Master Yoda's Aluminium processing foundry

Discussion in 'Furnaces and their construction' started by master53yoda, Aug 18, 2020.

  1. master53yoda

    master53yoda Silver

    I have finally gotten caught up with the orders since I started the new foundry 2375 LBs This is the first week I had ingots that weren’t already sold.

    I have been promising a video, but I need someone to do the video as they will need to move from one place to another while making the video as the process of pouring is none stop once I start. And the container is only 8 ft wide. So what I’m going to do for now is put together a descriptive of the foundry fabrication, system operation, and a pour process with pictures. For those interested I can also provide the programing for the Arduino controller, I still have not addressed a problem using the furnace temp to control the firing rate on the waste oil feeds but I have identified the problem. My description of operation will assume that the temp control is operational.

    Behind the Foundry

    The foundry is contained in a twenty foot container, my reasoning for this type of housing is that we live in a dry forested area and I needed the peace of mind of being able contain even the worst case fire (close the container). The doors to the container face the road. I also have a live 3/4 “water house inside the container, I keep the floor wet as well as the road wet in hot dry weather.

    The waste oil tank is external of the container and the oil line is shut off with a normally closed solenoid valve. By killing the power, it locks out the oil feed.

    The propane feed also has an external shut off valve. Propane pressure is maintained at 10 PSI so that a ½” line is capable of feeding all 3 of my burners if I wanted to fire everything at once. It also allows me to use ¼” feed lines and solenoid valves to the burners without starving the burners.

    I use a 4,000-cfm evaporative cooler (swamp cooler) for cooling and ventilation, it completely ventilates the container in 20 seconds.

    The next post will be about the titling furnace design including the bugs I found, and the modifications needed to correct them.

    I will try to get one post out a day until I’m done.

    Art B
     
    Last edited: Aug 18, 2020
  2. JCSalomon

    JCSalomon Copper

    Do the open doors of the container provide sufficient ventilation?
     
  3. master53yoda

    master53yoda Silver

    yes wide open that is like an 8' x 8' duct you can feel the air standing in the end of the container
    that is about 1 ft per sec or about 3/4 mph wind not much really
     
    Last edited: Aug 18, 2020
  4. master53yoda

    master53yoda Silver

    When I started the design on the furnace, I had a couple of requirements,


    1 My previous locations had access to Natural Gas, the new foundry had propane, due to cost I would need to go to waste oil as the main fuel source.


    2 The size of my old furnace was based on a 60 gal compressor tank and could not take a cylinder block or transmission housing, It also was limited to about 3 cylinder heads or 100 lbs. before I had to shut it down and clean it out if the cylinder heads still had valves etc. In the last year I had some 1000lb + months and in those months I would run my scrap source out of cylinder heads and at times had to take complete engines, that required time to tear them down and when melting the block the furnace lids would not close substantially increasing melt times. The new furnace is based on a 200 gal. round fuel tank. It can handle 8 cylinder heads +, or a complete v6 engine if I need to do that.


    3 The firing rate needed to be high enough to do an initial melt from a cold furnace in about ½ hour, It would need multiple burners in order to fill the furnace volume with flame. Calculated heat needed was 5 to 7 Gallons per hour. The burners would need to have automatic firing rate control and the ability reduce the firing rate to about 1 to 2 gallons per hour for melt pool temp maintenance.


    Some pictures


    Rear burner on high fire mid burner on low fire
    [​IMG]

    7 cylinder heads and rocker arm/cam cover

    20200806_133337.jpg

    Fluxing furnace

    20200519_085149.jpg

    Center of 3 mold sets in pour position

    20200819_200113.jpg

    mold set in dump position and Quench tank
    20200819_200129.jpg



    The small furnace is based on 23 gal. compressor tank it is in 3 pieces the bottom has the burner and crucible base. The side is attached with three flange bolts for easy disassembly for cleaning out after a crucible failure or over fill. IT can be fired on propane or waste oil. I use it on propane as it only needs to fire for the first 2 or 3 crucibles as once the main furnace gets everything melted and the pool up to about 1500F I can just fill the crucible and flux, skim, dump ingots into Quench, and pour. With three sets of molds, from that point it is nonstop until the tilting furnace is empty.
    A crucible holds about 11 lbs. and will fill one set of molds. Process time per mold set is about 2 to 3 minutes. The mold set is front hinged and the back has a chain used to dump the mold. The metal plate behind the mold is to direct an ingot overflow into the quench tank. The quench tanks start boiling with the 4th mold dump that's about 125LBs. I then have to empty the tanks of ingots drain and refill the with colder water. That takes about 5 minutes then it is back to pouring.

