read in Centigrade. Those I've had all read to 1365 degC (2489 degF), despite being marked 750 degC or 1300 degC.


    You will need to spend considerably more on the thermocouple.


    I use Mineral-Insulated thermocouples of 6mm (1/4") diameter and 600mm (24") length with a plastic handle and curly cable, ending in a miniature plug to suit the TM902C and my other (much more expensive but no more accurate) instruments.


    I also have a couple of longer 6mm MI type K transition joint thermocouples. These also work well and are off-the-shelf items, just needing the plug fitting. The sheaths on those I have are type 310 stainless steel, rated for 1100 degC (1202 degF), but they seem to work fine up to the 1365 degC maximum for the sort of timescales they see when used for checking/adjusting my forge.


    There are cheaper unsheathed thermocouples intended for fixed installations. They need a sheath if they are going to be in a forge for long and ceramic sheaths are available for this.These are good if you can put them in the right location to measure the temperature that you are interested in. Most of the forges I have had any dealings with have had temperature variations throughout the internal space. A long handheld probe will let you measure and understand the variations, helping to find a useful location for the fixed thermocouple if that is the way you intend to go. The ceramic sheaths are pretty good insulators and therefore slow the response to temperature changes. This means they tend to damp out fluctuations and can make a forge appear much more stable than it really is. It's one of the reasons I like to check the temperature stability and distribution before going that route.


    Most thermocouples are made to order. I try to specify "Grounded junctions" for the handled ones. This reduces the response time significantly when compared with "insulated junctions", but does have minor disadvantages in other areas.


    If you are in an industrial area, there is a good chance there will be a thermocouple supplier fairly near. If not, Omega are about the biggest name in temperature control and it's worth talking to them (www.omega.com) if you need to order with shipping. They have knowledgable technical support and I like their Omegaclad XL sheath material. Either way, I'd strongly recommend talking to the technical guru at whichever supplier you use and listening to their advise.

  3. I think you've missed the all-important 2 Alan.


    It's CO2 (Carbon Dioxide) that's heavier than air and will extiguish a candle. Its Relative Molecular Mass is 40. The RMM of air is about 29.


    CO (Carbon Monoxide) is a little lighter than air and is a flammable gas. Its RMM is 16.


    The differences between the RMMs of CO2, Air and CO are not all that great. As a result, little or no separation will normally occur over the sort of timescales that we are usually concerned with. This is partly due to the fact that, in our application, both CO and CO2 are produced along with lots of heat and even hot CO2 is lighter than cool air, so will tend to rise.


    Neither CO2 nor CO is healthy to have around: CO2 is usually present because the Oxygen in the air has been burned with a Carbon-containing fuel (or respired by a life-form; it is present in a concentration of 3-5% in the air you normally breathe out). The main risk from CO2 is usually considered to be the lack of remaining Oxygen in the air, causing asphyxiation.


    CO kills at much lower concentrations. It binds tightly to the Oxygen-carrying locations in the red blood cells, preventing them doing their normal job of carrying Oxygen to where it is needed. It's worth bearing this in mind; CO attaches to the Oxygen-carrying sites much more readily than it leaves them; going outside for a short breather during a forging session will not actually do much to reduce the blood CO level once it has been raised.

    • Like 1
  4. What's the Hones Buzzer burner originally from, Sam?


    It looks like Hones make several that could be useful, but the blurb seems to suggest they need secondary air in most applications (I'm not sure about the high-pressure version), so may run way too rich for a forge. I've been playing with Amal atmospheric injectors lately, which are designed for low pressure but work very well on high-pressure Propane when fitted with much smaller jets.


    I could be telling Granny how to suck eggs, but if yours doesn't play right off the bat, you may need a smaller gas jet.

  5. 3' long flames are not a problem where there is the space to accommodate them. Maybe a brick kiln or large pottery kiln. In general, we have small workpieces which set the minimum size of our forges, but any extra volume means extra surface area, which means heat loss, which means gas consumption, which means running cost.


