Experiments with getting the oil furnace to melt iron

Discussion in 'Furnaces and their construction' started by Mark's castings, May 9, 2023.

  1. Today I ran some more experiments in getting my particular 28cm/11" bore furnace hot enough to melt iron. The plan was to have 50 PSI diesel feed my small brass pressure sprayer nozzles which gives a spray fine enough for a cold start on the diesel and then once it was warm enough withdraw the nozzles so that they sprayed the conical sheet stainless air nozzle on the inside: it should be red hot as it would be inside the furnace chamber and hopefully the diesel hitting the metal would vaporize it directly into flammable gas... or so the theory ran. At any rate withdrawing the nozzles made the temperature plummet each time it was tried. Even with the narrower 38mm/1.5" air cone the furnace seemed sluggish compared to normal and was slow to reach 1000 degrees C, where with the old larger air cone it could reach barely 1305 degrees C. I did notice that running it really rich, like on the verge of black smoke rich while at any fuel flow setting seemed to keep the temperature rising. Once I set the fuel flow to it's normal rate of just under 20 litres an hour and throttled down the air flow, I had the temperature rapidly rising and the furnace looking like a video of those Pakistani foundry furnaces with 60cm/2' yellow flames out the top. By now the furnace had been running for over an hour, and burned 20 litres and I switched over to the reserve tank and the furnace hit 1363 degrees C / 2495 deg F and the temp display went blank after that as the K type thermocouple melted at 1365 and climbing. I shut it all down and poured a 150mm/6" long iron ingot into a preheated steel mould with lots of solid iron oxide dross from the brake rotors in the crucible too.

    The ingot was cut in half with a grinder and seems sound inside, I'll have to drill it and see how hard it is.

    running rich furnace.jpg

    furnace temp.jpg

    iron ingot section.jpg
    Last edited: May 9, 2023
    Tops likes this.
  2. Petee716

    Petee716 Gold Banner Member

    Reviewing my last failed attempt to pour iron, I believe too much air was the culprit. The iron had gotten liquid and I jacked up the fuel and air to try to super heat it, and it went back to thick mud. By then I was tired from previous pours in other metals that day, and I was irritated, so I dumped my mud and called it a day. I haven’t tried it since, but I’ll give it a go.
    Relatedly, your story reminds me of what I’ve been told are major factors in combustion: the 3 T’s. Time, Temperature, and Turbulence. I understand the conventional wisdom regarding the Pakistani flame cone-that it’s an indication of the fuel burning outside the furnace- but those conditions could in some way be allowing more turbulence and dwell time in the furnace for the richer fuel remaining in the furnace. Of course there are other factors to consider, and I may be way off base. I know that while Denis abides by that philosophy, Ironsides’ videos seem to show a somewhat high cone at times. Both with great success, although with different burner types, etc. Throwing the heat in any which way we can gets us by for melting our aluminum and copper alloys but we have to get dialed it for consistent iron. Maybe we’ll get some schooling here.

    Tops likes this.
  3. I'd pretty much run things for an hour without getting much over 1000 degrees C, was getting sick of it and ready to dump the contents on the ground, when all of a sudden I had things dialed in and the temperature rapidly rose to 1200 deg C and kept on rising over 1365 deg C. In that last 10 minutes or so, it was running hot...not efficient or clean, just hotter than ever. I may swap in the previous air nozzle and run at the same super rich settings which I've never done before... it goes against instinct but the thermocouple doesn't lie until it melts at least. I need to get the cheapo S type running although I suspect only the junction is platinum-rhodium and the rest of the wire will melt at more mundane temps. I'll settle for consistent iron melts and then aim for low fuel consumption, I burnt about 25 litres in diesel but could probably halve that by running it rich from the start.

    Edit: The ingot drills like nice grey iron too, without added ferrosilicon.
    Last edited: May 9, 2023
    Tops likes this.
  4. Melterskelter

    Melterskelter Gold Banner Member

    It looks to me like your fuel burn rate is way high. My optimal burn rate is .18 to .19 liters of diesel per minute for a furnace about 20 inches deep and about 14 inches diameter. I would suggest just trying to run it proportionately at something like that rate with sufficient air to minimize the flame visible coming out the vent---just a faint color that is just barrely visible on a sunny day and best seen against a dark background. (Burning 5 gallons of diesel requires you to heat about 600 pounds of air while heating much less iron.)

