French electrical generator rotor forging failure

4–5-inch-thick cast-iron generator housings the size of a small house — the size-extreme example for sand casting. ## Figures referenced (recurring numeric anchors, not cases) - Steel mill yield improvement: 65% → 90% → 97% with continuous casting (`§7.p3`). - $6,000 single-crystal turbine blade unit cost (`§4.p6`, `§4.p9`). - 100 ounces/year world iridium production (`§5.p6`, `§7.p18`). - ~50 years of unsuccessful continuous-cast steel-strip research (`§7.p18`). - 75% area reduction as minimum for breaking down macrosegregation in heavy forgings; 90% preferred (`§8.p16`–`§8.p19`). - $6M wire contract, $300M overall project, $2M savings claim — trolley wire forensic case (`§9.p6`–`§9.p11`).

You also can make big cast-iron parts the size of this room. If I have a generator rotor forging — we talked about generator rotor forgings yesterday — I have to have a housing to put it in. Once I take that steel rotor and wind it with copper, I need a housing to hold the other windings to make my big generator. That generator housing is about four or five inches thick cast iron, the size of a small house. There's a top and a bottom, and it's a big sand casting — just bigger than the little nipples you buy at the hardware store.

Cited as supporting precedent for the molybdate water-treatment failure mode. Same type of high-strength steel, same type of cooling water treatment, same stress corrosion cracking initiated from pitting — a documented French utility failure that should have been in the inspection company's literature review.

Yeah. They started checking their water treatment. The company that did the water treatment so it wouldn't corrode, they were sort of on the hook too. They were not supposed to put things in there. They put molybdates [?] in the water, which some French researchers had shown — for the same type of steel and a great big generator rotor forging — they had a failure in France of a great big generator rotor forging in a utility, an electrical generating plant. It had stress corrosion cracking in the same type of steel with the same type of water treatment, because of pitting. So if you started looking back in the literature of the previous five or ten years, yeah, the water treatment they were using was sort of out of date. And they hadn't done inspections — the inspection company was sort of on the hook too, and they were worth about a million dollars. Hey, just throw your money in. They didn't have enough money to be held responsible.