There is a number in Campbell's handbook that rearranges how you see every foundry you will ever walk through. Liquid aluminium must not fall much further than about twelve millimetres – roughly the height of a small coin stood on edge – before it passes half a metre per second and begins to damage itself. Above that speed the melt's surface folds over and swallows its own skin. Foundries, meanwhile, have spent the better part of a century pouring metal from half a metre, a metre, sometimes more, and then wondering why castings leak, crack and fail their fatigue tests. Campbell spent a career pointing at the falling stream and explaining, with the patience of a man not always listened to, that the industry was manufacturing its own defects.
The mechanism is almost embarrassingly simple once you have seen it. Every liquid metal wears a skin. Aluminium in particular clothes itself in oxide the instant it meets air, and a quiet, gently rising melt keeps that skin where it belongs, on the outside. A turbulent one folds it inwards – and the fold goes in dry face to dry face, two ceramic surfaces that can never weld to each other. Campbell named the result a bifilm, a crack held in suspension, invisible and waiting. As the casting solidifies, these entrained films open into porosity, feed hot tears, and hand fatigue its starting notch. Hence the book's quietly radical thesis – most of the classic defects of castings are a single defect wearing different costumes, and it was poured in at the very beginning.
Every casting carries a complete record of the way it was poured. The metal remembers.
For most of the industry's history, defects were treated like weather – regrettable, random, and nobody's fault. Campbell's life work is a long, courteous, occasionally exasperated demonstration that the weather is man-made. Fill from the bottom. Keep the melt below its critical speed. Design the running system as though the surface of the liquid were sacred, because it is. His ten rules for good castings read less like engineering guidance and more like a code of conduct, and the Cosworth process he helped develop – bottom-filled by electromagnetic pump, unhurried, almost silent – showed what the code delivers when it is actually obeyed. The deeper point is historical. A craft humanity has practised for five thousand years became a science within living memory, and the man who did much of the converting is still being argued with.
The taproot
Why should anyone outside a foundry care? Because casting sits upstream of nearly everything. By the industry's usual estimate, some nine in ten manufactured goods contain at least one cast component, almost always buried where no customer will ever see it – the engine block, the pump housing, the valve body, the gearbox casing, the machine-tool bed whose mass and damping make precision machining possible at all. A modern wind turbine swings on a ductile iron hub weighing tens of tonnes. A hip implant is very often an investment casting. And at the extreme end of the art, the blades of a jet engine are grown as single continuous crystals of superalloy and then run, hour after hour, in gas hotter than the melting point of the metal they are made from. We even name the ages of civilisation after cast metal – the oldest known casting is usually said to be a small copper frog, made in Mesopotamia more than five thousand years ago, and the Bronze Age is, at bottom, a casting technology with an empire attached.
If manufacturing is the root system that prosperity grows on, as I argued in the Venice and Korea essay, then casting is the taproot. Machining, forging and fabrication mostly begin where a casting or an ingot left off. Lose the foundries and the rest of the industrial food chain starves quietly, a supplier tier at a time – which is worth remembering in a country like Poland, where the foundries are still here, still competitive, and still treated as somewhat unfashionable.
That unfashionableness is the strangest part of the whole story. Foundries carry an old reputation – dark, dirty, dying – while the reality on a modern casting floor is applied physics and chemistry conducted at seven hundred degrees, with solidification simulation, X-ray inspection and metallurgy that would not embarrass a laboratory. The status of the industry and the sophistication of the industry parted company decades ago and have never been reintroduced. Campbell's handbook reads, among other things, as an attempt at that reintroduction – the field's declaration of intellectual dignity, nine hundred pages long.
I came to the book with an unusual reading history, because I had spent more than a decade alongside an aluminium foundry before I opened it – interpreting X-ray and NDT training, sitting with inspectors as they traced indications across radiographs, translating arguments about porosity between engineers on two sides of a language. Reading Campbell was like being handed the grammar of a language I had been interpreting for years. The shadows on those radiographs were, more often than anyone in the room quite knew, the ghosts of turbulent pours – bifilms unfurling exactly as the handbook says they must. It is a strange pleasure, years into a specialism, to be shown the sentence structure underneath the vocabulary.
So the next time you board an aircraft, start a car, or walk past a wind turbine, spare a thought for the least visible industry you depend on. Somewhere inside each of those machines is a piece of metal that remembers exactly how carefully, or carelessly, it was poured – and an old, patient, five-thousand-year-old craft that deserves rather more of civilisation's respect than it gets.