    I'm not familiar with pottery kiln burners, but a lot of industrial Venturi burners only draw in part of their combustion air as primary air, the rest comes in as secondary air. The primary air part-burns the gas in the central cone, leaving the part-burned gas to mix with the atmosphere and complete the burn as the outer bushy part of the flame. In a forge, the only route for the air to get in is usually as the primary air. Any secondary mixing can only happen once the part-burned gases leave the forge and can mix with the atmosphere, where they burn as the dragons breath.


    If starting with a commercial Venturi burner, it is often necessary to substantially increase the proportion of primary air for use in a forge. The easiest way to do this is usually to fit a smaller gas jet.


    In addition, many commercial Venturi burners run on low gas pressures, which give low gas velocity through the Venturi throat and less "draw" than high-pressure jets would give. Again, "we" are likely to need smaller jets than would be typical for industrial use.


    I have been playing with "Amal" atmospheric injectors lately and have been getting reasonably good results with the 1" size used with 0.6mm MIG tips as gas jets running on Propane at 0-4 Bar (0-60 PSI). I've been able to maintain stable temperatures all the way from from 750 degC (1382 degF) to 1410 degC (2570 degF). I think there's a bit of scope for further tuning with proper jets, but I needed to get somewhere fairly close as a starting point.


    The Amal Gas Injectors leaflet can be found at http://www.amalcarb.co.uk/downloads.aspx


    I have been using a forge constructed from 20 IFBs to play with, giving a chamber 6" x 6" x 13 1/2". Running on a 19Kg (circa 50 Lb) cylinder, freezing has only become an issue when the cylinder is close to empty and the ambient temperature is close to freezing.


    The time to heat is longer than I'd like; forging temperature takes a minute or two, but welding temperature needs a good half hour. My IFBs are cheap grade 23s and are about twice as dense as the Morgan Thermal Ceramics JM23s I use for my electric HT ovens. I'm pretty sure the thermal mass is the reason for the slow starting. The JM 23s are 3 times the price here and I was only aiming for a quick-and-dirty test bed for the burner.


    I'll be building a round one next, lined with CF blanket, for better insulation and lower thermal mass. If I could get it over here, I'd try bubble Alumina for a flux-resistant lining on the bottom over the blanket, and aiming to keep it reasonably thin in order to minimize the thermal mass.


    With available materials being limited, I will probably have to go with dense castable instead of bubble Alumina.


    Whatever you end up going for, I'm convinced the difference between very good and just so-so is the burner. Get that less than excellent and you'll have a lousy forge, regardless of how well-engineered it may be structurally.

  6. The sightly unhelpful answer is that it depends on exactly what you are intending to use it for.


    As a display,it will show the temperature of an attached thermocouple (so long as you use a thermocouple it recognizes) in your choice of units. The type K thermocouple is the most common general-purpose thermocouple and it will recognize that.


    As far as using it for control goes, it has a DC pulse output to drive a Solid State Relay, which is the sort of output that is often used for electric HT ovens/furnaces/kilns.


    If you would want to switch a gas solenoid valve instead, for a temperature-controlled drum HT forge for example, you would probably do better with a relay output.


    72 x 72mm is a fairly unusual size. 48 x 48mm and 96 x 96 mm are more common, with 48 x 48 usually being significantly cheaper. I tend to stick to the more common sizes because it makes it easy to change them later on as your needs change. A basic controller, like the one in the link, will generally be all you need for HT of carbon steels, but many people feel a ramp/soak controller is worthwhile for stainless steels.


    I was not able to find a manual for that particular controller online within a few minutes of starting to google it. I would not buy it for this reason alone.


    To elaborate, if you run into difficulty, you can call for help on the interweb with a link to the manual, and there's a fair chance that someone who knows a bit about controllers in general can chime in and help, without necessarily having used the particular model you are using. Without the link to the manual, you are on your own.


    Trying to set up your first controller using a manual translated from Chinese using Babelfish or Google Translate can be a tad frustrating.


    For an information-only temperature display, you are probably better with a battery-powered handheld unit, freeing you from mains power. I use the TM902C, bought from ebay, but I think in degC. If you really need degF, you'll have to find something appropriate. Something to be aware of is that type K thermocouples can measure up to 1365 degC (2489 degF). A 3 1/2-digit display showing degF with a maximum count of 1999 will cut off the top of the useful range.


