Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Posts posted by timgunn

  1. First thing to say is that I've built a few forges and they seem to work ok. I've spent most of my working life dealing with gas burners and temperature control and I think I understand it fairly well. However, I'm no smith and there are plenty of guys out there with more experience than me.


    That said, my take on it is this:


    There are a number of factors that determine how much gas you really need and those rules of thumb are just rules of thumb.


    Most of the stuff that is out there on BTUs and forges is aimed more at blacksmithing than at bladesmithing. Welding temperatures for mild steel, and particularly for wrought iron, are quite a lot higher than those for typical bladesmithing steels.


    2200 degF is around 1200 degC and it is certainly possible to weld at that temperature. However, when I've measured the temperature of forges being used for making Damascus, by smiths who are good at making Damascus, they have been around 1300 degC (2372 degF). Even then, it is high-Carbon steel that is being welded.


    Blades also tend to be able to fit through small openings, so it's quite easy to build a bladesmithing forge with small openings and a large volume. A typical blacksmithing forge will have a much larger front opening and often a large rear opening.


    In my experience, the area of the openings has much more effect on the heat input needed to maintain temperature than the forge volume does.


    For the (blacksmithing) rules of thumb, there is probably quite a strong correlation between forge volume and open area when building "normal" forges.


    The BTU rating of the burner is based on the heat output assuming all the gas is fully burnt. In reality, it seems quite usual to run with a rich (reducing) forge atmosphere to control scaling and for there to be a significant amount of Dragons Breath. The DB represents wasted heat: BTUs that are not actually released in the forge itself.


    When using MIG tips for gas jets, it should be noted that the tips are sized for a nominal wire size and that I have found the bore of the tip is typically .006-.008" bigger than the nominal tip size (Hybridburners give .007-.009" near the bottom of the page at http://www.hybridburners.com/BTU-charts.html#oneerchart) A .045" nominal mig tip would therefore be .051-.054" in the BTU calculator. The nicely-tapered lead-in probably also raises the discharge coefficient of the jet to at least 0.75 and possibly close to 0.8. This will also affect the BTU calculator results.


    When I was playing with jet sizes for burners, I built a forge using cheap IFB and with a 6" x 6" x 13.3" chamber. With a single 3" x 3" opening, I got a temperature of 1545 degC (2813 degF) on a 0.6mm (.023") MIG tip measuring between .030" and .031" diameter and 4 bar (59 PSI) gas pressure. BTU calculator gives just under 104,000 BTU/hr (.031", 59 PSI, 0.8 discharge coefficient) which works out to 214 BTU/cu.in (less than half the rule of thumb figure) even at over 2800 degF.






    And the effect on the Thermal Ceramics JM23 IFB I'd used to restrict the opening:




    The cheap IFBs held up to the temperature pretty well, despite only being rated to 2300 degF, but are pretty poor insulators compared to the JM23s.

  2. I'm guessing the burner throat is (maybe) too big, or you are not putting enough fuel/air through it.


    Does the following make sense and/or sound like it could explain things?


    In the burner tube, the mixture needs to be moving towards the forge faster than the flame-front travels through the mixture in the opposite direction.


    The speed of the flame-front is not a constant. It depends on the air:fuel ratio, the pressure and the temperature.


    As the forge heats up, the temperature of the burner tube tends to rise a little and the speed of the flame-front increases.


    At some point, the flame-front moves slightly faster through the mixture than the mixture is moving along the tube.


    Often, there is a pop (at least with a NA burner: I don't get to play with blown burners) as the flame accelerates down the tube to the gas jet, where the flame goes out, followed by a brief pause as the burnt gases clear the tube, then another burnback when fresh mixture hits the hot chamber and everything repeats. If the operator is quick, turning up the pressure often sorts things. If the operator is not quick, each successive flashback heats the burner tube a little and increases the flame-front speed until the flame stabilizes at the back end of the burner tube close to the gas jet.


    If that sounds like your problem, your best approach is to increase both gas- and air-pressure to increase the mixture speed. If you have found your sweet-spot and would be wasting gas by doing this, you'll need to reduce the burner throat diameter so that when you increase the pressure to increase the mixture speed, you get back to the gas- and air-flow that you have now.

  3. Kaowool looks fine.


    Burner looks like it should be OK.


