01Executive summary
Four milestones in metallurgy and machine tools without which the steam engine would have remained a curiosity — and one of them was driven by an environmental crisis.
Abraham Darby smelted iron with coke at commercial scale at Coalbrookdale from 1709, breaking the industry's dependence on charcoal and therefore on forests. John Wilkinson's boring mill of 1774 made a cylinder round enough for Watt's engine to work. Henry Cort's grooved rolls of 1783 and puddling furnace of 1784 produced wrought iron directly from pig iron with no charcoal at all. Joseph Hall's oxide-lined furnace of 1838 made the metal markedly better.
02Coke smelting: a resource crisis forces a substitution
Iron making had been an art for centuries and a good craftsman could produce reliable metal in small quantities. Workers understood the difference between wrought and cast iron, knew steel, and knew that some ores worked better than others without knowing that sulphur and phosphorus were the reason. And they knew that no fuel was as good as charcoal.
That last fact was the problem. Wherever the industry took hold, charcoal burning destroyed the forests. In Britain this was not an aesthetic concern: admirals and statesmen were alarmed for the Royal Navy and for naval supremacy if the oak went. The pressure to find another fuel came from timber scarcity, not from any deficiency in charcoal as a reductant.
Coke was not adopted because it was better. It was adopted because charcoal was destroying a strategic resource, and the substitute had to be made to work. This is the shape of a great many material transitions, and it is worth naming: the replacement is frequently inferior at first and wins because the incumbent has become unavailable or unacceptable. Anyone assessing a substitution purely on technical merit is measuring the wrong thing.
Earlier attempts failed. A patent went to Simon Sturtevant and then to John Rovenzon, who seems to have done little beyond writing about it and coining the term "pig" iron from the appearance of the metal running from furnace into mould and branches. Dud Dudley claimed from 1665 to have made good merchantable iron with coal thirty to forty per cent cheaper than charcoal iron, in the face of hostility from persons who on one occasion cut his bellows. He never set out how he converted coal to coke or reduced ore with it, and the consensus among experts is that he was a wishful thinker who never achieved what he claimed, working at it doggedly through illness and poverty until his death in 1684.
Thirty years later the first Abraham Darby succeeded. By 1713 he and his son-in-law Richard Ford were producing five to ten tonnes a week at Coalbrookdale on the Severn, using coke to supplement charcoal, and could make their cast iron hard and brittle or soft and tough as they chose. Commercial coke-smelted cast iron began the year after Newcomen's engine reached the market, which is a coincidence of timing that shaped the century: steam raised the demand for iron, and iron made the engines possible.
Stronger air blasts brought higher furnace temperatures, releasing impurities in both fuel and ore and making coke effective without charcoal at all. By 1745 the second Abraham Darby and Richard Reynolds were using coke regularly for a superior cast iron. By 1775 the third Abraham Darby, with John Wilkinson, had improved the coke furnace enough to cast the arched ribs for the first iron bridge.
03Wilkinson's boring mill: the tool that made the engine
Newcomen's early cylinders were not bored. They were rough castings, polished by hand, with much friction and lost power, and variations in diameter that caused the piston to bind or the vacuum to fail. Brass cylinders could be made thin but grew expensive as they grew larger. Cast iron cost far less and had to be made thick — too thick for the rapid temperature cycling the Newcomen process required.
John Wilkinson patented a cylinder-boring mill in 1774 that introduced the guide principle to machine tools: the cutting head is carried on a guided bar rather than following the bore it is cutting, so errors are not reproduced. It was considerably better than Smeaton's and, in plain terms, indispensable to Boulton and Watt. With it, Watt could guarantee a cylinder of 1.8 metres diameter "not farther distant from absolute truth than the thickness of a thin sixpence at the worst part" — roughly four tenths of a millimetre.
The separate condenser is famous and the boring mill is not, and the engine does not work without both. A cylinder that is not round leaks past the piston, and an engine relying on a vacuum cannot tolerate leakage. Watt's improvement was unusable until someone could make the part. Every series in this set contains one of these: photolithography behind the integrated circuit, the pressure exchanger behind desalination, full-scale fatigue testing behind the modern airframe. When assessing why an idea did or did not succeed, look for the tool that had to exist first.
Wilkinson, one of the first British millionaires, contributed more than the mill. He ordered one of the first Watt engines in 1775, pioneered applying steam power to rolling iron, and in 1787 launched the first iron vessel — a barge some 21 metres long built of riveted iron plate.
04Cort: rolling and puddling
Coke smelting gave cast iron. Cast iron is strong in compression, weak and brittle in tension. Wrought iron is tough and works in tension, and converting pig iron into it still required charcoal and heavy hand labour. Henry Cort removed both requirements within two years.
| Patent | Date | Mechanism | Result |
|---|---|---|---|
| Grooved rolls | 1783 | The heated metal is squeezed between shaped rolls rather than only hammered | Impurities expelled and the metal formed directly into plates, bars and rods |
| Puddling furnace | 1784 | A reverberatory furnace in which flame and gas swirl over molten pig iron resting on sand, so metal never touches fuel | Wrought iron produced from pig iron with coke and no charcoal at any stage |
| Boiling process | 1838 | Joseph Hall lines the furnace with iron oxide; oxygen combines with carbon, raising the melting point so the metal becomes pasty | Slag separates far better and phosphorus and silicon pass into it — a much superior metal |
The key idea in puddling is separating the metal from the fuel. In a direct-fired process the charge absorbs impurities from whatever is burning next to it, which is precisely why charcoal was required — it is clean. A reverberatory furnace heats by radiation and hot gas from a separate hearth, so a dirty cheap fuel can be used without contaminating the product. Cort's rolls then let him cut, pile, hammer and re-roll the spongy mass, working slag out as he went.
It is worth noting the sequencing. Cort tried to get a Boulton and Watt engine in 1779 and could not, because the engine had not yet been adapted to heavy machinery of that kind — Watt had not perfected rotary motion. Cort continued on water power. The technologies in this series were arriving close enough together that one was frequently waiting on another.
Like Dudley before him, Cort was opposed by ironmasters of the old school. He was also deceived by his partner Adam Jellicoe, whose stake in the business was money stolen from the Navy while a deputy paymaster. When Jellicoe died suddenly in 1789 the investigation stripped Cort of his patents and forced him into bankruptcy, losing property worth around £25,000. He sought government re-employment repeatedly over the next decade, received a pension of £160 a year, and died in 1800 disgraced. The value of his contribution was never in question. It is a reminder that the historical record of who profited from an innovation is a very poor guide to who produced it.
The strategic consequence was immediate and was recognised at the time. By 1788 British production exceeded importation, with two-thirds of domestic supply smelted with coke. Writing after the loss of the American colonies, Lord Sheffield argued that Cort's improvements together with the Boulton and Watt engine would prove more advantageous to Britain than the thirteen colonies had been. Whatever one makes of that as consolation, as a prediction about where industrial power was about to come from it was shrewd.
05Takeaways for current practice
- Substitutions are usually forced by scarcity, not chosen for merit. Coke won because the forests were going.
- Look for the tool that had to exist first. The separate condenser is useless without a mill that can bore a round cylinder.
- Separate the product from the contaminant. Reverberatory heating let a dirty fuel make clean iron, which is the whole of puddling.
- Guide the cutter rather than letting it follow the work. The guide principle stops a machine tool reproducing its own errors, and it generalises to any feedback path.
- Who profited is a poor guide to who contributed. Cort died bankrupt and disgraced with his contribution unquestioned.