    For example yesterday I poured 65 pounds of molten iron after burning 21 liters of diesel in an hour and 45 mins. I could just see a 8 inch cone of faint orange light above the chimney. This setup is extremely consistent for me. I can reliably melt 3 pounds of iron with one pound of dieselt to 2550F poutring temp. So, even if it does not at first seem right, just give it a try for an hour. What's to lose. I think any reasonably efficient system of vaporizing diesel should provide similar results. And It might be possible for me to tune mine yet better. But once I lose visible flame I am not too sure how much past the optimal settings I might have gone.

    Last edited: May 9, 2023
    Tops likes this.
  5. Hi Denis, it's kind of weird, in the end it was running at "normal" for this furnace rate of consumption: 15-20 litres per hour but with the airflow turned down to ridiculous levels. It is a high mass furnace so it does take some energy to warm up the concrete for the first run but there's other factors at play. For any given fuel flow, varying the airflow makes the flame length lengthen and retreat (higher airflows) so I may just be tuning things to make the hottest part of the flame hit the thermocouple rather than having the hottest furnace chamber. I did spend an hour trying different fuel flow rates and all of them seemed need a rich mixture to get hottest and that did waste a lot of fuel, it may well be that ditching the circular flame track under the plinth and just running rich will get the best consumption results. It just seemed counter intuitive that a rich flame runs hottest but does match a lot of small third world furnace videos.
  6. Melterskelter

    Melterskelter Gold Banner Member

    OK, Mark. Maybe there is something fundamentally different about your furnace and its operation. But, it took me a long time to figure out that, yes, I could melt iron with a "fat" flame but I could melt it a lot faster (half the time) and for less than 1/2 the fuel with a thin flame.

  7. I'm at the point where ANY sort of successful iron melt will do... efficiency can come later with experience. I took some photos of the aftermath in the furnace, not much in the way of soot and the thermocouple took a beating: even the ceramic beads melted from the flame impingement. Running the hot settings right from the start and reverting back to the original air cone nozzle should give a more normal fuel consumption and burn time.

    The thermocouple tip, with only one wire left in the ceramic blob:
    Ded thermocouple 1.jpg

    Ded thrmocouple.jpg

    The furnace lining with the usual iron oxide coating, no soot would deposit at operating temps anyway.
    furnace clean.jpg

    Soot should be impossible to deposit so not a real indicator of air/fuel mixture.
    furnace lid.jpg
  8. Tops

    Tops Silver Banner Member

    For something like this, would it work to have the thermocouple embedded in the liner instead of out in the flame?
  9. Of the two thermocouple wires, one alloy melts sooner than the other and falls apart at not much more than 1350 degrees C. Embedding the thermocouple would help protect against oxidation of the wires. About the only real solution is to use something like an S type thermocouple with platinum/rhodium/rhenium/tungsten alloy wires which would be much more oxidation resistant and higher temperature rated to about 1600 degrees C. I have a cheap S type but I suspect there's only a small amount of platinum right at the junction so the attached wires would melt anyway. It only claims a 1300 degree C maximum operating temperature.
    Last edited: May 10, 2023
    Tops likes this.
  10. Melterskelter

    Melterskelter Gold Banner Member

    Recent discussion here got me to thinking about leaning out my mix a bit more. So, today, a bright sunny day (affects visiblity of flame) I leaned the fuel out a tad more and got the exhaust to show only the faintest of red color or no color at all. And, I melted 60 pounds with 17.5 liters diesel and in 92 minutes to pouring temp of 2560. This is about a 10% improvement. This suggests that in my setup I should be trying for a bit leaner yet. Today the flow rate was .19L/min. and I turned the air up to get the flame to disappear. FWIW.

    Tobho Mott and Tops like this.