    Meter looks fine. Personally, I search for TM902C on ebay, filter by shipped price and buy the cheapest. The TM902C reads in degC only but that's not a problem when you are aiming for a target temperature you've decided on beforehand. I've had 20 or so and they've all worked to 1368 degC (just under 2500 degF), regardless of the range marked on the front. I've had them on the calibrator at work and accuracy has been just as good as brand-name instruments at 20 times the price. At around 5 bucks delivered, they are a bargain.


    The probe is really the awkward bit. It will not usually be cheap.


    The one in your link has a 6" probe and 3' of cable. If you fit it through the casing, you'll be able to measure the temperature about 4" from the wall wherever you fit it. Being a type K, the junction will measure to 2500 degF. The rest of the thermocouple assembly will probably not take that temperature though.


    Over here, I'd use something along the lines of the 405-041 with a miniature plug 724-102


    http://www.tcdirect.co.uk/Default.aspx?level=2&department_id=190/2and http://www.tcdirect.co.uk/Default.aspx?level=2&department_id=280/1


    I appreciate this doesn't help you guys in the states much, but it should give the general idea. I've been struggling to find a US equivalent of TCDirect with online pricing, though I suspect my UK-centric search settings don't help


    It's well worth reading up on thermocouples, but it seems pretty difficult to find a good basic thermocouple primer anywhere. Most seem to get quite technical, quite quickly. Probably the best I've found is the Labfacility Handbook:




    About the biggest name in temperature control worldwide is Omega: http://www.omega.com/temperature/


    They can supply pretty much everything you need. They are not the cheapest (IME) but have genuinely outstanding technical support that more than justifies the premium (again, IME). If you need good advice, they are well worth a phone call. They have an online thermocouple configurator that will give prices, but there's not an online pricelist that I've found.

  4. I've not tried 3/4" kaowool (or even seen it). If it's the high-density stuff (128 kg/m3, 8 lb/cu ft) I think it "should" work, but will give you a bit less headroom for getting the rest right than the thicker stuff. I'd rather expect the lower densities to give marginal insulation at 3/4", though if you get everything else pretty much perfect, it should still work OK.


    I'd recommend a long type K mineral insulated thermocouple with a 1/4" sheath in type 310 stainless, Inconel or Super Omegaclad XL. That way you can put the tip anywhere you want it and can move it about to profile the temperature distribution. Once you are happy with the distribution, you can put the probe in a representative position and use it for the adjustments. I find 6mm diameter (1/4") is stiff enough to use horizontally at 1000mm (40") long at HT temperatures. Vertically, there should be no problem at all except for keeping the handle and the hand holding it cool. getting a 6' transition junction thermocouple and bending it is probably the best solution.

  5. Would stainless wire work to hold the blanket up?

    Depends on the stainless and the the temperature you are working, but tentatively yes.


    What kills stainless in heat cycling is loss of the Oxide layer. When it gets hot, an oxide layer forms. When it cools, the mismatch between the expansion coefficients of the Oxide layer and the underlying Stainless Steel causes the Oxide to spall off. After a number of cycles, depending on thickness, there's nothing left. Different stainless steels will hang onto the Oxide layer through different temperatures. 316 keeps the Oxide layer intact through around 850 degC/1560 degF, so is good enough for most Carbon steel HT.


    310 stainless has higher (25%) Chromium for a stronger Oxide layer and higher (20%) Nickel, which brings the thermal expansion coefficient closer to that of the Oxide, and will keep the Oxide layer intact through cycles up to about 1100 degC/2000 degF.


    Finding 310 is usually difficult, so Nichrome or Kanthal are often used because they can be found in small quantities on ebay.


    316 in conjunction with rigidizer works quite well for blanket retention: by the time the stainless wire fails, the shell is rigid.

  6. There are many more things going on in a gas forge than there are in a gas cooker and it is very dangerous to assume that a forge will be as safe as a cooker simply because both burn gas.


    The "well-tuned" part is where most of the complexity comes in when referring to a gas forge.


    For heating things with flames, the flame needs to be hotter than the thing that it is heating because heat transfers from areas of high temperature to areas of low temperature.