    Till tomorrow
    Art B
     

    Attached Files:

  5. Petee716

    Petee716 Gold Banner Member

    You've outdone yourself Art. Your process description sounds well thought out and productive, but exhausting! Your previous equipment and system was remarkable and well done itself, but you've really stepped it up, especially with your furnace upgrades. How are you lifting your scrap in there?

    Pete
     
  6. Very cool setup! I'm envious.

    One comment on cooling though, why dump boiling water? It only takes about 62 BTU to raise one gallon of water 1 degree Fahrenheit. It takes about 60,000 BTU to boil a gallon of water. So if you have a 5 gallon container at 42 degrees and take it to boiling you have to add 52,700 BTU, less than the energy required to boil one gallon of water at 212F. So if you can use containers which will hold boiling water and use tongs to get ingots out of the water all you have to do is keep adding water, hot or cold, to keep cooling ingots. For what it's worth.:D
     
  7. master53yoda

    master53yoda Silver

    thanks Pete, with the exception of complete a engine I load the furnace by hand. for a complete engine I will use a roof mounted electric hoist. with the tilt all the way down it is only 35" to the opening.

    Andy, once they get that hot the plastic quench tanks get soft and start distorting, they also are close to being full enough that not all the ingots get submerged, I have always just used the boiling as sign that i needed to empty the quench tanks. I understand about the BTUs needed to boil the water is about 900btus per lb and it only take 180 btus to get a lb of water from 32 f to 212f.
    If I am close to having the tilt furnace empty I do just add water to keep every thing submerged. I also get splashed sum times when i dump the molds 212 splash is hot.

    the ingots are at about 700 to 750 when i dump them.
     
  8. Jason

    Jason Gold

    New aluminum ford trucks beware!
     
  9. 30,000 miles and they're ready for Art.
     
    Jason likes this.
  10. Petee716

    Petee716 Gold Banner Member

    In the printshop we've learned to ask the question "is it a Ford?" There's little worse than the expression on a contractor's face when he realizes why his flashy new set of vehicle door magnets won't stick to the truck!

    Pete
     
  11. :D:D:D:D:D:D:D:D:D:D:D:cool:
     
    Jason likes this.
  12. master53yoda

    master53yoda Silver

    I haven't forgot this I had a 160 lb order and had to do a pour, I'll work on a right up and pictures tomorrow
     
  13. GTS225

    GTS225 Silver

    Seems your setup is ripe for a recirculating cooling system. A pump, pulling from your quench tanks, pushing through a couple auto radiators outside, and returning to the tanks, with fans blowing air through them just might get your water usage down a big percentage.

    I suppose that depends on your water supply. If it's a nearby stream, you're golden already.

    Roger
     
  14. master53yoda

    master53yoda Silver

    I use about 10 gallons of water per pour, The tanks are full of cold water to start with, if I did a second pour I would probably use closer to 30 gallons. We have a well that puts out 20 GPM so water supply isn't an issue.

    Art b
     
  15. master53yoda

    master53yoda Silver

    Ok next segment

    I think I will go into the controls for the next segment

    In pre-"retirement= got really busy" I worked with programmable controllers in the HVAC industry, so this is control system is sum what similar. One of the primary things that an electronic control system designer needs to remember but often forget is that safety issues MUST be redundant, because it isn't if the electronic control fails it is WHEN it fails, prime example is firing rate controls cannot be the main control for burners, or steam valves etc. they are a secondary modifying control. In the case of this furnace the waste oil control can shut off the waste oil, so it could 99.9% of the time be used as a primary control but, in a loss of power instance the servo remains at its last setting, the combustion fan has shut down, the oil feed supply is compressed air so it keeps right on spraying oil into a hot furnace without proper combustion air and it may flame out, then the power comes back on and the combustion air starts back up with a furnace full of oil vapor above ignition point. Hopefully, the furnace stays in the container when it relights on its own. On a smaller furnace it would most likely be a blow the lid off event but in the case of a 200-gallon furnace there is a bit more volatile mixture.

    So what I did to resolve this problem is tied in a 12VDC Normally Closed primary oil valve with the combustion fan, now the valve shuts of the oil with a power outage. The main reason I will stay with a siphon nozzle system is that for 4 or 5 minutes the siphon air should be enough air to keep it from flaming out.