    The TM902C only works with type K thermocouples. I like this, as it means there's no chance of setting the wrong thermocouple type and getting wrong readings.


    Don't think that cheap means inaccurate. My TM902Cs agree with each other, with instruments costing over 20 times the price and with my calibration checker.


    The readout is only part of the story and the likelihood is that you'll need to spend more time and money sourcing the thermocouple itself than you will the display.


    I'd suggest a phone call to Omega to discuss your thermocouple and controller needs. Read up on things first, by all means. Make sure you have worked out (and made notes on) exactly what you are trying to do, then put your credit card out of reach, phone them and ask for their recommendation. See what they come back with and think about it before actually ordering.




    Price-wise, I find Omega expensive on hand-held instruments, but good on thermocouples. They are also good on ramp/soak controllers.


    Even for basic controllers, having a proper manual, written in English by someone who writes manuals for a living, can be worth paying a premium for. The support and advice also has a value, which you'll probably be able to gauge from the initial phone call.

  7. I'm assuming you are using a type K thermocouple and the D1S-VR controller with the SSR output to run an electric kiln/oven/furnace. I also assume you may want to treat Stainless steels at some point, so will want to be able to go above 1000 degC (1832 degF), and that you do not need alarms.


    I have not used the Sestos controller myself, so this isn't necessarily reliable. However, based on my experience of other controllers and the manual, such as it is, I think you will probably want all factory default settings except for the following:


    Ctl This is the output cycle time. Default is 4 seconds. I'd suggest 2 seconds.

    D1P Decimal point position. Default is 1, set to 0 (1 limits max display to 999.9)

    d1H Display High range. Set to 1300.


    Once you have the settings in, I recommend you run an autotune at a fairly typical tempering temperature first. Maybe 250 degC. I think the way to do this is:

    Power the elements with the setpoint at zero (how you do this will depend on your setup; it might be close the door, press a button, plug it in, or any one of a number of other ways), raise the setpoint to 250 degC then immediately press the AT button for 2 seconds. It should start flashing the AT display. Let it do its thing; I'd expect it to take several minutes.


    If successful, the autotune should complete and the controller should just keep working and hold the 250 degree setpoint.


    You can then adjust the setpoint as you see fit. Watch for the temperature overshooting the setpoint and see if you are happy with it.


    Other than that it's quick, the reason I suggest tuning at a low temperature is mainly due to overshoot; if you tune at high temperature, I find you tend to get overshoot at low temperature. If you tune at low temperature, you seem to get a slower approach to the setpoint when it is set at high temperature. You take a choice, but slow approach to Austenitizing doesn't seem to do as much harm as overshootin the tempering temperature.


    Of course, if you only use it for Austenitizing and temper with something else, you should tune at high temperature.

  8. For general workshop lighting, I like fluorescents with "daylight" tubes (6000 K or thereabouts). They are a little harsh for some people, so you may prefer "cool white". I always try to use High Frequency fittings (with electronic ballasts) to minimize the stroboscopic effect when rotating machinery or power tools are likely to be used in the room.


    Local lighting for rotating machinery really needs to be completely flicker-free. It tended to be Tungsten-Halogen in my day, but I'd imagine LEDs running on DC should get the job done.


    With low ceilings and therefore low lights, I tend to use weatherproof outdoor fittings in the workshop and garage. Not for the weatherproofing, but because they offer some mechanical protection to the tube. Little shards of glass and white powder raining down while I try to find the off switch for the grinder in the dark, isn't an experience I particularly wish to repeat.


    I also try to have two independent light sources in the room (a window counts if there is one), on separate circuits (no common fuse, breaker or RCD/GFCI) for the same reason.

  9. Dave, there are a few things to consider that may not be too obvious at first.


    Your motor rating plate shows a current of 11.8 Amps. However it is a 3-phase motor, so that is 11.8 Amps per Phase.


    I think (and I may be wrong) that the maths give 1.73 times the phase current needed for perfect conversion from single-phase. That would give 20.4 Amps of single-phase supply needed, plus a bit extra because the conversion will not be perfect. I'm guessing a bit, but I suspect it'll need around a 25A supply.


