Preheat is the single most audited variable at the weld joint before the arc strikes. Getting it right protects against hydrogen-induced cracking (HIC) in the heat-affected zone (HAZ), and documenting it correctly keeps the WPS defensible when an inspector or owner's engineer reviews the package. Yet field preheat practice varies widely — the same fab shop may run three different measurement tools across shifts with no written method.
This article covers what AWS D1.1:2025 requires, what the commonly used measurement tools actually measure, and how to build a defensible documentation chain that survives an AISC or third-party audit.
What AWS D1.1:2025 actually requires
The standard establishes minimum preheat requirements based on base metal specification, carbon equivalent, thickness, and filler metal hydrogen designation. Those code-minimum values feed directly into the WPS — the WPS minimum preheat is at least as high as the code minimum, and may be higher if the PQR or engineering judgment requires it.
Two spatial requirements define where the preheat must exist:
Zone width: Minimum temperature must be present in the base metal within 3 in [75 mm] from the point of welding in every direction — not just at the groove face. On a thick wide-flange column splice this means heating the flange 3 inches back from the CJP, not just warming the groove with a torch for 30 seconds and touching the face.
Measurement timing: Temperature must be verified immediately before striking the arc. If there is a delay — welder stops for filler change, break, equipment issue — the temperature must be re-checked before resuming. A measurement taken 20 minutes before welding begins tells you the temperature 20 minutes ago, which is not what the standard requires.
Rule library based on AWS D1.1:2025; verify against your governing edition.
Temperature-indicating crayons (temp sticks)
Temp sticks — sold under various trade names, most recognized as "Tempilstik" — are the most common field verification method in structural shops. Each crayon is rated to melt at a specific temperature, typically within ±1% of rating. The technique: mark the base metal at the measurement location with the crayon before or during heating, and observe whether the mark liquefies.
What they measure well: The surface temperature of the metal at the moment of application. They're simple, require no calibration, leave a visible record, and cost essentially nothing.
Where they fail:
- They indicate whether the surface reached the rated temperature; they don't confirm the temperature is maintained at depth (through-thickness). On heavy plate (1½ in [38 mm] and above), surface temperature can read adequate while the interior is still cold — the HAZ you care about.
- A mark applied after the metal has already cooled past the crayon's melt point will not liquefy at a lower temperature. The welder must apply the mark while temperature is rising.
- Marking the groove face with a crayon introduces carbon and wax contaminant into the weld zone if not ground out — always verify 3 in [75 mm] away from the joint, not at the joint itself.
Best practice: use the next-size-down crayon as a "floor" indicator, and the target preheat crayon as the "pass" mark. If 300°F crayon melts and 350°F does not, you have bounded the surface temperature in a range.
Contact thermometers (thermocouple-based pyrometers)
A contact pyrometer — a thermocouple probe pressed against the base metal surface — gives a direct temperature reading in seconds. Modern digital units read to ±2°F or ±2°C across the range relevant to structural welding (150°F–700°F [65°C–370°C]).
Advantages: Instantaneous digital read, direct display, calibrated to NIST-traceable standards on most quality instruments, records can be logged electronically.
Limitations: The probe tip must be held firmly against the metal — a loose or angled contact reading will be lower than actual temperature. On rough mill-scale surfaces or uneven welds, getting consistent contact can be tricky. Probe tips wear and calibration drift over time; a pyrometer that hasn't been calibration-checked in two years is no longer a valid reference.
For AISC-certified shops or those supplying prequalified WPS packages to owners with strict QC requirements, contact pyrometers with documented calibration intervals are generally preferred over temp sticks for formal hold-point records.
Infrared (IR) thermometers
Non-contact IR thermometers read thermal radiation emitted by the surface and calculate temperature from emissivity — the fraction of radiation a surface emits relative to an ideal blackbody. This is where field use frequently goes wrong.
Emissivity matters enormously:
- Hot-rolled mill scale: emissivity 0.85–0.95 → IR thermometers read accurately on this surface.
