A thin hotface medium/low mass beer keg (drum size) furnace for melting Iron

Discussion in 'Furnaces and their construction' started by PatJ, Aug 24, 2017.

  1. PatJ

    PatJ Silver

    I do want to try some limestone as a flux, to see if I can increase fluidity before the melt reaches pour time ( I have heard that it works that way).
    I would add it to the crucible prior to beginning the melt.
    I bought some limestone powder on fleabay.
     
  2. PatJ

    PatJ Silver

    Iron is so different looking than a silvery aluminum melt.
    Iron is very red, and at the end it lays down perfectly flat much like aluminum, and has that weird red swirl, and without UV/IR glasses, it looks white hot (at least until your retina goes away, then it looks just like aluminum and so does everything else).
     
  3. PatJ

    PatJ Silver

    Scavenger started a thread several years ago called "The beach that is cast iron", or something very similar to that (located elsewhere).

    I must agree, iron can definitely be a beach, especially if you intend to machine it easily.

    And if you pull the crucible out of the furnace, add the ferrosilicon, and dilly-dally for a few seconds too many, all is lost.
    With aluminum, if I overheat it, I can go to lunch and come back and still hit pour temperature.
    With iron, you better be pouring within perhaps 60 seconds of pulling the crucible out of the furnace, perhaps 120 seconds (I know ironsides wheels his iron all over the place before he pours, but he seems to get his iron quite hot before he does that).

    The synthetic ladles that the art-iron folks use are insulating, and the cupolet iron is also probably superheated, but those guys/girls stand around for several minutes before pouring, and spend several minutes pouring a large number of molds, and the last mold is almost as fluid as the first.
    A clay-graphite crucible with iron cools off very fast after it is out of the furnace.
    It would almost be worthwhile to pull the crucible from the furnace and then set it in an insulating ladle, but nobody does that, and it would not be necessary in our setting.

    I think some add the ferrosilicon while the crucible is in the furnace, with the burner off.
     
    Last edited: Aug 15, 2018
  4. PatJ

    PatJ Silver

    The temperature probes are reading in the 100F range.
    It is difficult to see the decimal place on the readout.
    (Edit: The temperature probes melted. I don't recommend using them on the back side of the hot face).

    The extra hose length on everything is from previous experiments, and is a trip-hazard; all that needs to be trimmed off.

    I used blue painter's tape to secure the thermocouple wires to the stainless outer furnace shell, and even though the refractory shell was bright red hot on its outside face, the painter's tape survived unscathed and unmelted (protected by the two layers of 1" ceramic blanket).
    I avoided using painter's tape of the highest thermocouple wire. Did not want to push my luck.

    One observation from a previous dual-oil burner setup with burners at 180 degrees is that there were no cool spots, and no flame climbing the back wall that you can see when only using one burner.
    Dual 180 degree oil burners are used commercially, and it is a viable design with very smooth and even flame throughout the furnace from top to bottom, and round the crucible, but not necessary for a backyard setup.

    The iron scraps were very strategically stood on end parallel to each other with a little space between each, and not jammed, so that when they expanded, they did not crack the crucible.



    rImg_1087.jpg


    rImg_1089.jpg


    rImg_1088.jpg



    rImg_1093.jpg



    rImg_1097.jpg



    rImg_1106.jpg
     
    Last edited: Nov 29, 2018
  5. PatJ

    PatJ Silver

    I think this was the last temperature reading that I photographed.
    I stopped reading temperature at this point since I figured the thermocouples would fail if they got much hotter (I don't think they failed, I will have to see what their maximum rating is, but I am sure they are not rated for iron temperatures).

    It looks like 2172 F was the highest reading.
    I don't think the last reading had good thermocouple/refractory contact, or else the thermocouple or wire/joint failed.



    r20180815_203216.jpg
     
  6. PatJ

    PatJ Silver

    The melt began at 7:30 PM, and I spent 10 minutes playing around with the blower and rigging up the temporary damper.