    Why I brought this up is the safe operation of most devices gets more critical as the device is scaled up or made large. We can get away with a lot on crucible furnace because of the restricted size of the firebox. We fire the furnaces mostly at peak firing rate or flame exhausting the outlet. The only time my melting furnace fires at peak is when there is metal that hasn't yet melted. once the melt process is complete the firing rate must be reduced between 60 to 75% or the furnace temp runs away with itself, it happens very fast 3 or 4 minutes from 1500 to 2300 and keeps right on climbing.

    I use a thermocouple that senses the melt pool temp as long as there is metal that isn't melted it will remain very close to 1250F. Once there isn't any more unmelted metal that temp starts climbing when it hits 1350F I shutoff the rear burner and start throttling the mid burner and limit the pool temp to 1500%. This temp ramp could be reset to work with other metals such as brass or copper if needed. There are also Potentiometers that will allow me to take manual control when i want to or if the electronic sensor fails. A sensor failure is any reading that is out of range, theses sensors show 32F if a wire is unhooked and gives a "non" if shorted. Initial startup is uses potentiometer control for the first 10minutes as a cold start high fire can smoke.

    Let me know how far you want me to go into the actual sequence of operations for the system.

    Control pics



    ID control panel.jpg control switching.jpg
     
    Last edited: Aug 25, 2020
    oldironfarmer and Tobho Mott like this.
  16. Nice description. It is worth repeating that any automated systems can be very dangerous. Similar to just walking away from your furnace while it is running. It is also a new field for most people to try to understand what failure modes can be unsafe. Most people understand spring brakes on trucks. They are fail safe - if the truck loses air the springs apply the brakes.

    Your logic for avoiding a runaway during a power failure is great. Generally I would like to see an additional spring to close safety valve so both valves have to fail (in case the primary valve decides to stick open on power failure). That's over kill for a home setup, probably.

    While I understand the desire for automated control, my personal preference would be to continually monitor the furnace and manually control the burners. I'm seriously considering building a furnace large enough to melt blocks and wheels to avoid disassembly and cutting. I was not satisfied with oxides generated and the labor involved in the burn barrel melter.
     
    Redwolf947 and master53yoda like this.
  17. master53yoda

    master53yoda Silver

    I agree that personal monitoring is a safer method, my problem is that when I start pouring it is a continuous loop mostly away from the main furnace, my clue that the melt process has completed is when the pour into my crucible starts getting hotter, by this time the furnace is in run away mode and has been for at least 5 minutes, in the automated mode I listen for the rear burner to drop out and the mid burner to start backing off. So in my case I'm still monitoring the furnace but mostly it is being aware of the automated system working. This furnace even with the safeguards I use, I don't feel that i can just walk away from it firing, I stay within hearing range as the sound changes if there is a problem. Thanks for your feedback,

    Art B
     
    Redwolf947 likes this.
  18. I an not a safety engineer, and I am not an expert in safety controls. However, I have worked with several safety engineers and this is what I have observed. For engineering controls, it seems one typically uses a redundant sensor on 2 control lines. If either or both of the redundant lines drop low the emergency/safety shutoff is triggered. The shutoff valves. or contactors should also be redundant so that both independently shut off at the same time. They are in series on the energy supply line so so that if one fails "on" the other will shut off the energy supply. Safety rated relays start at about $100 new but can be had for $25 on eBay. I have probably butchered this explanation so here is a diagram. Note in instead of an e-stop one could use temperature sensors, flameout sensors, etc. naturally one can have multiple types of trigger a stop, but they should all be redundant (2X of each sensor type) in case one fails.

    https://www.automationdirect.com/adc/shopping/catalog/safety/safety_relay_modules

    [​IMG]
     
  19. Well, after writing the last post, I checked eBay and promptly bought a for my Bridgeport milling machine, an Allen-Bradley 440R-M23084 Monitoring Safety Relays w/ Delayed Outputs for $16 with free shipping. It is way overkill, but I wanted a time delay so that I could trigger the VFD to electronically break the motor to a stop before killing the power 240 VAC power bus a fraction a second later.

    Alex Madsen
     
  20. I have started to dig into this topic more now that I have a Safety Relay for my Bridgeport milling machine. This is a nice summary explanation of Safety Circuits and when dual-channel safety is appropriate.
    https://www.automationinc.com/files/2018-06/Dual_Channel_Safety.pdf

    I have not gotten around to reading ISO 13849 yet. Perhaps if I need some bedtime light reading and have an extra $232.00 burning a hole in my pocket.

    ISO 13849-1:2015 Safety of machinery. Safety-related parts of control systems. General principles for design
     

Share This Page