Heat Treat Math

How to Calculate Heat Treatment Furnace Metrics: Cycle Time, Load Density, and Energy

A step-by-step walkthrough of the five calculations that govern furnace heat treatment, with worked examples in real units and clear input sources.

Start with soak time, the backbone of every recipe. The rule of thumb for through-heating steel is 1 hour per inch (25 mm) of section thickness at temperature, plus a ramp allowance. A 2 inch shaft needs roughly 2 hours at soak, but total cycle time also includes ramp: heating from 70 F to an 1550 F austenitizing setpoint at a conservative 300 F per hour adds about 5 hours. The Annealing Cycle Time calculator handles this split, but you can hand-check it: total = ramp time + soak time + any furnace recovery after door open. Always pull section thickness from the thickest part in the load, not the average.

Batch capacity determines how many parts ride one cycle, and it is pure geometry plus a packing factor. Take usable hearth area (length x width minus fixture footprint), divide by part footprint, then multiply by a 0.60 to 0.85 packing efficiency to leave gas-flow gaps. A 36 in by 48 in hearth is 1,728 sq in; parts occupying 12 sq in each give 144 theoretical positions, but at 0.70 packing you plan for about 100. Feed those numbers into the Batch Heat Treat Capacity tool. Stacking height and basket weight limits often cap you before floor area does.

Load density ties weight to the furnace's heating capability and is where over-packing quietly wrecks uniformity. Divide net charge weight by usable hearth volume to get lb per cubic foot. A load of 1,200 lb of steel in a 30 cubic foot chamber is 40 lb/cu ft. Above roughly 45 to 60 lb/cu ft for atmosphere furnaces, interior parts lag the setpoint and you must extend soak. The Furnace Load Density calculator flags this. Track it every batch, because thermocouple placement on the surface will not reveal a cold core.

Furnace energy per part starts from the theoretical heat to raise the charge, then divides by efficiency. Q = mass x specific heat x delta-T. For 1,200 lb (544 kg) of steel, specific heat near 0.12 BTU/lb-F, heated 1,480 F: Q = 1,200 x 0.12 x 1,480 = 213,000 BTU. Real furnaces run 20 to 45 percent thermal efficiency once you add wall losses, atmosphere burn-off, and fixture mass, so actual input is 213,000 / 0.30 = 710,000 BTU. Divide by parts per load for energy per part; the Furnace Energy Cost calculator converts that to therms or kWh.

Furnace utilization is a time ratio, not a money figure. Utilization = productive furnace hours / total available hours. If a furnace runs loaded and at temperature 128 hours in a 168 hour week, utilization is 76 percent. Idle-hot standby and empty ramp cycles count against you. The Furnace Utilization calculator separates loaded-hot time from idle-hot and cold, which matters because an idle-hot furnace at 1,650 F still draws 40 to 70 percent of its full-load energy just holding temperature through the refractory.

Hardness variation quantifies whether the cycle actually delivered spec across the load. Compute the range (max minus min) and standard deviation from your test coupons. If nine coupons read from 58 to 62 HRC, the range is 4 HRC and you would target a standard deviation under about 1.5 HRC for a controlled load. The Hardness Variation calculator does the statistics; feed it Rockwell readings taken at corner, center, and top positions. Wide spread points back to load density or thermocouple lag, closing the loop with the earlier calculations.

Tie the chain together with one worked load. Suppose 100 gears, 12 lb each, austenitized at 1,550 F. Charge weight is 1,200 lb; at 40 lb/cu ft in a 30 cu ft chamber, density is fine. Soak for the 1.5 in tooth section is about 1.5 hours; ramp adds 5 hours, so cycle time is roughly 6.5 hours. Energy is 710,000 BTU, or 7,100 BTU per part before quench and temper. Run each stage through its calculator and the numbers reconcile, giving you a defensible process sheet rather than a guess.

One caution on inputs: delta-T uses the true starting metal temperature, not room air. Parts pulled from a cold-soaked warehouse in winter may enter at 40 F, adding measurable load. Specific heat also rises with temperature, so for high-precision energy work use a mean value near 0.14 BTU/lb-F above 1,000 F rather than the room-temperature 0.11. Getting these two inputs right shifts calculated energy by 10 to 15 percent, which is the difference between a cycle that hits temperature on schedule and one that runs long every time.

Published 2026-07-01.