    3kW/4HP is about the top end of the power range for VFDs running from a 220V single-phase supply. Many manufacturers stop at 2.2kW/3HP and to go any bigger, a 3-phase supply into the drive is needed. A few years ago, I needed a 3kW/4HP single-phase 230V drive and the only one I could easily get in the UK was a Siemens G110.


    I've heard pretty good things about the HuanYang drives sold on ebay, but haven't had a play with one yet. There are plenty listed as 3kW/4HP, but all the photos accompaying the listings show the 2.2kW/3HP version.


    If it's for a grinder, you'll need to ensure grinding dust can't get anywhere near the innards of the drive. There are several options, from mounting the drive in a separate "clean" room, through filtered-ventilated enclosures, sealed enclosures and factory-built sealed drives. Dealing with the heat produced by the drive is the most awkward part of this. I assume from your location that you are not going to have to deal with tropical temperatures, so a drive in a sealed enclosure may do the job very well. The enclosure will probably need to be fairly big though.

  10. As I understand it, motors tend to be designed for mass manufacture. Most of the bits of an 1800 RPM 145tc motor end up the same as those on the 3600 RPM 145tc motor built on the same production line. Only the stator windings will be different, and these don't rotate.


    All the moving parts of the 1800 RPM motor will therefore have been designed for 3600 RPM.


    Electrically, the motor will have been designed for 60 Hz (possibly both 60 Hz and 50 Hz, if aimed at a world market) and optimized for that frequency. Moving away from the design frequency will gradually increase internal losses until the motor becomes too inefficient to be useful.


    Most motor and VFD manufacturers seem to recommend a speed range of 10 Hz to 100 Hz for a 4-pole motor. My limited personal experience suggests they've got it about right.


    For a motor rated for 1800 RPM at 60 Hz, 10-100 Hz gives a speed range of 300-3000 RPM.


    A 3600 RPM motor could probably be run over the same 10-100 Hz frequency range, giving 600-6000 RPM. Motor life would probably be shortened dramatically, as everything above 3600 RPM would exceed the design speed.


    If using a TEFC motor, you need to be aware that you'll need to derate below about 25 Hz, as the cooling airflow is reduced at low speed and the motor is more prone to overheating. In most cases, you just need to be aware of it; I don't imagine you'll be leaning hard on the platen for long periods at such low speeds, so the derate will probably be taken care of naturally. (If I'm wrong on this, please somebody let me know soon, before I build my grinder)


    I have run 4-pole motors at 120 Hz, 3600 RPM, but they feel as if the torque is down. I've tried running below 10 Hz on a basic V/F drive (Voltage and Frequency are varied together) and the motor felt "coggy". Using a slightly higher-specced "Sensorless Vector" drive kept the same motor feeling smooth at 3 Hz. SV is worth knowing about if you need very low speed, but I don't think it's needed on a grinder.

  11. I'd be inclined to build from scratch, rather than substantially modify an existing unit, unless it is very clear from the outset that modifying will be cheaper and/or easier.


    I've built 5 electric HT ovens from scratch, 4 of them for work up to 18" long and one for work up to 42" long. All will work vertically or horizontally. I think the long one would probably work best at about 45 degrees; vertically, it's a long way to have to lift out a full-length workpiece.


    For your questions:


    First: Maybe, but it's fun.


    Second: You can buy elements from many places. As a general rule, the fewer constraints you give the supplier, the cheaper they are. Simple coiled 16 ga Kanthal A1 works well. I get mine from an ebay supplier in Portland, Oregon (I'm in England). No complaints.


    Third: What type of firebricks? You want the soft Insulating Fire Bricks. 2300 degF grade are ideal, as they are cheapest, softest (for shaping) and least dense. They have the lowest heat capacity and are the best insulators. If you have hard bricks, look for the soft ones. On my first couple of ovens, I filed the element grooves with a length of allthread. From number 3, I wised up and used a router to cut the grooves.


    Fourth: If you use the controller output to drive a contactor (basically, a big relay) and switch the 220V through the contactor, it should work OK. You'll need advice from someone familiar with US wiring codes though.


    Fifth: It's better to ask someone who knows you well :D I've built HT ovens that work and haven't killed or maimed anyone, which suggests it's not difficult.