    The last video stopped at 8:16 when the memory card fill up (such a bummer).
    I poured at 8:30.

    It felt like several hours passed while I was waiting on the melt, but it was only 1 hour, and only 50 minutes if you discount the 10 minutes required to get the blower damper installed.


    The melt size was 10 lbs iron.

    I made two ingot sizes; one thick, and one thin.
    I will break them tomorrow and see if any chill (shiny hard spots) are visible.
     
  7. PatJ

    PatJ Silver

    A photo looking down into the melt, early in the melting process.
    The solid scrap pieces are still visible.


    r20180815_194302.jpg
     
  8. PatJ

    PatJ Silver

    Here is the furnace at 7:49.

    You can see how hot the 1" thick refractory shell was towards the top (bright red).
    The lid did not seal very well unfortunately, but the lid did have a lip on it that protruded down, and that stopped much of the horizontal flame leakage between the lid and the furnace body.

    The material on top the lid is two rigidized layers of 1" thick ceramic blanket (rigidized with sprayed on thinned furnace cement).

    No apparent leakage around the burner/tuyere connection, and the extended tuyere seemed to have stopped any leakage at that joint (a problem with previous furnaces without an extended tuyere).


    r20180815_195458.jpg
     
  9. PatJ

    PatJ Silver

    I took this photo of the iron in the sand, and then took a photo of the clock and was surprised at how hot the iron remained this long after pouring (16 minutes after the pour).
    I guess the sand acts as an insulator, and if the metal takes a long time to heat up, then it will take an equally long time to cool down.

    The top of the molten iron in open-faced molds has what some call "the marching ants", which I guess is graphite or something dancing around on top the molten metal before it solidifies.
    Very interesting to watch.
    I may have a video of that from one of the ornamental art pours. I will look.



    rImg_1121.jpg


    rImg_1124.jpg
     
  10. PatJ

    PatJ Silver

    Molten iron is some strange looking stuff.
    Here are two photos from an art-iron pour at the local metal museum, and you can see zebra stripes in the metal as it is poured, and this is similar to what is seen with the swirling when you hit pour temperature.


    IMG_3762.jpg



    IMG_7995.JPG
     
  11. PatJ

    PatJ Silver

    Here is the first video of the iron melt yesterday.
    I had never used this leaf blower before, and it produced far more air than I ever imagined, even when set on its lowest setting.
    I ended up having to tape a piece of carboard over about 2/3 of the intake to reduce the airflow.
    Edit: I think the leaf blower produced too much air because the lid opening was too large.
    With the new lid and the smaller lid opening, the leaf blower works well on its lowest speed setting, with no need to restrict the blower air intake.


    I started the furnace by lighting a rag with some diesel on it first.
    Beware; the lit rag flies out of the furnace at some point, so be careful it does not land on something combustible.


     
    Last edited: Nov 29, 2018
  12. PatJ

    PatJ Silver

    Here is the second part of the video for the iron pour yesterday.
    The video card filled up 10 minutes before pour time.

    Next time I will have a spare video card standing by.

    Unfortunately a kind of boring video, but you can see how hot the furnace was getting.
    The outer stainless beer keg shell never got hot.
    I am convinced that a 1" thick shell is the way to go with iron, but we will see how well it holds up over time.
    It does have stainless needles mixed into the refractory, and those are suppose to help hold things together.

    Edit:
    I condensed the two videos into one, which is the one above.


    .
     
    Last edited: Nov 29, 2018
  13. Jason

    Jason Gold

    Ya missed the pour???
     
  14. No, we missed it. He saw it all, I guess. Unless he's like me, I get scared and close my eyes as I'm tipping the crucible.
     
  15. PatJ

    PatJ Silver

    Yes, it was a serious pisser.
    10 minutes before pour time, I saw the dreaded message on the back of the camera "CARD FULL".

