Possible explanation for my oil fired furnace tuning issues.

Discussion in 'Furnaces and their construction' started by Mark's castings, Jul 13, 2021.

  1. So I've been trying many furnace tuning sessions to melt iron as fast as I did for only two really great sessions that were relatively fast. My furnace is a tight fit for the A25 crucibles with about an inch gap between the wall and the crucible, this is not an issue for LPG/propane gas furnace it was copied from but I'm running diesel fuel and can't reach the desired temps needed for iron service. I've had maybe a dozen sessions that only just barely reach molten iron temps. I have a 2Hp blower that is double the desired size and throttle it down with butterfly valve to a usable volume.

    So recently I experimented with a grille fitted to the airflow, I'd heard third-hand about these being used in older commercial furnaces from a discussion between a World War Two foundryman and a guy who successfully solved his furnace issues by fitting a drilled plate to the airflow. So I got some 5mm thick aluminium sheet and drilled a whole bunch of 6.35mm/0.25" holes and fitted it to the 4" bore PVC air pipe of my furnace. On the whole it made a slight improvement and for some reason the furnace ran noticeably quieter: probably the 4" bore allows air to spiral and swirl it's way through the pipe. The grill would provide some resistance to flow and also even out/ laminarize the airflow which cuts the combustion deep rumble mostly. After I sent a photo of the grille to Peter: it jogged his memory about the early 1980's grille construction conversation and he recalled that the holes should be closer to 3.2mm or 1/8" diameter. So this would provide some significant flow resistance and slow the airspeed as well as evening it out.

    air grille.jpg

    Anyway as I'm getting ready to make another grille with 1/8" holes, it percolates through my maggoty thought processes that the successful furnace runs had some 90 degree and a couple of 45 degree bends in the airflow pipe plumbing. So a bit of online research later and I find a nice chart that shows how much pressure drop (and airflow speed) a single 4" bore by 3" centerline bend radius 90 degree elbow can cause: about a 53% reduction in pressure/flow. So for my successful iron pipe setup with a 45 degree Y, a 45 degree bend and a 90 degree bend, that would be a pressure drop of 37% per 45 degree bend and 53% per 90 degree bend: 0.63 x 0.63 x 0.47 = 0.19 of the original pressure.

    furnace air plumbing 2.jpg

    The not so successful runs had a 45 degree bend, a 15 degree bend and a 45 degree Y which would be: 0.63 x 0.85 x 0.63 = 0.33 of the original pressure. Looking further at Peter's successful gas fired furnace it has a 1Hp blower feeding about 15' of 2.5"/65mm bore PVC with four 90 degree bends, three of which are wide radius gentle bends, so the length and number of bends would cause some air flow resistance. So obviously I have to duplicate the original successful two iron sessions air pipe plumbing and see if the furnace works again and maybe rig up a water manometer or two. I'll report back on the results soon hopefully.

    air plumbing 1.jpg

    Bend loss coefficients for a pipe (Babcock & Wilcox Co., 1978).
    airflow-bends-coefficient.gif
     
    Last edited: Jul 13, 2021
    Chazza likes this.
  2. rocco

    rocco Silver

    This reminds me of something I heard of a long time ago, although in a totally different context, someone was moving air though a section of 3 or 4" round duct, the duct work had several bends and turbulence in the airflow was causing problems, he solved his problems by packing a short section of the duct with drinking straws just downstream of the final bend in the duct.
     
  3. That's a classic example of making a flow laminar and smooth.
     
  4. Melterskelter

    Melterskelter Gold Banner Member

    Mark,

    I cannot get past the idea that a 2 Hp motor on a backyard furnace blower seems like it is 1.75 Hp too large. I am running a repurposed originally battery-powered 20v leaf blower that I throttle down with a sliding gate valve. It is now powered by a 20v DC power supply. But, my point is that the volume of air it moves must be a fraction of yours. I burn .19L diesel per min. An A25 loaded with an initial charge of roughly 30 to 35 pounds of iron is sweating iron drops in 30 mins or so. The entire 60 to 65 pound charge is melted within an hour or so. Pouring temp may be 40 mins or so after the charge is melted. My furnace seems to be tuned best when the flame out the chimney is not a yellow color but has more of a reddish hue and is not very bright. I always look at it so that a dark green fir tree provides the background rather than a bright sky. FWIW. I know for sure my furnace is burning cold when the flame is bright yellow and 1.5 feet long.