    For cooking, the thing being heated (the food) tends not to need to reach a temperature much above about a couple of hundred degrees Centigrade; around 400 degrees Fahrenheit. None of the common fuel gases will sustain a flame at temperatures this low, so it's fair to assume that if there is a flame present, it is hot enough to cook with.


    The steeper the temperature gradient, the faster the heat transfer, so higher flame temperatures are better for efficiency. The highest flame temperature for each of the common fuel gases tends to be at around the stoichiometric air:fuel ratio. This is the air:fuel ratio at which all the fuel and all the Oxygen are consumed with neither fuel nor Oxygen remaining.


    Moving either side of the stoichiometric ratio, the flame temperature reduces because there is either excess air or excess gas to be heated to flame temperature and this takes some of the energy released during combustion.


    If we have excess gas, we don't get complete combustion and produce some Carbon Monoxide, a toxic gas. If we have a small amount of excess air, the flame temperature is still very hot and there is enough energy to cause some of the excess Oxygen to combine with Nitrogen and form Oxides of Nitrogen, often abbreviated to NOX, which are toxic gases.


    The OP lists auto racing and modifying among his interests, so he is probably somewhat familiar with some of this. An automotive exhaust Lambda sensor usually supplies a signal to the engine management system, which it uses to adjust the fuelling and maintain a Lambda value of 1. Lambda is just the scientific notation for the stoichiometric ratio. An exhaust reading below 1 is a rich mixture, above 1 is a lean mixture.


    If the mixture is very lean, the flame temperature is much lower. There is not enough energy to react the excess Oxygen with Nitrogen and little or no NOX is formed. Carbon Monoxide will tend not be formed either.


    The gas cooker will therefore be set up to burn very lean. That way there is minimal risk of causing large numbers of American mothers to keel over while cooking the thanksgiving turkey. The reduced heat transfer efficiency is generally felt to be an acceptable price to pay for this.


    Wikipedia lists the CO concentration near properly-adjusted gas cookers as 5-15 PPM whilst OSHA place a long-term exposure limit of 50 PPM of Carbon Monoxide on workplaces.


    We tend to need higher flame temperatures for forging than a very lean flame can provide. For welding, we tend to need temperatures that can only be reached with mixtures quite close to stoichiometric.


    When we have Oxygen present and high temperatures at the surface of our steel, we get scaling and decarburization. We tend to consider this undesirable, so usually run our forges with excess gas instead.


    This means that we are intentionally operating our forges to produce Carbon Monoxide.


    Most of the Carbon Monoxide produced in the forge leaves the mouth of the forge and mixes with air, where it burns with Oxygen to produce Carbon Dioxide. This burning is often referred to as the dragons breath. Because the mixing is uncontrolled, parts of the mixture will cool before combustion is complete, so some of the Carbon Monoxide will not be burnt and will remain in the air.


    I am less familiar with solid fuel combustion than I am with gaseous fuel combustion, but production of some level of Carbon Monoxide is pretty much inevitable when using a coal, coke or charcoal forge, or even when burning wood. Wikipedia lists the CO content of the exhaust from a typical home wood fire as 7000 PPM.


    The received wisdom is to put forges outside.


    If US workplace safety legislation is anything at all like ours in the UK, there will be an obligation on your employer to assess the risks and control them to an acceptable level: killing the workers is generally frowned on. I assume the OSHA workplace exposure limits will need to be met for all substances likely to be present, including Carbon Dioxide and Oxides of Nitrogen.

  7. I have a 2' diameter factory-motorized natural wheel, running in a water bath. I think it probably dates from the 1950s and it turns at a heady 42RPM thanks to a reduction gear on the motor and a reduction v-belt drive. That's so close to a surface speed of 3 MPH that I'm sure it was the target speed when they specified the drive.


    It works a lot better wet than it does dry, but it is very, very slow going. It always seemed pretty good for recovering badly-abused cutting edges but was no good at all for removing stock.


    At that speed, the water doesn't get thrown off. However, it does get stripped off by contact with the workpiece and then runs everywhere. In a warm climate, it would probably be fine. Here in Lancashire, it's not pleasant.


    I don't have the 3-phase, 400V supply needed to run it any more, so it's not been used for 3 or 4 years. I can't honestly say I miss it.