    Sixth: If it's at all useful or sellable, I'd try to keep the kiln as it is for as long as possible.


    Some of my efforts:


    18" oven with facility for vertical use with salt pot.



    42" oven. usable horizontally or vertically.


  12. I took a punt on a far eastern diamond indenter off ebay a year or so back. Mine shipped from Hong Kong. I tried it in a friends tester and was getting readings 3 RC points below the ones he got with the indenter that came with his tester, on the same test block.


    As far as the weights go, the weight is the important thing. If you can weigh the weights from another machine of the same model, and make another weight that weighs the same, it'll work. A bag of lead shot would get the job done so long as the weight is right. If you are only going to be testing blades and are not going to be using the HRB scale, only one weight is needed.

  13. I've only seen 2 of the 55-gallon drum HT forges. Both worked very well.


    The first one was all-manual and had a surprisingly small burner; about 1/2" from memory.


    The other used two cheap Chinese propane torches and a PID controller; the (very) small torch was the continuous pilot, with the larger torch switched on and off by the controller. I'm pretty sure the gas jet in the big torch was 0.35mm/.014", which should give a fair idea of the heat input needed; I normally fit a 0.6mm (0.024") MIG tip, which actually measures about 0.7mm (.028"), in a 1" burner. Half the gas jet diameter should mean half the burner diameter, so a 1/2" burner would seem about right.


    I'd expect a 3/4" burner to work OK. It'll probably be running at quite a low pressure.

  14. Yes, look for Sodium Borate Pentahydrate.


    The active part is the Sodium Tetraborate. Ideally, you'd use anhydrous Sodium Tetraborate, but this is not usually as easy to obtain as the hydrated forms.


    The hydrated forms are the Pentahydrate, which has 5 water molecules associated with each molecule of Sodium Tetraborate, and the Decahydrate, which has 10 water molecules associated with each molecule of Sodium Tetraborate.


    Given that what we want is the Borax part, and that the water is undesirable, the Pentahydrate is better for our purposes than the Decahydrate. Both

    hydrated forms are available at similar prices-per-kilo, so the one with least water in is therefore usually better value. If you need to factor in shipping costs as well, the Pentahydrate will definitely be better value.


    When it comes into contact with hot steel, the first thing the associated water molecules do is boil off, making your Borax all frothy in the process, and causing a significant proportion of it to dribble off the work.


    Anhydrous Borax (I can't be bothered typing "Sodium Tetraborate" each time) froths very little, if at all. The Pentahydrate definitely froths and the Decahydrate froths more.


    The spillage that inevitably results from the hydrated forms frothing may be significant to you, especially if you have a gas forge with a refractory lining made from metal oxides (I'd expect less of a problem with solid fuel than with gas, but this expectation is based entirely on ignorance and supposition; I have not used solid fuel and have no real idea).


    "Metal Oxides" covers ceramic fibre blanket, most Insulating Fire Bricks, many hard fire bricks and a good few of the refractory cements and castables. Alumina (Aluminium Oxide) and/or Zirconia (Zirconium Oxide) are used in most of the refractory compositions we are likely to encounter.


    It's worth bearing in mind that the main reason we use Borax, or any other flux, is to dissolve the metal oxides that would otherwise contaminate our weld, and allow them to flow away from the weld. It would seem unrealistic to expect anything to work well as a flux, yet not damage a metal-oxide-refractory-lined forge.


    I have reduced the water content of the Decahydrate by heating it for a few hours in a domestic oven. Weighing it before and after and doing the sums, it seemed to get it down to the Pentahydrate, but no further. YMMV.


    Getting it up to melting (about 750 degC), letting it cool and pounding the resultant glassy solid to powder, gives Anhydrous Borax. Once made, it needs keeping in an airtight container to stop it absorbing moisture from the air. It's a lot of hassle, but some people, who's opinions I respect, say it's worth it.

  15. You can buy ready-wound elements. Many pottery equipment suppliers will wind them to your spec, albeit at a price.


    I got the coils for my ovens from ebay seller jrider12 in Portland, Oregon. He seems to have dropped off the radar lately, but ebay seller pmtoolco, also in Portland, seems to offer similar coils.