- Freshly ground or machined bright steel: emissivity 0.25–0.45 → at 300°F actual temperature, an IR unit set for 0.90 may display 150°F or less. This is a significant false-low reading that could lead a welder to believe preheat is inadequate and over-heat the steel.
- Painted or primed surfaces: emissivity varies by coating. Verify the coating manufacturer's data before relying on IR measurements.
When using an IR thermometer, verify the instrument emissivity setting matches the surface condition. If uncertain, default to a contact thermometer or temp stick for the critical preheat verification and use IR for monitoring during heating only.
IR tools are fast and non-contact, which is useful for sweeping a large preheat zone to confirm uniform heating. Use them as a screening pass, then confirm the minimum at the critical measurement point with a contact or crayon method.
Building a defensible documentation chain
A CWI sign-off on preheat should include at minimum:
- The WPS minimum preheat requirement (as-specified)
- The measured temperature and the location where it was taken
- The instrument type (and calibration sticker/date for contact pyrometer)
- Time of measurement relative to arc initiation
- Inspector's initials or signature
On projects with an ITP (Inspection and Test Plan) or ITL (Inspection Test Level) hold-point structure, preheat is typically a Hold or Witness point — meaning production cannot advance past first pass without inspector sign-off. For that sign-off to mean anything, it needs a real temperature record, not just a checked box.
Some shops run a pre-weld checklist embedded in the traveler package. Others use a dedicated preheat log per joint or per weld sequence. Either approach works as long as the record is traceable to the joint and the date. See CWI weld inspection hold points for a broader look at ITP structure.
Preheat on the WPS: what to document
Under AWS D1.1:2025, minimum preheat is a non-essential variable for most processes — increasing preheat above the minimum stated does not require a new PQR. But the WPS must still state the minimum. That minimum must be at least as high as the applicable code table value for the material, thickness, and process combination.
If the WPS states 200°F [95°C] minimum preheat and field verification shows 175°F [80°C] immediately before the arc, that is a WPS nonconformance, regardless of essential-variable classification. The CWI has authority — and obligation — to stop welding until the correct temperature is achieved and re-verified.
For a broader look at how WPS essential and non-essential variables interact with your PQR coverage, see WPS essential variables vs non-essential and AWS D1.1 Table 6.6 explained.
Common field failures and how to catch them
Torch chasing: Welder heats with a torch while holding the temp stick directly in the flame or immediately adjacent — reads falsely high because the flame is 2000°F+ and the stick melts from radiant heat, not base metal conduction. The stick should contact the base metal at the measurement location with the heat source removed or at a distance.
Surface-only heating on heavy plate: Oxy-acetylene rosebud or propane wand can bring a thick plate surface to 300°F in minutes while the interior is 70°F. For sections over 1½ in [38 mm], soak time after heating is critical — allow the surface and interior to equilibrate before measuring. Heating too fast then immediately reading is a common way to get a compliant surface reading on a non-compliant through-thickness condition.
Checking away from the joint: A temperature reading 8 in [200 mm] from the groove face meets neither the 3 in [75 mm] zone requirement nor gives meaningful data about the temperature at the weld zone.
No re-check after delay: A joint properly preheated at 7:30 AM, left while the crew broke for 45 minutes for a safety meeting, and then welded at 8:15 AM without re-verification has an undocumented temperature at weld time. Ambient loss on a winter day at 35°F ambient can drop a preheated joint 50–80°F in 30 minutes depending on mass and surface exposure.
Linking preheat practice to your WPS workflow
Consistent preheat verification starts with a WPS that states clear minimum preheat values — not a range like "200–400°F" with no floor enforcement. The WPS review checklist for CWIs walks through what inspectors should confirm in the WPS document itself before production begins.
For fab shops tracking multiple weld procedures across structural projects, keeping preheat records organized by joint type, material, and WPS reference is far easier with a digital system than with paper travelers. See pricing to explore how a structured WPS management tool supports the inspection workflow.
Field preheat verification is not just a paperwork exercise — it is the last defense against hydrogen cracking in joints that can't be safely reworked after structural erection. A half-inch of discipline at the pre-heat check before every arc start is worth more than any amount of corrective documentation after a failure.