Using the lost wax method, Le circe perdu..

This project, running now for past half a decade, had seen different setups, components. Succesful castings have been achieved but with a vast amount problems, i.e. challanges.

During this project, i became aquinted with a handfull of helpfull industrial experts who advised me, and without i couldn't get so far. I would like to thank the poeple at Eriks, Meindert van der Berg, HEC, my neigbours and friends who had the patience to listen to and helping with new setups and experiments.

Finally it worked out, i was able to maintain temperatures of around 1100degrees celsius for a sufficient amount of time to melt bronze.

The liquid metal was contained in a graphite crucible, obtained from 'De beeldhouw winkel' in Scheveningen.

A short picture guided overview:

First experiment. Using regular 'Gronings aardgas' available from the connection in the kitchen. This took place in Nootdorp, in the backgarden of a rowhouse. The natural gas  burner was obtained by decomposing a hot water 'geiser'. Measurements showed that even after 5 hours full blast the temp didnt exceed the 900 degrees Celsius. In a second experiment i used a propane/butane roofliners torch (d= 50mm, massflow 2.4 Kg/hour), which generated local temperatures of 1050degrees measures with a K type thermocouple. A inconvinient effect was the noise originating from the turbulant flame. It sounded like a fighterjet, causing my poor neighbours closing their windows and locking their doors...




The next thing was more sophisticated in design. It uses a plastic bottle as pressure reservoir. A bicycle ventile could be connected to a wheel-tube-pomp to bring it to its operating pressure of 700mBar. This airpressure is used to push the oil in the wiskey bottle out into the evaporator at the front of the burner. I thought to evaporate oil, and feed back its gas to burn it in the evaporator. Unfortunatly it was before the lectures 'Heat Exchange', so i didn't take in account the effect of film boiling. It worked nice for burning spiritus, but when using oil the evaporation proces became unstable, the operation of it uncontrollable and product was not a hot flame but really nasty smoke. Experiments where also run in Nootdorp.




After more research on the internet i found the construction drawings of a so called Babington Burner. In this 1971 Patented design presurised air is blown through a small (0.25mm) hole. On the outside of the airpipe a thin film of heated oil is running. The air blows through the film, seperating the oil in countless very small droplets that easily ignites. In the pictures above a ordinary fridge compressor is used, run at 2.5 bar. 4 of these babington nozzles are situated at the back of an old VW engine cilinder, which now functions as a flame tube. With this thing i was able to melt bronze, but the varying thickness of the oil and the unregulated heating element made its running need a constant adjusting of different valves and motor speed.

The burning of oil instead of gas seamed to be a good choice. The energy content is much higher and the flame produced in the pictures above was so hot that its radiance could be felt as a warm glow in the face. The oil burned is just waste engine oil. It costs nothing and when burned with the right amount of air it generates absolutely no smoke or odors.


Bronze doorknocker. The onreadable little signs mid left was supposed to be 'Anno 2003', on the right side there is just readable the number of the house '143, 145'. Fabricated using the
Babington burner.

As more and more meltin applications arise, there was need for a more robust and friendlier to control burner.  So after finding information about the professional oil burners, that normally use a Danfoss nozzle to turn heated (70degrees) and pressurised (7 to 10 bar) oil into a fine mist, i builded such a thing just to try. Unfortunately, the pump, motor and motorcontroller turned out more expensive that the previous designs, and as they generated a oil mist, that needs an additional device to mix the spray with air in the good ratio, a good combustion was never realised and the setup was dismantled when the need of an apple grinder required its parts. When i think back, i am still amazed that it was so difficult to get good mixing. I quite sure that for our line of work, this could have been usefull once its proper working has been secured...

A next design was based on the page of Lionel, (, but in my workedout prototype (oil dripping in airpipe) the air pipe didn't get heated enough to evaporate the oil. I spent limited time on this, roused by the following setup:

I learned by a simple experiment, that when a roofliners butane torch, filled with parafin oil in a separate resevoir, produced nice flames when the reservoir was tilted and the oil was atomized by the speed of the gas expanding through the nozzle. This developed in a gas burner in which pressurized oil was fed. A small .60 Bing nozzle was were the expansion took place.

Test setup: Around 2.5 bar gas pressure, with increasing amount of oil added. Note in this setup, no additional mixer is involved. The gas just mixes with the air as it expands and sucks the air.


As i moved to a other house with a paradisical garden, with a little effort, a nice melting furnace was constructed. The inner stones with the yellow tan are special firebricks (Purchased in Delft, Koolschijn at about 2,50 each) and could withstand at least 1200 degrees celsius. The alumium thing is a preheat stage, obtained just with luck at Meindert v d Berg, Rotterdam. The exhaust gases now heat the intake air and effiency could be reached. I found somewhere that this preheating increased the capacity of the early industrial furnaces with a factor 3. The reference is however lost. The picture down in the middle is a test with a half build furnace, and altrough it looks realy hot, the temperature was limited by the amount of false air that was sucked between the stones at the inlet. Afterwards i filed the space around the pipe with cement. Mr. E.S. is smilingly gazing at a red hot glowing copper pipe.