    Prices, shipped USPS flat-rate to the UK, were less than I could buy the Kanthal wire for from anywhere else. The coils I have are wound from 16ga (AWG?) wire.


    There are a number of far-East ebay sellers offering kiln elements very cheaply now, but these appear to be of very thin wire. The detail is sparse on the ebay listings, but they look like they are cut from a long length of continuous coil. If so, the working life will be shorter due to the smaller diameter and they'll probably need the ends uncoiling to make "tails" for connection.


    I asked to have my elements wound for 13A (to suit 13A UK fused mains plugs) at 110V and used 2, wired in series, for each of my 18" ovens; one on each side.

  16. There are a few pretty close UK equivalents to the AISI 10xx steels.


    CS70 is equivalent to 1070, CS80 to 1080 and CS95 to 1095. EN9 is a pretty close match for 1055.


    As Jake and others have said, a good starter steel if you want to make things that will cut, is BS1407 silver steel. Downsides are that it's prone to grain growth and benefits from careful attention to normalizing. It only tends to come as precision ground rod, so is expensive by weight.


    On the other hand, it's widely available in a good range of sizes. You can buy the diameter you need and immediately make stuff, rather than spending much of your time reducing your stock (you've intimated this may be a consideration for you with your light hammer). It can be bought in 330mmm, 1M and 3M lengths; If you buy the 1M lengths, you'll get at least one, probably two, goes at making something from each length before you need to use the tongs; I find it hard enough to learn one thing at a time and having enough length to use it as a handle really helps as a beginner.


    Cromwell are a decent source for BS1407 silver steel.


    For patternwelding steels, Mick Maxen at your third link really knows his stuff.


    If you've not looked at the BritishBlades forum, it's worth checking it out. Quite a few of the members are on here too, but BB is obviously UK-biased and many of the local solutions you'll be looking for can be found there.


    Dave Budd's work might also be worth a search. I gather his background is archaeology and he uses iron-age equipment. He's not large, but manages to do an impressive amount of work with a small hammer.

  17. I just cut the grooves straight in and pinned the elements fairly closely using little U-shaped staples pushed into the bricks. The staples were bent from 1mm Kanthal wire.


    The outside diameter of the coiled elements was only just under 10mm, so they were a pretty good fit in the grooves. They did feel a bit springy as I was fitting them in the first one, and I wasn't sure the staples would be enough to hold them in use. Everything seemed to stress-relieve on the first firing and I've not worried about it since.


    The element seems to sit quite happily in the groove. Not the best photo and the oven's upside-down in it; that's the thermocouple on the "floor". There seems to be an Oxide layer on both the element and thermocouple, so it's been up to temperature. This one's an 18" oven.




    The odd angle is because the oven was standing on end at the time; the plan was to allow it to be used with a salt-pot, should I one day feel the urge.




    The 42" oven hadn't been fired when this photo was taken



  18. Salem


    I'm no expert, but I've built 5 electric HT ovens so far and I've found out a few things on the way.


    The soft bricks tend to suck all of the moisture out of whatever jointing compound you use, so it's quite easy to stick the bricks together with a thick layer of the compound, but it gets more difficult as the joint gets thinner. Because I want a close joint, I've tended to fit them dry. Where I've been concerned about gaps, I've just slathered the outside of the joint with watered-down fire cement (Satanite and AP Green seem to be unavailable here, so I can't comment on them). If you do the inside as well, you need to avoid getting the mix on the elements.


    Cutting depends a lot on the actual bricks you use. JM23 branded bricks are the easiest and offer the best insulation. They cut with a cheap hardpoint tenon saw, no problem. They are quite chalky in texture. I've also used low-iron GD23LW bricks; bubbly in texture and horrible to cut; I killed a tenon saw in 3 cuts and used a new blade from a mechanical hacksaw for the rest of the oven. By the end, there was no set left on the saw blade. They are quite a bit harder than the JM23s and a little denser, so not such good insulators. I've also found unbranded 23-grade bricks. Denser still and even poorer insulators, the texture is more like a sand/cement mix and ease-of-cutting is somewhere between the other two types. The JM23s are more expensive here in the UK (they are only made in the US and in Italy) but I feel they are worth the extra cost.


