Preheat is one of the oldest controls in structural welding practice. Raise the base metal temperature before striking an arc, and you slow the cooling rate of the HAZ — which reduces hydrogen cracking susceptibility, lowers residual stress, and improves HAZ toughness. AWS D1.1:2025 encodes minimum preheat requirements based on base metal chemistry (carbon equivalent), thickness, and process.
But specifying a preheat temperature on the WPS and actually achieving it in the field are different problems. The gap between what the WPS says and what the welder actually does — and what the CWI can verify and record — is where preheat nonconformances live.
This article covers how preheat is measured, where it must be measured, which tools are acceptable, and what a defensible preheat record looks like.
Rule library based on AWS D1.1:2025; verify against your governing edition.
Why measurement location matters
A base metal heated with a propane torch does not heat uniformly. The surface directly in front of the flame is hotter than the base metal 6 inches away. The through-thickness temperature gradient depends on heating time, section thickness, and heat source.
AWS D1.1:2025 requires preheat measurement on the base metal surface at a minimum of 3 inches (75 mm) from each edge of the weld joint. This location is chosen deliberately:
- It is away from the direct flame impingement point, which would give an artificially high reading
- It is within the zone that will be influenced by the heat of welding
- It approximates the temperature of the base metal at the HAZ before welding starts
A measurement taken at 1 inch from the weld edge, or worse, inside the joint groove immediately after heating, will read higher than the 3-inch measurement. That higher number may look like the preheat is achieved when the through-thickness temperature at 3 inches — the condition AWS D1.1 is actually requiring — has not been reached.
On heavy sections (2 inches and thicker), through-thickness temperature equilibration is the critical issue. The surface temperature reaches the minimum preheat temperature well before the interior of the base metal does. AWS D1.1:2025 allows sufficient time for the heat to soak in — in practice, this means maintaining the heat source long enough that the surface temperature at the required measurement location has been stable for several minutes. A CWI who verifies temperature while the welder's torch is still pointed at the base metal is not verifying the soak-through condition.
Temperature measurement tools
Temperature-indicating crayons (Tempilstik and equivalents)
Tempilstik crayons are wax-based sticks manufactured to melt at a specific temperature within a rated tolerance. They are the most common field tool for preheat verification. Method:
- Remove the heat source from the joint area.
- Draw a mark with the crayon at the required measurement location (3 inches from the joint edge).
- If the mark melts, the base metal surface is at or above the crayon's rated temperature.
- If the mark does not melt or smears but doesn't fully liquefy, the temperature has not been reached.
Tempilstik crayons are a pass/fail indicator. They do not give a continuous temperature reading. A shop using crayons must stock the specific temperature corresponding to the required minimum preheat — and the crayon temperature should be one calibration step below the required minimum, not above, to confirm the minimum has been achieved.
Crayons are inexpensive and require no calibration. They do not inherently generate a paper record — a CWI must note the observation in the inspection record manually.
Contact (probe) thermometers
Contact thermometers with a metallic probe give a direct surface temperature reading. They are more precise than crayons, faster to read, and generate a numerical value that can be recorded directly. A CWI carrying a calibrated digital contact thermometer can take multiple readings across the joint area, document each one, and produce a record that shows spatial uniformity of the preheat.
The probe must make solid contact with the base metal surface. A probe pressed lightly against rough mill scale will read lower than the actual surface temperature. Some inspectors lightly grind or clean the measurement spot before contact measurement on heavily scaled surfaces.
Calibration of the thermometer is a quality-system requirement. A contact thermometer that has not been calibrated within the interval specified by the shop's quality plan — typically annually — does not produce a defensible record.
Infrared (IR) thermometers and pyrometers
Infrared thermometers measure surface temperature without contact by detecting emitted thermal radiation. They are fast and can measure temperature at locations that are difficult to probe. However, their accuracy depends heavily on the emissivity of the surface. Mill scale, bare steel, and oxidized steel have different emissivities. An IR thermometer calibrated for one surface condition will read incorrectly on another.
AWS D1.1:2025 does not prohibit IR thermometers, but CWIs using them must account for emissivity correction. Some IR tools allow emissivity adjustment; without that adjustment, the reading may be meaningfully off. Using an IR thermometer on bare machined steel without emissivity correction can underread surface temperature by 50–100°F or more at typical preheat temperatures.
For code-compliance verification, contact measurement or temperature-indicating crayons are more defensible and easier to explain to an AHJ or owner's representative than an IR reading with a complicated emissivity discussion.
Thermocouples
For production environments where continuous temperature monitoring is required — large-diameter pipe welds, heavy section connections, high-restraint joints requiring extended preheat and controlled post-weld cooling — thermocouples attached to the base metal give a continuous record. This is less common in structural fab than in pressure vessel or pipe fabrication, but it appears on critical projects with specific temperature control requirements.