Bolting the Cylinder Head

The cylinder head for the Great Western was returned from the foundry on the fourth of November, 1837, and I had it bolted to the bed of the boring mill by noon. We had specified a wall thickness of one and three-quarter inches. The first observer - the patternmaker, the moulder, the man who pours - reads the drawing and gives you a casting that matches the drawing. That is the first observer’s whole duty, and he discharged it. The second observer is the one who cuts into the metal.

I set the boring bar and took a first cut of three-eighths. At fourteen inches from the flange the tool broke through into a blowhole the size of a pea. A pea-sized void does not appear on any drawing. It appears when the iron cools unevenly around a core that shifted, and it sits there, invisible, until the boring bar finds it. The first observer certified the casting against the pattern. The pattern was correct. The casting was not.

Here is the question the student always asks: why not inspect the casting before boring? Because the defect is interior. No lamp, no caliper, no gauge reads a void that has not yet been opened. The only instrument that finds it is the cut that destroys the surface. The second observer does not confirm the first observer’s work. The second observer is the load test.

The first observer tells you what was intended. The second tells you what is there. Design is a claim about the future. Machining is a measurement of the present. I will not accept a cylinder head on the patternmaker’s certificate again. I am ordering every future casting sectioned at the foundry before it leaves the yard.