Mira Kostic eventually left Saginaw to teach at Lawrence Tech. But the calculator lived on. Well into the 1980s, old-timers would pull yellowed Saginaw Thermal Calculators from their toolbox lids, ignoring the new digital infrared guns. “Batteries die,” they’d say, spinning the cardboard disk. “This never does.”
By aligning the part’s “minimum section thickness” with its “mass,” and reading across to “time since quench,” a line operator could instantly estimate the core temperature to within ±15°F. No electronics. No batteries. Just laminated cardboard, brass rivets, and a clear plastic cursor. saginaw thermal calculator
Within six months, scrap rates from thermal cracking dropped 43%. Dutch had the tool laminated in greaseproof plastic and chained to every quench tank. Mira’s design was so effective that the plant manager sent copies to GM’s Hydra-matic and Detroit Diesel divisions. By 1962, over 2,000 Saginaw Thermal Calculators were in use across the Midwest. Mira Kostic eventually left Saginaw to teach at
The story took a twist in 1965. A quality auditor noticed that Mira’s formula consistently overpredicted cooling for hollow shafts. She went back to the data, found a second-order boundary layer effect, and issued a — a small correction table printed on the back. Operators grumbled about flipping the card, but the new accuracy caught a latent problem: an oil quench tank that had been slowly contaminated with water. That discovery alone saved a $250,000 recall. No batteries
[ T_{core}(t) = T_{furnace} - \left( \frac{k \cdot t}{ (V/A)^{0.85} } \right) ]
Then a junior process engineer named Mira Kostic did something unexpected. She asked for a slide rule, a pad of graph paper, and three weeks of logged cooling curves from a dozen part geometries. Management thought she was wasting time. Dutch gave her the green light anyway.