    For my first 2 ovens, I hacksawed either side of the element grooves, broke out the centre bit by tilting the saw blade, then filed the base of the grooves with a suitably-sized piece of studding (allthread).


    For the last 3, I used a router, together with a shopvac and respirator. Infinitely easier and works for all 3 types of (soft) brick I've come across. I'd expected it to kill router cutters fast and bought several cheap ones, but I've probably done over 20 ft of 10mm wide groove and I'm still on the first one.


    It is possible to use the same oven for Austenitizing stainless and tempering Carbon steels, but it's not as straightforward as it might initially appear.


    I have found that there is a tendency for the oven temperature itself to overshoot the setpoint at tempering temperatures, as a result of the very high heat input needed to get the stainless Austenitizing temperatures. Added to this is the tendency to heat the workpiece by radiation fron the elements themselves, rather than from the oven as a whole.


    I've got it all down to what I think is an acceptable standard of control by paying attention to detail: a fast-responding control thermocouple, a short cycle time on the controller output and a ramp/soak controller. I'll admit to being slightly obsessive about temperature control, and to having access to instrumentation that few others on here will have, as part of my day job.


    As a good alternative to going high-end on the control, one very good knifemaker I know buries his hardened blades in dry sand in a fish-kettle for tempering and sticks the whole kaboodle in his HT oven . It gives a huge thermal mass and effectively eliminates the overshoot issue, but obviously adds many hours to the tempering time.


    As Daniel says, the biggest problem with using the same oven for both really seems to be the cooldown time. It takes several hours to cool the oven from around 800 degC (1472 degF) to around 200 degC (392 degF) and it's a long time to leave fully-hardened blades. For swords, a spell at intermediate temperature in the kitchen oven isn't even an option.

  19. Sounds like you've got it covered, so I'll just sit back and wait. I'm told patience is a virtue.


    Any information you can provide on the control system you mentioned would be gratefully received. My last 4 builds have used the AutomationDirect Solo controllers (They seem to be the same as the Omega CN 7200, 7500, 7600, 7800 series controllers). It's a good controller, but doesn't seem very user-friendly when it comes to the profile programming, especially for non-geeks. I need to get something pretty soon for my next build and if there's something that may be better available, I'd like to give it a try.


    My last build was a 42" sword-length oven with around 6 kW of input power. With hindsight, I could have gone a lot less. I basically just doubled up on the elements I had used on my earlier, 18", builds. This was partly because I only wanted to make one change at a time, and partly because I already had the elements, I'm basically lazy and, faced with a choice of 3 kW, 6 kW or winding fresh elements, it seemed like a no-brainer at the time. I really should try it on just one set of elements and see how hot it'll get. I've successfully taken the 18", 3 kW, version to 1176 degC/2150 degF, which was the highest HT temperature recommendation I could find at the time (S30V treated for maximum wear resistance) and it was still rising quite fast as it approached that temperature.


    I find lots of power tends to make accurate control at low temperatures a lot harder.


    Still on my to-do list is a tempering oven to go with the sword oven: a 4' long piece of 4" pipe, a fairly low-powered catering rod element (I'm guessing 1000W would do it) clamped to the outside of the pipe, along with a thermocouple and the whole thing insulated with Rockwool slab. The plan is to run it from the control box that runs the sword oven, once the blades are quenched and the sword oven is finished with.

  20. Looks the business, Dan.


    Nice use of the TIG shrouds. I just might have to steal that idea for a future build.


    Any chance, as you write it all up, you can major on the control/electrical side of things?


    Most of the oven/furnace write-ups I've seen have tended to gloss over all the boring-looking technical bits. Whereas most of the guys I've spoken to, who are considering building one, are happy enough about the physical build, but rather nervous about the electrics and control.


    Hopefully, all the questions will be answered anyway during the WIP, but there are one or two that spring to mind already:


    That looks like a lot of element length you have there. What sort of wattage are you going to be running, and what sort of temperature range are you intending to run it in? Carbon steels only, or stainless as well? Are you looking to use it for tempering too, or does the cool-down time rule that out?


    Overnight annealing suggests a ramp-and-soak controller. Have you found one that's affordable and easy to use? If so, please share.