  8. Unless you are pretty good at scrounging, electric is probably out of your budget.


    A Don Fogg style 55 gallon drum HT forge is cheap, about as simple as it gets and sounds very much like Jesus' suggestion.


    Basically, it is a 55-gallon drum, mounted horizontally and lined with a single layer of 1" Kaowool blanket, with a small burner mounted into the original 2" bung hole low down at one end. High up at the other end, a fairly small (maybe 2" wide and 3-4" high) opening is cut for workpiece access and exhaust gas egress. In the two I've seen, the workpiece is hung in wire loops which are just extra to the wire holding the blanket in place.


    The burner is not a tight fit in the bung hole: the temperature adjustment is basically done by putting a thermocouple in and adjusting the gas until the temperature is where you want it to be. It's the balance between the amount of burner flame and the amount of extra air coming in around the burner that gives the temperature control, so it does not need a fancy forge burner, a cheap torch will do the job.


    I'm pretty certain the large volume is a big factor in getting even heating (radiative heat transfer follows an inverse-square law). Temperature can be held within surprisingly tight limits once correctly adjusted.


    The design is very elegant: just about any attempt to improve it seems very likely to make it perform worse: more insulation, ITC100 and multiple burners are just three that I've heard suggested by folk with an "if it ain't broke, keep fixing it until it is" outlook.


    If a 55 gallon drum is too short, I'm pretty sure any other suitably-sized cylinder should work, even a couple of 55-gallon drums joined together. I think the diameter drives the airflow by providing the effective chimney height for the draft, so a long, skinny cylinder might not work as well. I think standing the drum on end would probably not improve things at all either, but I could be wrong.


    I've had a play with a PID-controlled version as well. It worked very well but needed a pilot burner plus the main burner and it took a fair amount of faffing about to get reliable ignition of the main burner. It did not seem to perform any better or worse than the manually-adjusted version. Given the potential consequences of filling such a large volume with gas/air mixture and then lighting it, I'm inclined to feel that PID control does not save enough adjustment time to be worth the risk and effort. Obviously YMMV.

  9. Can anyone advise what the best fire proof wadding to use in my gas forge. I have some that came with the forge but it is melting away quickly where the heat hits it. If not wadding is there any other way of protecting the inside of the forge. Also where can I buy some



    Whereabouts are you? As Alan says, some stuff is harder to find in the UK. Other stuff is easy and, on a little island, pretty close.


    What is the forge? Pics or a good description would help. Pics are usually better: since you are asking the question, there's a fair chance you are new enough to this that you don't yet know the right questions to ask.


    What do you use it for? The temperature makes a difference and if it involves flux, the answer will be very different to if it doesn't involve flux.

  10. I've not often seen it with torches, but it's a fairly common effect with Naturally Aspirated burners that draw primary air.


    Normally, the speed of the mixture moving away from the nozzle is higher than the speed the flame can move through the mixture all the way to the burner exit. After the exit, the mixture can spread out over a larger area, slowing down as it does so, to give a narrow band somewhere near the burner exit where the flame speed matches the mixture speed and you get a stable flame.


    Rather unhelpfully, the flame speed changes with the air:fuel ratio, temperature and pressure. Things can be going fine until the forge reaches a particular temperature, then it starts misbehaving.


    What happens with my Venturi burners and is probably happening with your torch is this:


    The Flame-front radiates heat, some of which goes forward and is absorbed by the unburned mixture. The flame-front speeds up slightly, still radiating heat and generating a bit of a pressure-wave in front as it speeds up. The hotter, higher-pressure mixture burns faster, the flame front accelerates and it keeps accelerating until it runs out of mixture when it gets so close to the gas jet that the gas and air are not mixed. The flame goes out for lack of anything to burn. The gas and air continue to flow, mixing as normal, until the mixture hits the hot chamber, ignites and repeats the process.


    The usual way to deal with this is to increase the mixture speed by turning up the gas pressure. What do you have by way of a regulator?


    With a choked Venturi burner, there is also adjustment of the air:fuel ratio available using the choke. The fastest flame speed is generally at the stoichiometric mixture (effectively a neutral flame) and adjusting to give a rich (reducing) flame (closing down the choke) will usually slow the flame speed to some degree.


    If your gas jet is restricted at all, your mixture is likely to be lean and this may be the cause of the problem.


