This is just a very quick post before heading off to work: I ran some experiments to increase the path of the flames in my furnace. I've noticed the flames normally travel about 15cm/6" away from the fuel nozzles before performing a right angle turn and going vertically up the side of the crucible and out the lid. I made a mild steel disc that's a neat fit to the furnace bore with a section cut out of the rim. This forces the flames to travel a longer path around the plinth before exiting and may set up a circulating flame around the plinth to preheat the fuel even further. The steel is badly eroded after an hours use with signs of iron oxide spray up the bore of the furnace, so I know I hit 1540 degrees C / 2800 deg F as that's the melting temperature of iron oxide. I'm including some "before" and "after" photos of the furnace, the "before" photo is well after an hours run time of a 2 hour run and the "after" photo with the swirl plate is about the 45 minute mark. I think the next step is to cast a refractory disc a few inches thick to sit on the plinth with a uniform gap around the rim to evenly distribute the flames as the hot spot caused the cheap clay graphite crucible to fail. The iron oxide seemed to eat into the outside of the crucible like acid until it cracked and leaked. The furnace does appear to run noticeably brighter/hotter with the steel plate fitted and at higher blower airflow you can see the flames swirling under the plate with no crucible, so hot flames and gasses are hitting the fuel spray and heating the fuel which should increase the flame propagation speed. By increasing the distance travelled in the furnace I should be able to make use of higher rates of burn for slower burning fuels like oil. The fuel consumption was 25 litres or about 6 US gallons for a short ten minute run and then 56 minutes continuous, the amount of iron was 14 Kgs or about 30 lbs. Steel disc on the plinth, the tuyere air opening is on the bottom left of the bore so it has to travel about 10" to reach the opening in the disc. Furnace running at close to full throttle after 90 minutes with no steel disc: Furnace at close to full throttle at the 45 minute mark, with the steel disc, the crucible failed at the 55 minute mark: The crucible leaked at the point just above the finger, while the side opposite is still fine where the iron oxides didn't hit it.