  21. If you really need to contiuously measure welding temperature, you'll almost certainly need to go for a Platinum-based thermocouple with a ceramic sheath.


    The Omega thermocouple John uses is really about the upper limit for base-metal thermocouples. Where I feel it is most useful for the novice forge-welder is in establishing whether the (gas) forge is hot enough to weld high(-ish?) carbon steels.


    My feeling is that it's unlikely to help the experienced guys much, if at all. I originally recommended it because I feel that it offers the most bang for the buck. Over here, the additional cost for the Omegaclad XL sheath versus a 310 Stainless sheath from another supplier, works out at around 10 bucks or so. Once you factor in shipping and the readout, it adds less than 20% to the total system cost to get the extra 200 degC of usable range and this is enough to just get into the welding range.


    I come at this as a bit of a geek with an interest in tools and a background in gas burners and process control, rather than as a maker of knives. I had no idea at all what sort of temperature to aim for when building a propane forge a couple of years ago. I struggled to find objective information online, so I went to a hammerin, watched lots of folk making Damascus for two days, then measured the temperature of the forge using a type S (Platinum) thermocouple. It was doing 1280-1310 degC (2336-2390 degF).


    As is the way of such things, there followed a brief discussion on achievable temperatures, some burner adjustment and a melted forge lining. Peak temperature measured was 1470 degC (2678 degF).


    The thing I had wanted to achieve by the exercise, was a realistic value for a welding forge temperature for knifemaking. Several of the guys using the forge were very good smiths. A couple of them had guessed the temperature within 30 degC but one was 30 degC high and the other 30 degC low. Most were happy to admit they had no idea what the temperature was. Nobody had said it was too hot. Nobody had said it wasn't hot enough. If it had needed to be hotter, it could readily have been adjusted to make it hotter.


    Based on that experience, I'd say that for welding what the type K can do is tell you whether the forge temperature is in the right ball-park and give some indication of the temperature distribution. It'll be close to its limit though.


    Based on my limited understanding of metallurgy and welding, a base-metal thermocouple isn't going to be much use for measuring the higher welding temperatures needed for mild steel or wrought iron.


    My feeling is that continuous measurement in a welding forge is only likely to be helpful if you are sure the temperature distribution is even and that the thermocouple itself is measuring at a point that is at the same temperature as the working area, since the thermocouple is fragile and can't encroach on the working area itself. Effectively, it's only useful in a really well-designed and well-executed forge. I've not seen a forge that I feel is sufficiently well-designed and -executed to convince me that the continuous measurement would be useful, but I've seen plenty where a check measurement in the working zone would certainly be useful to some people.


    For HT, leaving the thermocouple in will not be a problem. For forging, it may not be a problem either, but a lot depends on your setup and how you work. For welding, I'd say you are better off taking it out. Personally, I'd take it out for forging too.

  22. There are a lot of things involved in putting together a PID controlled gas forge that works reasonably well. Few of them are immediately obvious to most people. Not all are immediately obvious to me, and I'm a seriously geeky type who deals with PID-controlled gas equipment for a living.


    The Don Fogg oil drum HT furnace lends itself well to simple PID control, but is about the only design I've seen that does (without going to analog control, which is getting rather specialized ).


    For the Don Fogg setup I've played with, a separate pilot burner was used and ran continuously. The main burner was switched on and off by the PID controller. The pilot burner was a small gas torch and the main burner was a larger gas torch. Both were teed off the same regulator, with the solenoid valve in the line to the big burner. A baffle plate was found to be needed in front of the burners to achieve reliable lighting of the main burner. Both burners were cheapies of far eastern origin. The PID controller was a cheap one off ebay, but had a relay output to switch the solenoid valve, rather than a DC output to drive an SSR.


    It wouldn't do anything that the manual version couldn't do, but was quicker, and needed much less fiddling, to get to temperature.


    The temperature of most gas forges is varied primarily by adjusting the gas:air mixture ratio to vary the flame temperature. The gasfeed can be fixed and the airfeed adjusted, or the airfeed can be fixed and the gasfeed adjusted. Either way, automating this system needs a means of varying flow smoothly across a range. This is generally more complex and expensive than simple time-proportioning (on/off) control.

  • Create New...