    The first 2 things to check, in whichever order is easiest, are therefore crud in the gas jet and operating pressure.

  11. My thoughts on a dedicated tempering setup are as follows:


    A length of reasonably thick-walled box section, big enough to take the anticipated work (say 4" x 4" x 1/4"+ wall).

    A length of reasonably thick steel flat on the bottom of it (say 4" x 3/4").

    This needs to have a hole drilled in it to take the thermocouple, but it can be drilled from the side so it's not a big deal.

    A rod-type electrical heating element on the bottom of that. Diameter seems to be around 8mm or 5/16" for the ones I've looked at.

    Two strips of steel flat, whose thickness is the same as the diameter of the rod element, on either side of the element (say 5/16" x 1 3/4").

    Another, wide, strip of steel flat to go on the bottom (say 4" x 1/2"). Drill through and bolt everything together tightly, maybe with heat-sink compound on the joint surfaces.

    Fit Mineral-Insulated thermocouple and wire up rod element.

    Insulate with around 4" or so of Rockwool (mineral wool) slab.

    Control with a PID controller.


    The idea is that the insulation is good enough and the mass of steel is great enough to give even temperature

    distribution inside the box-section, albeit slow to get there.

    The thermocouple is between the element and the workpiece, so will prevent overheating of the chamber.

    The element does not have line-of-sight to the workpiece, eliminating any radiated heat issues.

    The heat input is even along the length of the oven.

    The elements are available pretty much any length and wattage from ebay, and are fairly reasonably priced.

    My best guess at required Wattage would be around 500W per foot of chamber length.

    A very basic PID controller will run this setup and the low power/large thermal mass/good insulation should

    make it very stable.

    A downside to this stability is likely to be a very slow autotune.


    The main reason I've not actually built one yet is that it'll be too heavy to move and I don't have the space for it as a permanent fixture.

    I wanted to build one to go with a 42" HT oven I built a while back, but the HT oven needed wheels to move it,

    and there was no way I was going to be able to manhandle a deliberately heavy tempering oven.

  12. OK. A couple of things jump out.


    First is the Propane cylinder. It's a Patio Gas cylinder inended for patio heaters and, as far as I can tell, only takes a 37mbar clip-on regulator.


    Second is the amount of Carbon on and in the forge. This shows a very rich burn: you are not getting enough air in to burn the gas. You may not be getting enough gas in to make temperature either, but the lack of air is the first thing to address.


    You don't show the burners you have tried. I'm guessing they were atmospheric burners. If you are going to use the Patio gas, I'd expect you to have to go to a blown burner.


    If you are intending to use an unblown burner, you'll really want to start with a "normal" Propane cylinder with the female threaded fitting, and a high-pressure regulator giving 0-4 bar, 0-60 PSI.


    If you've been building burners to "known good" plans, the cylinder and regulator change may be enough to fix your problems.


    Otherwise, can you provide some more information about the burners, preferably with pictures?

  13. Borax is very viscous at Carbon Steel HT temperatures. I tried it once in a mini salt pot and it was just too thick to use for O1. I tried low-sodium salt next (Class 5 HT salt, but from my local supermarket) and it was much better.


    Borax might be usable at higher temperatures, but I'm not sure. In the forge, it seems to go properly runny at about welding temperature and I think that might be too hot.


    It does have a tendency to dissolve refractories too, which can be something of a nuisance.


    I'm sure there are other good reasons Borax is not used, but the viscosity was enough for me.

  14. The C706 and C715 actually look to be CuproNickels; Copper/Nickel alloys with no Zinc.


    As I understand it, the Nickel Silvers are Copper/Nickel/Zinc alloys. I gather the 18% Nickel C752 is the one to use for the "silver" colour. The 10% Nickel stuff tends to look a bit brassy, with a definite yellow tinge.

  15. I cannot be sure; do you have a link to the one you can get?


    The ones we use are a 3/8 brass valve with coil to suit the supply and an asymmetric timer that fits between the DIN connector and the coil. The timer is universal supply. Price here is 30-35 GBP, so I'm guessing it's basically the same thing.


    They look a lot like this: http://www.cccme.org.cn/products/detail-8063679.aspx

  16. We use them at work a lot; usually on 3 HP piston compressors with around 40-gallon tanks running 24/365 to provide control air on remote sites. The motor/pump duty cycles are usually in the 10-25% range.