    The pour was not anything great anyway, just two ingots, but it was very fluid.

    Not to worry, I am going to do it again soon, and this time I have two memory cards, and will put in an empty one 10 minutes before pour time.
     
  16. crazybillybob

    crazybillybob Silver Banner Member

    So how did the cast Iron turn out? did you break or machine it, was there any chill. I'm really thinking about making my next furnace (also a beer keg) tough enough to handle CI.

    Thanks,
    CBB
     
  17. PatJ

    PatJ Silver

    I have not had time to break the ingots, but will try to do that today (rain allowing).
    I did not use ferrosilicon for this iron, but I would anticipate that the ingots will be like the previous iron castings I have made without ferrosilicon, ie: anything 3/4" and thicker will machine easily, and anything less than 3/4" thick, and especially as you get down to perhaps 1/4" thick will be tool steel hardness.

    I had a model engine builder in the UK assure me that I could not make machinable iron engine castings without using ferrosilicon, and he also said I would have lots of slag inclusions, etc.
    I figured he knew what he was talking about since he has built more engines from castings than anyone I know.
    As it turns out, he was wrong, since I did not have any slag inclusions in the two iron flywheels I made, and no hard spots either.
    So I can only guess that some/most of the suppliers for the UK market cast iron engine parts do some very sloppy work, and /or uses some very poor quality material/procedures.
    I am sure mass production of small iron parts would not lend itself to quality, at least not in this day and age, and if you look at the commercially available model engine castings, some leave a LOT to be desired, and are more lumps of iron that hardly even resemble the final part.
    He seems to run into a lot of bad engine castings, as do others on the model engine forums.

    I did use bound sand for my cast iron parts, and paid attention to the sprue/basin/runners, gating, risers, etc., in order to minimize turbulence and the possibility of sand getting into the castings.
    I did not skim the melt particularly well, but if your runners have dead ends, and your gates are shaped correctly, the melt will self-skim as it fills the mold, thus getting rid of any slag that may still be in the melt.

    I have had others tell me that engine parts will not be machinable without ferrosilicon, but it is simply not true. As my buddy said (the guy who got one of these unmachined flywheels), this iron cuts like butter.
    Photo proof below; one of two.




    rImg_2771.jpg



    rImg_2781.jpg



    rIMG_5088.jpg
     
  18. PatJ

    PatJ Silver

    Here is an interesting guy who does an iron pour.
    I think he is the UK.




    He posted a few videos and then basically vanished from the face of the earth.
    Best I can tell he is a mechanical engineer.

    He makes lots of mistakes, ie: no UV/IR eye protection, no leathers, lace-up tennis shoes while pouring, pouring iron all over his wooden flask, clunky crucible retainer.
    But he did build his own burner, and it looks like his iron is quite hot, so that is a significant accomplishment by itself.

    One of his comments states:
    "Just gonna be honest it seems you have no clue what you are doing!"

    I have to admit I often feel like I have no clue about what I am doing with iron, so I can relate to that comment.

    But digging through this guy's material, he mentions adding limestone late in the melt in order to break up slag, so it makes me wonder if this is the appropriate time to add this material.
    He does appear to have visited some commercial foundries and so I assume he got some good tips from the pros.

    Edit:
    In his description section, he also mentions adding 120g (4.2 oz) of ferrosilicon to a 15 kg (33 lbs) melt, which does not seem like enough ferrosilicon, but I am certainly no ferrosilicon expert either.

    Edit02:
    He describes his visit to an iron foundry:
    http://thesupercorenetwork.com/metal-casting/visiting-an-iron-foundry/
     
    Last edited: Sep 14, 2018
  19. Melterskelter

    Melterskelter Gold Banner Member

    So, PatJ, you used sodium silicate bound sand for your mold, I take it. I am about to try sodium silicate molds for iron myself. One concern I've had is how much strength the sodium silicate binder will have with molten iron in it. If the information I have is correct, sodium silicate melts around 3200F and the iron should bring it within 600 or 700 degrees of that. I would expect it to lose quite a bit of strength. what have you observed regarding strength of the bound sand with iron in it?