  11. I've been thinking about possible reasons for why such a super rich mixture gets hotter in my particular furnace: one reason mentioned ages ago by Kelly I think, is that a rich mix burns faster than a lean mix. It would compensate for a low dwell time in my particular furnace, also the higher the temperature, the exponentially faster the fuel will burn. For now my priority is to reliably melt iron with fuel economy running a close second in priority.
  12. Jason

    Jason Gold

    Can't help you with the iron thing, but do I need to dig up my fuel vs heat bell curve for you? lol
  13. The analogy I had in mind is a circus bear riding a minibike round the ring: it's not that it rides well but that it rides the bike at all o_O
    Jason likes this.
  14. Jason

    Jason Gold

    lol.. Funny mental picture!

    In a normal world, any deviation from peak exhaust gas temperature be it rich or lean will result in a lower temperature. It's super easy to demonstrate this principle in a piston airplane with a good egt gauge and a fine touch on the mixture knob. The engine manufactures want you to find peak temp, then enrichen the fuel usually by 75degrees C. to keep from burning valves or elevating cylinder head temps too much. (400c) Now, with a good set of injectors and individual cylinder head temp gauges, it is possible to peak the egt and the LEAN some more! The loss of power is less and cooler cylinder temps can be achieved. Sloppy rich is never good, it wastes fuel and accelerates cylinder wear washing off the oil.
  15. I remember reading about Charles Lindbergh doing a lot of flights to prove leaner mixtures weren't going to cause immediate engine burnout and saved the US Army Air Corp a lot of fuel during WW2.
  16. Looking up the melting points of Chromel and Alumel used in K type thermocouple wires and I would have to have reached 1399 deg C alumel melting point, but not 1420 deg C that the second chromel wire melts at.
  17. Ironsides

    Ironsides Silver

    I have a large tree stump in my backyard to burn out so I tried an experiment using a vacuum cleaner to get the stump to burn out quicker. What I found was the charcoal started to glow brighter as the blast was closer to to the stump. At 150mm it reached a peak of 1100C using a cheap infrared pyrometer. Moving it closer to the charcoal it started to glow less and putting the blast about 25mm to the charcoal put the charcoal completely out. So it seems that using any fuel (gas,liquid or solid) there is a happy point where the fuel burns the hottest with regards to air fuel ratios.
  18. I'm working on a similar theory: the burning fuel has to pass through a narrow gap of about 2.5cm/1" between the crucible and furnace wall. I did some calculations and worked out that by using an A6 crucible (9Kg bronze) I can go from the A25's diameter of 22cm down to the A6's 13cm which would halve the velocity of the flame by doubling the surface area of the air gap (from 1" to 3" air gap). If that works, I'll look at increasing the bore of a new furnace to 40cm minimum up from the existing 28cm.

    The last run I did barely melted the iron but slagged the new thermocouple in transparent white/straw coloured flames. I got rid of the refractory disc that forms a circular doughnut chamber under the crucible and this lost 40 degrees C, but let me ditch the super rich fuel settings and also lowered fuel consumption. The narrower air cone gave an even flame that formed a double helix which I'd never seen before: you can see two flames 180 degrees apart exit at an angle instead of the usual one.

    I did get a 3Kg iron slug cast that cleaned up ok and machined well. The surface finish was great due to the low iron temps and some sea coal in the green sand.

    twin helix.jpg
    Last edited: May 26, 2023
    Petee716 likes this.
  19. Ironsides

    Ironsides Silver

    Have you got a A6 crucible? If you have, do an iron melt to prove that a larger air gap will get a hotter melt. If I use a A6 in my furnace I have to be very careful not to overheat the iron because there is a small amount of iron to heat up. The perfect size crucible for my furnace is an A12 but I have used a A20 in it and it seems to get the best fuel used for every kilo melted.
  20. It's good to know the smaller sizes can be heated higher, I didn't know it was possible to overheat iron :D. I ordered an A6 from Skamol in Melbourne last week so it should be here this week. I've been watching your video of the North Pines Trades Guild iron casting session and comparing that smaller furnace built in a 90 Kg cylinder. It looks like it's bore is comparable with mine and they are using a crucible about A8 or so. It's visible at the 7:23 minute mark.

    Last edited: May 29, 2023
    Tops likes this.

Share This Page