For hobbyists it is possible to build a gas fired metal melting furnace for small amounts (50..300 grams) with the following features:

I started long ago in 1973. I used the bottom third of a 200 liter oil drum, filled it with sand and left a hole in the sand of about 20 cm wide and deep, put some ordinary (nonrefractory) bricks around it and that was the furnace. I used anthracite as fuel with a reversed vacuum cleaner as fan. To light the coal I burned some wood first. It reached temperatures until 1600 C (2900 F) and I used stainless steel 'crucibles' for melting copper ... which leaked after two times of use, because the alloy components of the stainless steel dissolved in the liquid copper. And I started with home made foundry sand which I later replaced by 'real' foundry sand for better results. Later on I also got a 'real' 500 cc graphite crucible. The melting capacity was about three kilos (6 1/2 lbs). The furnace wall bricks were completely cracked and 'slaggified' on the inside.
I suspended this hobby for 20 years and I resumed it in 1992 with a large furnace from refractory concrete which is able to heat 500 cc crucibles for melting 3 kilos of copper in 45 mins with a roofmaker's propane burner. The furnace was just an oil drum of about 30 cm (12") diameter and height and I lined this with a layer of 7 cm refractory concrete. From the top of the drum I cut about 7 cm, filled it with that same concrete (leaving an exhaust hole just off the center) and that was the lid. Now I call this furnace the 'roar furnace' because it roars so loudly that once on a summer day when I was melting bronze a neighbor thought that a jet airplane was flying low. I use it very seldom, only when I make large castings. A year later I also made smaller castings and started silver- and goldsmithing. I bought high resolution oil-based sand (purchased at a silversmithing supply store) for small (< 100 grams) bronze and sterling silver castings and home made sand for coarse (cast iron) and larger castings (bronze). The small furnace can be used in the garage too because I have a hood with a vent. The description below is about the small furnace. In October 1995 I discovered that one of the cheapest buiding materials (air - spaced gypsum) turned out to be one of the best and most insulating refractories. I could make the furnace so hot that it is able to melt cast iron and shorten the melting times of copper, silver and gold alloys further.
These items are required:
![]()
The furnace should have a cilindrical shape, the best size is between 1.2 and 2 times the diameter of the crucible (1.5 is optimal) and 1.2 to 2 times the height of the crucible. In the bottom of the wall a hole for the torch should be made with tangential entry, which forces a rotating movement of the flame and hot gases around the crucible. When the flame enters straight to the center, the cooler part of the flame hits the crucible, thus cooling it. The torch hole should fit exactly, otherwise hot gases exit here with overheating (and melting if it is made of brass ) the torch nozzle and more heat loss. It should be open on top. The lid should have the same diameter as the furnace itself with a hole in the center with about the same diameter as the torch entry point. Here the exhaust leaves the furnace and the user can watch the melting process (again: use face shield when getting close).

Another option I tried where gypsum blocks for building inner walls.
These are sold in Europe under the name YTONG and have good refractory
properties (until 1450 deg C) and are remarkably cheap ( about $1 for a block
of
60x25x8 cm). They are available at every builder supplier. They consists of
virtually pure anhydrous calcium sulfate (CaSO4) and
are air-spaced and thus provide very good heat insulation. They are also easy
to work with. A worn-out ordinary saw can be used for cutting the material and an
ordinary screwdriver can be used for other workings, such as carving the hole for
the torch and crucible.
I tried one of 8 cm and made a cylindrical hole in it of ca. 5 cm
diameter and 8 cm length in it and in the bottom a hole in which just the
torch fits in. I put it on another one wit a flat top and put the crucible with
ca. 40 grams of copper in the hole. It melted within five minutes. But there are some
(minor) disadvantages:
For use as soldering blocks this material is excellent. It is soft (nails to fix the work can easily be pushed in), heat-resistant and cheap. Now I don't need the far more expensive soldering blocks sold by silversmith suppliers anymore.
The amounts of this blocks needed are small, so it suffices just to look at trash heaps at building sites.
Here are some guidelines for working with molten metals. Of course, the mold should be ready before melting the metal.
When one needs very high temperatures (> 1400 C), e.g. for melting platinum, iron,
etc. the following possibilities are available. However, the higher temp is needed,
the harder to keep it. This is because of the law of physics which tells that the
heat loss by radiation equals the fourth power of the temperature difference
with the surroundings. An example is the fact that a candle has a flame temperature
of 1500 C, but no object (except the very tiny soot particles which are responsible
for the candlelight) of a reasonable size can be heated over 1000 C.
Note: Except the oxypropane torch I've never tested or tried out these things,
but I'd like to someday.
Use always dark (welding) goggles when looking into such hot objects, otherwise
damage to your eyes will result.
Mail comments to
Klaas van Ditzhuyzen
Velp, The Netherlands
Author: Klaas van Ditzhuyzen, Velp, The Netherlands
ArtMetal Editor: Enrique
Vega