    Denis
     
  5. Hi Denis, the blower was built around stuff lying around: a 2.5Hp impeller and a 2Hp motor, I've since picked up a nice 1Hp three phase blower which is on a par with other furnace units I've seen that also heavily throttle the air down. Last time I had a good run I'd melted 15Kg (32 pounds) of iron in 62 minutes (poured in the mould) with 22 litres of Jet A1 fuel which would be 0.35 litres per minute. The Morgan furnaces use between 10-21 litres per hour of fuel oil to melt for 20-50Kgs of iron so I'm currently around double the fuel consumption but getting the iron liquid in 30 minutes or so versus 50-60 minutes for the Morgan furnaces with more iron in them. I'll need a lot more successful runs to be able to tune it efficiently but at least it's half as efficient as a Morgan furnace.
     
    Last edited: Jul 13, 2021
  6. Melterskelter

    Melterskelter Gold Banner Member

    Mark, your fuel burn rate just makes me wonder if your furnace is getting the air/fuel mix pushed through it too fast so that the furnace is functioning as a pre-combustion chamber. There must be a reason it so much greater than both my furnace and the Morgan for a 20 to 30Kg melt. Maybe you have already experimented with lower blower and fuel settings. But, if not, I’d be tempted.

    Denis
     
  7. The size of your blower is really not a factor. You can choke it down to a small wisp of air. With only a 1" annular space it seems you don't need much air. The elbows just affect how much you need to choke your blower and again, are not a significant variable.

    The air restrictor you showed may created turbulent flow. It all depends on the velocity. Turbulent flow is what you want in order to mix fuel well with the air to get fast combustion. Having many small jets at high velocity can produce a nice front of turbulent air. If the velocity in the pipe is not high enough to create turbulence it will soon go back to laminar flow. Elbows also introduce turbulence.

    I think your pipe is too big. Have you tried 2"? Get some high velocity and good turbulence and you'll get better combustion. High velocity without too much air flow takes smaller pipe.
     
  8. Hi Denis, there seems to be two different modes of operation with this furnace: I can remember texting someone that the iron was beginning to melt at the 19 minute mark on a good run and then hardly at all after an hour on the bad runs which would support your theory about it not burning fast enough for the specific conditions. I do have the toroidal shape precombustion chamber formed by the refractory disc sitting on the plinth which was also present for all good runs. There is even a third mode with the airflow and fuel turned up to maximum where there is a single blue flame spiralling up the furnace chamber and the furnace runs cool with unburnt fuel forced out the base of the furnace through cracks in the refractory by the pressure.


    Yes the pipe is too big: the flow without a cone air nozzle is so slow that the flames travel into the chamber, performed a 90 degree turn vertically and exited the furnace while heating only one side of the crucible. Having the air cone nozzle and the refractory disc on the plinth gives a uniform flame swirl round the furnace.

    Turbulence is quite possibly a factor, one thing elbows and bends will cause in a big way, I have tried a refractory lined pipe for the last 10" or so around the fuel nozzle that brings the flow down to 2.5" or so. I can try a run with 2-3 90 degree elbows to test the turbulence hypothesis and even fit some sharp edges on the corner as a "turbulator" like the lower elbow diagram.



    duct-hvac-friction-turbulence-acca.jpg
     
    Last edited: Jul 14, 2021
  9. Another thought I had today: as the blower air hits the right angle bend, will it begin to spin like a cyclone and spiral/corkscrew down the pipe?. Might have to try and see if a vane spins or smoke spirals or something.
     
    Last edited: Jul 14, 2021
  10. Petee716

    Petee716 Gold Banner Member

    It would seem to me that the only place turbulence should be a factor is where it mixes with the atomized fuel. All the rest would just be transport/efficiency issues. Your perforated disk would be most effective immediately upstream from your oil nozzle where it it would act as a diffuser providing a turbulent environment in the combustion chamber. Your changing furnace temperature is also a factor, as well as how much dwell time your mixture has to burn which is a function of your air volume and pressure. I may be talking a bit out of school here, but I've been told that time, temperature, and turbulence (3 Ts) are the critical factors.
    Just my 2 cents.