    Nice looking cast flywheels, BTW!

    Denis
     
  20. PatJ

    PatJ Silver

    For the first flywheel, I actually used petrobond, and I did have some surface defects in one flywheel because petrobond can erode due to the high heat.
    Luckily the defects were minor and I was able to use the flywheel.
    I have seen some use petrobond with iron, but I think either water-based green sand or bound sand (either resin or sodium silicate binder) would be a better choice (just my opinion).

    For my second flywheel I used resin-bound sand, and resin-bound sand is designed specifically for use with iron, and withstands iron temperatures easily even when parts of the mold are as thin as 3/4".

    I have not tried to use sodium silicate bound sand with iron, but the folks at the metal museum have stated that it works almost as well as resin-bound sand.
    They use commercial foundry sand (same sand that I use) and it is very dry, and is probably baked at the factory to remove all moisture.

    If you use sand with moisture in it, then is may not work with sodium silicate.
    The instructions that come with the resin binder state that it will not work if there is moisture in the sand.

    I have heard of some baking their sodium silicate molds to drive off any moisture, but I have not tried that.
    Another trick that I have tried is to lightly flame sodium silicate cores (or the interior of SS molds) using a propane torch, again to drive off any moisture.
    I did try flaming a core lightly with propane, and had some discoloration, but otherwise no adverse effects, but have not cast an iron part with it yet.

    That is my next test; to cast an iron part in a sodium silicate mold.

    Edit:
    To answer your question, I think a sodium silicate mold will be plenty strong for iron, since it takes a while to heat up the sand once you pour the iron.
    I would expect the surface of the iron in the mold to solidify first, and I think this would happen before the mold began to overheat.
    No doubt the sodium silicate mold will be very scorched by the iron, but the resin-bound sand is also severely scorched by the iron, but this does not introduce defects.

    And if you are using sodium silicate sand, it is essential that you use the correct amount of SS, else the mold will be weak, and if you are using carbon dioxide to harden the SS sand, it is also essential that you do not overgas the mold, or else again you will have a very weak mold.

    I initially thought that it would be a good idea to partially harden a SS core, and then put the core in a ziplock bag full of CO2 to "fully" cure.
    This is a very bad idea, and it ruins your core strength.
    I think I use 3% SS, and I gas it for 5 seconds only.
    Some cores that I made using these figures (I need to verify the 3%, but as I recall that is the number I use) have been on the shelf out in the shop for about 2 years, and I looked at one yesterday, and it is still perfectly rigid and suitable for use.

    SS cores that I made with either a higher percentage of SS, or that were gassed longer than 5 seconds did not last more than a few days, and they crumbled.


    And they do make a sodium silicate binder that works with a catalyst instead of carbon dioxide, and I have some of that, but have not used it.
    Generally cores are easier with SS and CO2, since cores tend to be relatively small, and so it is easy for the CO2 to fully penetrate the entire core.

    SS used with catalyst works better when the entire mold is made from SS bound sand, since the sand will self-hardening, and it hardens consistently throughout all parts of the mold.
    Just don't leave the pattern in the sand beyond the strip time, else the pattern tends to glue itself into the mold. And wax the pattern well before using it with bound sand.

    I use an automotive repair slide hammer to remove patterns from bound sand, since once the sand hardens, there is no tapping the pattern side-to-side to break it lose.
    I screw the end of the slide hammer into the pattern, and give it one or two light taps to free the pattern from the sand.
    If it is a large pattern, I use the slide hammer in two or more spots on opposite sides of the pattern, and gently break the entire pattern free of the mold before trying to withdraw the pattern from the mold.
     
    Last edited: Aug 20, 2018

Share This Page