Chapter 2 of 10 · 10 min read
The fire and its fuel
Understand why the blast makes a forge hot, where the fire oxidises and where it reduces, how much heat a welding heat needs and how little of the fuel reaches the bar, and how fuels compare.
A forge weld begins in the fire, and a bad fire spoils more welds than a bad hammer. The fire must bring the stock to welding heat evenly, quickly, without burning it and without coating it in scale or sulphur. The old manuals spent pages on the fire because their readers’ welds depended on it. The Davis-Bournonville course of 1919 began its whole course with the question of why a blacksmith’s forge, blown by a fan, is so much hotter than a fire on the ground, and Cravens’s treatise of 1921 put the same point as a principle: the amount of heat depends on how much fuel is burned, but the temperature depends on how fast it burns.
This chapter explains why the blast makes the fire hot, how a deep coke fire forms an oxidising zone and a reducing zone, how much heat a welding heat needs and how much fuel supplies it, and how the common forge fuels compare.
Heat and temperature
Burning fuel releases a fixed amount of heat for each kilogram burned. How hot the fire gets depends on how fast that heat is released into how small a space, compared with how fast the fire loses heat to its surroundings. A fire on the ground, fed only by air drifting in, burns slowly and loses heat as fast as it makes it, so it stays at a modest temperature. A forge fire, with air blown through it from beneath, burns the same fuel many times faster in a compact bed, and the heat released in the bed far outruns the losses, so the bed’s temperature climbs until it can bring iron to a white heat.
That is the answer to the Davis-Bournonville course’s question, and it is the principle behind every torch and furnace that followed. Cravens drew the lesson that it applied to every welding process, not just the forge. A larger flame is not necessarily a hotter one; a faster-burning one is.
The blast and the tuyere
The forge’s air enters through the tuyere, an opening in the bottom of the hearth, fed by a hand bellows, a hand-cranked blower or an electric fan. The blast must be enough to make the fire hot but not so much that it blows cold air through the bed. Cravens warned that too much air chills the fire or blows it out, and that excess oxygen attacks the iron and forms scale. A smith controls the blast with a gate on the air supply and adjusts it through the heat: strong to bring the fire up, steadier while the bar soaks, and enough at the end to reach welding heat without driving oxygen onto the bar.
The zones of a deep fire
A deep coke fire is not uniform. Air from the tuyere meets the hot coke at the bottom of the bed and burns it fully to carbon dioxide, releasing the most heat; that zone, just above the tuyere, is hot but full of oxygen, an oxidising zone. As the carbon dioxide rises through more hot coke, it reacts with the carbon to form carbon monoxide, absorbing heat and using up the remaining oxygen; the zone above is slightly cooler but contains carbon monoxide and almost no oxygen, a reducing zone. Hart explained in 1910 that the carbon monoxide and the coke of a smith’s fire are reducing in action, preventing the heated iron from oxidising and even cleaning scale from it.
| Where | Atmosphere | Effect on the bar | |
|---|---|---|---|
| Oxidising zone | just above the tuyere | oxygen and carbon dioxide | heats fast but scales the bar heavily |
| Reducing zone | higher in the bed, under a cover of coke | carbon monoxide, little oxygen | heats with little scale; where the bar should sit |
| Top of the fire | the surface and above | air drawn in from the room | cools and scales a bar left exposed |
The practical rule follows: the bar sits in the reducing zone, with plenty of hot coke under it to burn the oxygen out of the blast and a cover of coke over it to keep the room’s air off. A bar pushed down into the oxidising zone in front of the tuyere heats fast but scales badly and may burn; a bar lying on the surface of a shallow fire is heated from below and oxidised from above. A deep, clean, compact fire is the smith’s first tool.
Heat for a welding heat
The heat a bar needs to reach welding heat can be estimated. Steel absorbs a little under a megajoule per kilogram to go from room temperature to about 1,300 °C, including the energy it absorbs as its crystal structure changes on the way.
- Heated length
- 25 × 25 × 150 mm, about 0.74 kg of steel
- Heat to welding heat
- about 0.74 × 0.9 ≈ 0.66 MJ
- Coke
- about 29 MJ/kg; propane about 46 MJ/kg
- coke fire: 0.66 ÷ (29 × 0.04) ≈ 0.57 kg of coke
- gas forge: 0.66 ÷ (46 × 0.12) ≈ 0.12 kg of propane
- most of the fuel's heat warms the hearth, the air and the room, not the bar
The efficiency of an open forge is very low. The fire heats its own bed, the hearth, the surrounding air and everything it can see; the bar takes a few per cent of the heat. An enclosed gas forge, lined with refractory insulation, keeps more of the heat in and reaches a higher efficiency, though still a small one. Induction heating, which puts the heat directly into the bar, is the most efficient of all, but its equipment is costly and suits repeated work on similar sections.
Fuel per hour
A forge burns fuel all the time it is alight, not just while a bar is in it. An open coke fire banked between heats still burns a few hundred grams an hour; a gas forge left running burns its full rate. The shop’s fuel bill therefore depends as much on how long the fire is lit as on how many welds are made.
The differences are not large for a small shop, and fuel is rarely its biggest cost. Labour is. What matters more is how quickly each forge brings work to heat, how easily the smith can reach welding heat in it, and how clean the heat is, because a forge that wastes a smith’s time or spoils welds costs far more than its fuel.
