Thermomechanically controlled processed (TMCP) steel has changed the economics of high-strength structural fabrication over the past two decades. Grades like ASTM A913 (quench and self-tempered), A709 HPS 70W, and some mill-specific HSLA offerings achieve 65–100 ksi yield strength with a carbon equivalent low enough to weld without preheat at thin sections. But that favorable weldability comes with a condition: the microstructure that produces high strength is created by rolling temperature and cooling rate control at the mill, not by bulk chemistry. Apply too much heat during welding, and you undo part of what the mill did.

Getting a WPS wrong for TMCP steel doesn't produce an obvious weld defect visible to the eye — the fusion may look perfect, the surface is smooth, and the hardness check passes. The failure mode is a softened HAZ whose yield strength or Charpy toughness falls below the specified minimum, detectable only by mechanical testing. This article explains how HAZ softening works, why the WPS must control maximum heat input, and how to set the qualified range through PQR testing.

How TMCP Microstructure Is Created

In a conventional normalized or hot-rolled steel, carbon and alloying elements (manganese, silicon, niobium, vanadium) control microstructure. Increasing strength means increasing alloy content, which raises carbon equivalent and reduces weldability.

TMCP decouples strength from alloy content by controlling:

  1. Rolling finish temperature: The steel is rolled at a temperature where austenite grains are still fine and work-hardening is preserved.
  2. Accelerated cooling rate: After rolling, water quenching or accelerated air cooling transforms the austenite into a fine-grained bainitic or ferritic-bainitic microstructure at a higher cooling rate than air cooling alone would produce.

The result is a steel with 50–100% higher yield strength than its chemistry alone would predict, plus excellent low-temperature toughness. The downside: a weld's heat input drives the HAZ through a partial thermal cycle that can re-austenitize, then cool at a rate determined by the weld process, not the mill — reversing some of the controlled microstructure.

HAZ Softening: The Mechanism

The HAZ in a multi-pass weld has multiple thermal zones:

  • Fusion line and coarse-grained HAZ (CG-HAZ): adjacent to the weld bead; peak temperature exceeds 2200°F [1200°C]; grain growth is significant; for TMCP steel this zone can re-austenitize and then cool at rates determined by the weld process
  • Fine-grained HAZ (FG-HAZ): partial re-austenitization produces fine grain structure; often the toughest zone
  • Intercritical HAZ (IC-HAZ): temperature between Ac1 and Ac3; partial transformation creates a mixed microstructure
  • Subcritical HAZ (SC-HAZ): below Ac1; no phase change, but tempering occurs — this is where softening is most pronounced in TMCP steels

In a TMCP steel, the subcritical and intercritical zones can experience over-tempering: the controlled bainite/martensite microstructure softens as carbon diffuses and precipitate coarsening occurs. At very high heat inputs the hardness in this zone can drop by 20–30 HV, corresponding to a yield strength reduction of 6–10 ksi [40–70 MPa] — potentially enough to put a heavily-loaded connection below design yield.

The softening is proportional to heat input and to the number of passes that reheat a given location. SAW with multi-electrode setups and high deposition rates is the most common offender. Single-pass SMAW or GMAW root passes at low heat input rarely produce measurable softening.

Heat Input Calculation for the WPS

AWS D1.1:2025 uses the arc energy formula:

H = (E × I × 60) / (1000 × S)

Where:

  • H = arc energy in kJ/in
  • E = arc voltage (V)
  • I = welding current (A)
  • S = travel speed (in/min)

For processes with efficiency factors (SAW typically 1.0, SMAW ~0.8, GMAW/FCAW ~0.85, GTAW ~0.6), multiply the arc energy by the thermal efficiency to get heat input per AWS D1.1:2025 Annex C. For WPS documentation and TMCP steel compliance, use the arc energy figure as documented on the WPS and verified at production start.

A typical upper bracket for a TMCP structural steel plate (e.g., A913 Grade 65):

  • Thin plate (< 1/2 in [12 mm]): 40–55 kJ/in maximum
  • Medium plate (1/2–1 in [12–25 mm]): 65–80 kJ/in maximum
  • Heavy plate (> 1 in [25 mm]): 80–100 kJ/in maximum

These are indicative; the steel manufacturer's welding guidelines are the authoritative source. Nippon Steel, SSAB, and ArcelorMittal all publish process-specific heat input limits for their TMCP offerings. Always request the welding datasheet from the mill with the material and attach it to the project quality file.

Rule library based on AWS D1.1:2025; verify against your governing edition (the AHJ or contract may specify 2020 or earlier).