    Pete
     
    Mark's castings likes this.
  11. The perforated disc is located about 4 inches upstream of the nozzle as you suggest, I'd been theorizing that the elbows are either slowing the air speed, reducing pressure/flow rate, introducing a vortex or introducing turbulence or some combination of the three that has the same effect as the perforated disc, inducing turbulence in an indirect unreliable manner. So a turbulent airflow would be desirable for a furnace tuyere?. I'm thinking of testing for a vortex flow with a kind of bladed fan with blade pitch aligned with the bore such that it won't spin unless there's a spiral airflow to hit the blades at an angle.
     
    Last edited: Jul 14, 2021
  12. Billy Elmore

    Billy Elmore Silver

    It gets tricky in smaller furnaces. I had issues where I thought I needed to slow down the velocity to get it to burn in the furnace and not blow it out the top. I switched to a bigger pipe and tried it and found it got worse. Instead of going back to the smaller pipe I added more air pressure to the bigger pipe and got much better results. Im sure it was strictly a velocity thing for me as there is no difference except the diameter of the pipe. I also found that if I increased it too much I would blow the flame out but right before it would go out it would have a big flame out of the furnace and was pulsing.
     
  13. With 4" bore pipe, I can't get the flame to swirl round and take a longer path in the furnace even with the way oversize blower. I had to neck it down with a cone round the nozzle which helped swirl and only got an even flame by sticking the refractory disc on top of the plinth to form a chamber. I'd like to ditch the disc and still get swirl if possible.

    I've also had the furnace running like a pulsejet engine, lots of noise, fuel and air but it was fairly dark and cool running. This last run I tried it again and had a single long blue flame for the first time with a small amount of orange so it was running clean at least, just not hot. I'm only 1/4 through drilling the new perforated plate so far but I'll copy the successful setup and try it with and without the plate fitted. If it melts iron with the plate near the nozzle then I'll try some smaller blowers I have lying around. I also picked up a nice long piece of about 3" stainless pipe complete with four 90 degree elbows: wide radius gentle bends too. It would come in handy for reducing the pipe size from 4".
     
    Last edited: Jul 22, 2021
  14. Melterskelter

    Melterskelter Gold Banner Member

    3” may still be plenty big. I am using 2.35 ID pipe. For a given volume of air / min the velocity of air should be about 1.5x that for a 3” ID and 3x that for a 4.

    Denis
     
  15. It's good to know what successful furnaces are made from, this stainless seamed tube is a weird size about 67mm bore (2.64") and 72.5 O.D. (2.85") possibly for building fire sprinkler water plumbing. With the 4" PVC, I'd originally thought I'd neck it down with dense castable refractory and try and get the flame into the tuyere some distance to give the oil a chance to burn fully. As far as blowers go I now have a 1/2Hp, 1Hp and the 2Hp that I'm currently using. I'm going to switch to the 1Hp once I get hold of a suitable inverter and rewire the motor to 240V three phase.
     
  16. Melterskelter

    Melterskelter Gold Banner Member

    Taken 1 min ago:) 8116CA34-B2BD-4647-AAAD-97C88D508F7C.jpeg

    Denis
     
    Mark's castings likes this.
  17. Just finished machining the new air grille/baffle/linearizer thingy with 1/8" holes ready for testing soon. It could also go up to 3/16" if it restricts flow too much.

    one eighth air grille.jpg
     
  18. Peedee

    Peedee Silver

    What about drilling/reaming those holes at a 30 deg angle, multiple vortex pattern or am I over thinking this.....
     
  19. Petee716

    Petee716 Gold Banner Member

    Maybe it's against machining etiquette but leaving the burrs on the face pointing downstream could help the diffusion as well.

    Pete
     
  20. There's room for that but I'd have to do it by hand unless I buy a tapered end mill or something. I'll test it as-is in the hopes it gives a uniform airflow across the tuyere.

    My drills were getting a bit blunt towards the end as the holes were getting a uniform extruded rim maybe 1/16" that I filed off periodically. Once I finished, both faces were sanded to get a nice sharp 90 degree edge and it occurred to me that countersinking it would aid airflow. The holes are 5mm deep and 3.2 diameter so they are deeper than the width. Using a cutting lubricant wasn't really practical, maybe a shallow dish with a sacrificial backing under the plate. The holes drilled with lubricant are very smooth almost reamed in appearance.


    My main aim with this plate is to make the airflow uniform across the width of the pipe and even out the airflow so the spray nozzle has the best chance of mixing with air.
     
    Last edited: Jul 25, 2021

Share This Page