Fuels compared
Each fuel has its character. Cravens judged charcoal the best, because it is almost free of impurities, and coal and coke good if they are low in sulphur and phosphorus. Hart preferred coke for smithing, because the coke fire is cleanly reducing, and warned that soft coal usually carries high sulphur, which the hot iron absorbs to its harm, and smoky flames that leave soot on the work.
| Strengths | Drawbacks | |
|---|---|---|
| Charcoal | very clean; low sulphur; reducing | burns fast; bulky; sparks; costly per heat |
| Coke | clean, deep, reducing fire; steady | must be lit with a starter fire; ash and clinker |
| Bituminous coal | forms its own coke in the fire; widely used | smoke; sulphur in poorer grades; needs managing |
| Propane or natural gas | clean; steady; easy to control; no clinker | atmosphere depends on burner tuning; hard to reach full welding heat in some designs |
| Induction | fast, precise, no combustion | costly equipment; suits repeated sections |
Sulphur and other poisons
Sulphur in the fire is absorbed by hot iron and makes it hot-short: brittle at red heat, so that it cracks under the hammer. Cravens added that phosphorus makes iron cold-short, brittle when cold, and warned that copper, brass, lead, tin or any other non-ferrous metal must be kept out of the forge, because they spoil iron for welding. A coin or a scrap of brass dropped into the fire can ruin a weld made in it later. The smith keeps the fire clean, removes clinker regularly, and does not burn rubbish in it.
Clinker
Coke and coal contain ash, which melts in the hottest part of the fire and runs together into clinker, a glassy lump that forms over the tuyere and chokes the blast. A fire with clinker in it heats unevenly and is hard to bring to welding heat. The smith lets the fire burn down a little, lifts the clinker out with a poker while it is still soft, and rebuilds the fire before a welding heat. Clean, low-ash coke makes less clinker, which is one reason fuel quality matters.
Gas forges and their atmosphere
Gas forges heat by burning gas with air in a burner and blowing the flame into an insulated chamber. They are clean, quiet and easy to control, and many smiths now use nothing else. Their atmosphere depends on the burner’s mixture. With too much air the chamber is oxidising and scales the work; with slightly too much gas it is reducing and the work stays clean. Hart’s advice for gas flames for welding was to adjust them until a small test piece of steel heated in the flame stayed clean, neither scaling nor sooting. Many gas forges struggle to reach full welding heat on larger stock, and smiths who weld often choose burners and chambers designed for it.
The shop’s carbon monoxide
A forge fire, especially the reducing zone, makes carbon monoxide, a colourless, odourless gas that causes headache, dizziness and, at higher levels, unconsciousness and death. Coal and coke forges must have hoods and flues that carry the gases out of the shop; gas forges, which usually vent into the room, need good ventilation, and many smiths fit carbon monoxide alarms. Enclosed shops in winter, with doors shut against the cold, are where the risk is greatest. Chapter 10 deals with the forge’s hazards together.
Heating evenly
A bar must be at welding heat right through the part that will be welded, not just on its surface. Thick stock heats from the outside in, and its centre lags behind; heated too fast, its surface reaches welding heat or burns while its core is still red. The smith brings large work up gradually, turning it in the fire, and lets it soak so that the temperature evens out. Both pieces of a weld must reach welding heat together, which means putting the larger piece in first or placing it deeper in the fire.
Sizing the fire to the work
A fire must be large enough for the work it heats. A small, deep fire suits small stock, bringing a short length to heat quickly with little waste; large stock needs a wider, deeper fire so that the whole of the weld area sits in the reducing zone at once. Cravens noted that large pieces take longer to heat and stay hot longer than small ones, and that their surfaces flake off as scale when exposed. A smith welding heavy work builds the fire up well in advance, lets it burn through to clean coke, and may use a helper on the blast so that the heat can be brought up steadily without pausing. A fire too small for the work heats the bar unevenly, leaves part of the joint cold, and tempts the smith to force it with the blast, driving oxygen onto the bar.
The running case: changing fires
This is an illustration. The shop had worked with a coal forge for years, and the owner noticed that welds in some batches of coal were poorer, with more cracking under the hammer. The supplier’s coal had changed, and a test of the new coal showed higher sulphur. The shop switched to a low-sulphur coke for its welding fire and kept a gas forge for general heating. The change in weld quality was immediate, and the coke’s clean fire also cut the time spent clearing clinker. The owner now test welds a sample from each new delivery before using it for class work.
A failure: the shallow fire
This is an illustration. A student in a class, impatient to get started, built a shallow fire with little coke over the tuyere and laid the bars across the top. The bars came up to colour quickly, but they were heavily scaled on top and, when the student turned the blast up to finish the heat, one began to spark and burn at its tip.
The bars had been heated in the oxidising zone from below and exposed to the room’s air above. Rebuilt into a deep fire, with the bars buried in the reducing zone under a cover of coke, the next pair came to welding heat with little scale, and the weld took. The owner keeps a cut-away sketch of the fire’s zones on the shop wall for exactly that lesson.
Exercise: fuel for a day of welding
A class makes 24 scarf welds in 20 mm round mild steel, each heating about 120 mm of each of the two pieces to welding heat. Taking about 0.9 MJ/kg to reach welding heat, estimate the heat needed per weld and for the day, and the coke used at a forge efficiency of 4%. If the fire burns about 2.5 kg of coke an hour over a 6-hour class regardless, which figure dominates?
Each piece heats about π ÷ 4 × 20² × 120 ≈ 37,700 mm³, about 0.30 kg, so each weld heats about 0.59 kg of steel, needing about 0.59 × 0.9 ≈ 0.53 MJ. Twenty-four welds need about 12.8 MJ.
At 4% efficiency, that needs about 12.8 ÷ (29 × 0.04) ≈ 11 kg of coke. The fire burns about 2.5 × 6 = 15 kg over the class whatever is done, so the fire’s running time dominates, and the coke burned directly for the welds is already counted within it. Most of the day’s fuel keeps the fire alight; planning work so that the fire is used continuously, not left burning between demonstrations, is the way to save it.
