Heat input condenses the three numbers on the machine and the stopwatch — arc voltage, welding current, and travel speed — into the energy delivered per length of joint: HI = 60 × V × I ÷ (1000 × S), giving kJ/in with speed in inches per minute and kJ/mm with mm/min (the two related by the exact 25.4 mm-per-inch ratio). A GMAW pass at 25 V and 250 A covering 10 inches per minute comes to 60 × 25 × 250 / (1000 × 10) = 37.5 kJ/in, or 1.48 kJ/mm. AWS D1.1 heat-input limits and most North American WPS maximums are written against this gross arc energy directly.
European practice layers a process correction on top. EN 1011-1 defines heat input with a thermal efficiency factor k that estimates how much of the arc energy actually enters the plate: 1.0 for submerged arc (buried under flux, nearly all energy reaches the joint), 0.8 for stick, MIG/MAG, and flux-cored processes, and 0.6 for TIG, whose open arc radiates a large share of its energy away. The same 37.5 kJ/in GMAW pass is therefore a 30.0 kJ/in net heat input under the EN 1011 convention — check which convention a specification follows before comparing numbers.
The reason procedures police this number in both directions: high heat input slows cooling, coarsens grain in the heat-affected zone, drops toughness, and can soften quenched-and-tempered steels below their specified strength — the case AWS D1.1 clause 7.7 requires heat input control for. Too little heat input speeds cooling instead, raising HAZ hardness and hydrogen-cracking risk, which is why many WPSs carry a minimum as well as a maximum; the qualified window comes from the procedure qualification record.