Air/gas mixer. This thing creates is inserted in the end of the hot air carrying pipe.The 0.60 nozzle at the end of the pipe is positioned in such a way that it just fits the small hole at the beginning of the mixer. In the picture most left  a second plate is seen, positioned around 20mm away from the back, with a opening of around 10mm.

As this page is expanding with state of the art experience (and failures alias opportunities) i will try to share some knowledge, eventhrough this almost diarylike content is far from ideal. The main thing what can be leaned will probably be what things you better don't do. A insight gained from the red tank is that when you pass oil through a long pipe of small diameter (in this case 1mm diameter and 40cm long) is that the stabilty of the fluid flow will be very sensitive to variations in viscosity (read temperature). Altrough  i used a controller on the oil heater, every increase resulted, after say 20 seconds, that is the time rewuired for the heated oil to reach the end of the pipe, resulted in unexpected large oil output. The first thought was this coming from gas formed the heater element, but aforementioned explenation sounds more reasonable. To overcome this problem, a new transport pipe was build. Now the small diameter pipe is only used at the end, for a length of 3 cm's, so the effect of the decreased viscosity will take place, after a longer time and far less intense. The thing was made in a modular approach with extra valves en experiental engraving, like the antique physics stuff you may someting see. It was difficult to get the letters straight using a sharpened screwdriver as 'burin' (=engravers tool), but it is nice to use the authentic process. Experiment for next time would be using wax and etchant to get more straight results. For the connaiseurs of  this kind of crafts, visit the Zeppelin museum in Friedrichshafen, Germany. There you see the early Daimler engines with so many detail, that even they mechanical layman would recognize 'maschinenbau' as technological art..

Update oktober 2009.

During the Mechanical-Construction-Tuesdayevening, a weekly event in which  the repairing of benzine/diesel/oil burning things serves as a alibi for a evening of man talk, winedrinking and fun, i walked into a car repair shop opposite our "workshop", in search of some useable steel, when i got a glimpse of a real, very rare in the Netherlands, original oil burner! The burner was broken, so the day after i could trade it for a crate of beers. The pump was weared out, which was replaced with the Danfoss BFP 3 pump, which was used in before mentioned experiments. This burner also had a electrical flamewatch sensor, which cuts of the fuel if the flame is extingished. The connections of it were worn, rebending them to a position were they made contact  solved the problem.

These photo's witness the first experiments. The main setup looks a bit messy, but hey, work is in progress. The glas jar with the white content is borax, which is used as a slag binder. It also forms a thin layer on top of the melt that screens of oxygen. The first test was succesful, burning 6 ltr diesel in 1.5 hour, and obtained temperatures above 1100 degrees. Copper tubes could be fed with a interval of seconds.  The flame is sharp, stable and brightly luminant. Power is adjustable between 19 and 83 kW by adjusting the pressure screw. Air can be regulated by a small spring tensioned handle, just above the motor.

Update December 2009

Just a december picture of the snow covered furnace, waiting for fresh fuel and dry molds....

Writing about molds...

Now the burner is reliable enough to melt copper and bronze, the pouring of the melt in the preheated plaster molds resulted in excessive gas formation inside the mold. All tests objects were of such a poor quality they were unusefull. Further study revealed the water in the plaster gets chemically bonded, and is only released after long time expose (6 hours and longer, depending on the size) to temperatures of 600 a 700 degrees celsius. This i never did, only preheated in the upper furnace chamber for as long as the melting took. The picures below below are 3 casting test of (unofficial title) "Head of a girl" by Hanneke Wessels. Disfigured by cracks, pins, holes, too thin walls and nonfilling. All were cast with  hollow, melted out wax copies of the original model.

New molds have been made, below they are hardening in the window sill. For the baking out process i bought a old RVS pan and cut out parts of its baseplate. Standing on the kitchen furnace, it makes a nice hot drying chamber! Same fast experiments showed temperature readings of roughly 300 degrees celsius, which i think will rise further after some time....Hope it will be sufficient....

Update April 2010

The kitchen pan experiment didn't turn out succesful. It could not gain suffiecient heat to warm up as hot as i actually wanted it. So.... A new thing was made, a forced air, wood fired and preheated hot chamber! Check out the pictures. The hollow tank on top reaches temperature of around 800 degrees celsius. Enough temperature to free the plaster molds of the chemically bonded water. The whole thing is insulated with glaswool ( T.P., thanks for the idea, and C.K., thanks for the Robax glassplate!).