Preheat temperature

Preheat temperature is a floor, not a target: the parent material must reach it and hold it before the arc starts, and it stays relevant for the whole joint because interpass temperature — the same measurement taken between passes — is usually bound by the same minimum. Two separate mechanisms justify the requirement, and they explain why the number is not simply “hotter is safer.”

Slowing the cooling rate is the metallurgical reason. A cold plate is an efficient heat sink; raising its starting temperature slows how fast the weld and the surrounding heat-affected zone pass through the transformation range, favouring tougher microstructures over hard, crack-susceptible ones. Giving hydrogen time to diffuse out is the second reason, independent of the first: diffusible hydrogen picked up from moisture in the atmosphere, the consumable, or surface contamination needs time and temperature to escape the weld metal before it can concentrate at a hardened, highly stressed location and initiate hydrogen cracking.

The required value is not read off a single table entry. It is a function of the parent material’s hardenability — normally expressed as a carbon equivalent — of section thickness, of the diffusible hydrogen level the consumable is expected to contribute, and of the heat input the procedure will run. EN 1011-2 gives two routes to the number: a set of nomograms, and a closed equation covering the same variables. Both convert a composition and a set of welding conditions into a temperature; neither converts a temperature into a guarantee, and both assume the input values used were the real ones for that heat of material, not a nominal certificate value carried over from a previous job.

Measurement is where compliance actually lives or dies. A contact thermometer or a temperature-indicating crayon checks one point at one instant, chosen by whoever is holding it, and production pace creates every incentive to check the easiest spot rather than the coldest one. A joint that reads compliant at the accessible edge can still be below the minimum at its centre. Non-contact thermal measurement across the full joint removes the “which spot” question, at the cost of needing a known emissivity for the surface being read before the number can be trusted as an absolute temperature rather than a relative signal.