    The drains are unbranded, made in the Far East and are pretty reliable as a rule.


    The most common failure mode we see is little tiny balls of steel from when the tank was welded, getting trapped in the valve and stopping it from closing properly. It tends to be more of a problem with new compressors.


    Timer failures and coil failures are much rarer. We see perhaps one every year or 2, with somewhere between 40 and 60 installed; maybe 40-120 years MTBF.

  17. Using a Propane forge indoors is a fairly high-risk activity. Carbon Monoxide is produced in quantity and is a very poisonous gas.


    It can be managed safely, but there are a few things to consider and they may not be obvious until you've got some experience and done some research.


    Stick with plan A for now.


    Whatever forge you build, you'll find that you reach some limit at some point and need to build another.

  18. Are you getting Dragons breath?


    There seem to be 2 possibilities:


    First is that you are not getting enough gas in. This condition is likely to give the same effect as the forge just being run at lower pressure than you want and whether or not there is any Dragons breath depends on the normal burner mixture.


    Second possibility is that you are not getting enough air in. If there is a restriction to the air supply, there will be a lot of Dragons breath.


    The dragons breath should narrow down where to start looking.

  19. Search for "TM902C" on the well-known auction site. If you think in Farenheit, find an conversion table and print it out as the TM902C reads in degC and only works with type K thermocouples. It's around 5-6 bucks, delivered. It's cheap, accurate and battery-powered


    You'll still need to get a suitable type K thermocouple with cable and a miniature plug, but you'd need that for the PID controller as well (though not the plug). The one that comes with the TM902C is only good to 400 degC/752 degF


    Type K tops out at 1370 degC/2500 degF (as do all of the TM902Cs in my experience, though I have some that are only marked 750 degC). If you want a thermocouple that goes hotter, you are probably talking Platinum-based and fairly serious money. I've not tried Tungsten-Rhenium, so don't know if thet would be a realistic option.


    The Auber link shows a range of 0-2300 degF for a type K input. That misses off the top 200 degF of the type K range. I think this might reduce its usefulness significantly.

  20. I have done a few remote boxes like this. It works for most drives I've come across, but one or two need separate start and forward/reverse inputs. It's just a wire to move in the remote box to make it work on these.


    The cable needs to be at least 7-core. I make mine long so that I can use the same drive for different machines by moving the box to the appropriate machine and plugging its motor into the drive.


    For the same reason, the forward/reverse switch is keyed; If the machine is not reversible, I can take the key out.


    I lucked into a bunch of multiturn pots. The one on this box is one of them and takes 10 turns from min to max. After using a single-turn it feels horrible to me, but the other guys using my remote boxes don't seem to notice.


    The pot works as a voltage divider, so the actual resistance value is not critical. It does need to be a linear trim though (not logarithmic, as is used in a lot of audio applications).










  21. That doesn't look like a sealed VFD in the pic, though it's not one I recognize. I guess you were controlling the speed from the front panel in the vid. If it's not sealed, I'd strongly advise a remote control box at the grinder and a sealed enclosure round the VFD.


    Conductive steel dust in the power electronics tends to get exciting. Briefly. And irreversibly.

  22. Not using PID control means you have to do all the control manually.


    As one of the advantages of salts is the huge thermal mass involved and it's ability to damp out temperature fluctuations, things tend to happen slowly. The intention is to control over a very narrow temperature band. The PID controller, if correctly set-up, can see the small temperature changes occuring over relatively long timescales and respond appropriately.


    If you do the control manually, you need to see the temperature readout change, usually by whole degrees, before you can respond.


    Even if you are good at it and alert, the PID controller (again I stress it needs to be correctly set-up) will typically keep the temperature within a significantly narrower band than you will be able to.


    Personally, I'd use a Venturi burner instead of a blown burner. The reasons being; Firstly that I can build a really good Venturi burner, with a commercially-made gas mixer, for significanly less than I'd have to pay for a blower or a big enough solenoid valve to cut off the air supply. Secondly, a Venturi with pilot/full settings only needs one (small and therefore cheap) solenoid valve and is its own pilot. Thirdly, they are what I know and I like them. Obviously YMMV.

  • Create New...