Where Heat Input Appears in AWS D1.1:2025 Essential Variables

For a tested WPS under Clause 6 of AWS D1.1:2025:

  • Table 6.6, row 11 (heat input): An increase in heat input beyond the range established during PQR testing is listed as an essential variable change requiring requalification. For most carbon steel WPS documents, the fabricator sets a range and qualifies both ends. For TMCP, the maximum is the binding constraint; the minimum is less critical unless HAZ hardness (cold cracking risk) is also a concern.
  • Table 6.8 (CVN supplementary essential variables): When Charpy impact test results are required by the contract or code clause, Table 6.8 row 2 imposes a stricter regime. Heat input changes are a supplementary essential variable under Table 6.8 — increases that shift HAZ properties outside the tested range trigger requalification regardless of whether the basic weld mechanics change.

For a prequalified WPS under Clause 5: AWS D1.1:2025 does not list heat input as a prequalified parameter subject to formal limits. However, for TMCP steels the prequalified route is generally inappropriate unless the steel manufacturer's documentation confirms compliance at the prequalified WPS parameters. Most TMCP steel for bridge work (ASTM A709 HPS 70W) triggers an automatic requirement for a tested WPS under the applicable bridge code (AWS D1.5).

Setting the Heat Input Range on the WPS

The PQR test program for a TMCP steel WPS should be designed to bracket the production range:

  1. Identify the maximum production heat input. Work backward from your worst-case joint: heaviest section, largest groove, highest-productivity process (SAW or FCAW-G at high wire feed speed). Calculate heat input at realistic production parameters.
  2. Run the PQR at or slightly above that maximum. The test coupon absorbs the worst thermal cycle. If the HAZ mechanical properties pass at this heat input, production will pass too.
  3. Run CVN specimens from the CG-HAZ and the SC-HAZ. The CG-HAZ is where toughness drops most; the SC-HAZ is where yield strength drops most. AWS D1.1:2025 Table 6.8 specifies CVN specimen locations; for TMCP steels, consider requesting additional specimens from the SC-HAZ even if not explicitly required by the code, as a conservative approach.
  4. Record the maximum qualified heat input on the WPS. This becomes the production limit. Add a note to the WPS: "Maximum heat input X kJ/in [per mill welding datasheet for [Grade/mill]; do not exceed without requalification."

A minimum heat input is also worth recording when cold cracking is a concern (very high CE steels, hydrogen-sensitive grades). For typical structural TMCP with low CE, the minimum is less critical.

Process-Specific Guidance

SAW: The highest heat input risk. Multi-wire SAW can exceed 200 kJ/in on thick plate — well above what most TMCP steels tolerate. For TMCP materials, single-wire SAW with controlled parameters is safer; discuss with the mill whether tandem SAW is permissible and at what heat input limit.

FCAW-G and GMAW: Mid-range risk. High wire feed speeds push heat input up, but the process is inherently faster-traveling than SAW, which limits the total heat delivered. The WPS should specify a maximum current/wire feed speed combination, not just a current range.

SMAW: Lowest heat input risk for TMCP softening. Individual pass heat input rarely exceeds 60 kJ/in even with large electrodes. The greater risk with SMAW on TMCP is under-heat: too-cold starts before arc stabilization can cause hydrogen cracking if the steel has even moderate CE and preheat is at minimum.

GTAW: Very low heat input; not a softening concern. GTAW is typically used only for root passes on TMCP pipe; the fill and cap passes will dominate HAZ behavior.

Pass Sequence and Interpass Effects

On multi-pass groove welds in thick TMCP plate, each subsequent pass reheats the HAZ of the previous pass. The cumulative thermal effect matters:

  • Grain refinement by subsequent passes is beneficial in the CG-HAZ — the reheating from later passes can break up coarsened prior-austenite grains.
  • Over-tempering by subsequent passes in the SC-HAZ compounds with each additional pass — the zone that doesn't re-austenitize keeps softening.

For thick sections, a temper bead technique (an additional pass deliberately placed to temper the coarse-grained HAZ) can improve HAZ toughness without PWHT. This must be documented on the WPS and supported by the PQR test. AWS D1.1:2025 does not prohibit this technique; it simply needs to be captured in the qualified procedure.

Documenting Maximum Heat Input on the WPS Form

Use Annex M of AWS D1.1:2025 as the WPS format. The electrical parameters section should read:

  • Voltage range: XX–XX V
  • Amperage range: XXX–XXX A
  • Travel speed range: XX–XX in/min
  • Maximum heat input: XX kJ/in (based on PQR and mill welding datasheet)

Add a note column: "Exceeding maximum heat input constitutes an essential variable change per AWS D1.1:2025 Table 6.6 and requires WPS requalification. Do not interpolate beyond the qualified range without written authorization from the RWS."

This language is not boilerplate — it tells the welder, the foreman, and the CWI that the upper parameter limit is a quality-hold condition, not just a preference.

For related reading on how heat input is tracked in the rule engine, see Arc Energy and Heat Input in AWS D1.1 and SAW High Heat Input HAZ Grain Coarsening. If your project requires CVN impact testing at the WPS level, see Charpy CVN Testing and AWS D1.1:2025 Table 6.8.

Need a WPS built around TMCP or HSLA steel? WPS Welding drafts code-compliant procedures with heat input ranges and CVN essential variables documented out of the box.