Weld Distortion Control: What Goes in the WPS

Weld distortion is one of those problems that's easiest to address at the WPS stage—and most expensive to deal with after fit-up is done and the welder has made a dozen passes. A plate that warps 3/16 inch on a base plate connection costs rework time; on a crane girder, it costs a shop's relationship with the erector.

The WPS is where distortion controls belong. Not as boilerplate disclaimers, but as specific, enforceable parameters that the CWI can actually verify.

Why Distortion Happens: The Short Version

Weld distortion is caused by differential thermal expansion and contraction. When metal heats up, it expands; when it cools, it contracts. The weld zone cools from above the melting point to ambient, contracting significantly. The surrounding base metal, which was cooler to begin with, constrains that contraction—leaving residual stress and, depending on joint geometry and stiffness, measurable distortion.

The variables that drive distortion severity:

• Heat input per pass: Higher heat input means more thermal energy deposited, more expansion, more contraction on cooling.
• Joint volume: More weld passes means more cumulative thermal cycling and shrinkage.
• Joint restraint: A fully restrained joint distorts less in gross geometry but develops higher residual stresses.
• Welding sequence: Controls where heat goes and what's cooling while other areas heat up.
• Interpass temperature: Higher interpass temperature means the base metal starts each pass already warm, reducing the temperature gradient but keeping the part in a thermally expanded state longer.

Each of these is a WPS variable. Each is controllable before the first arc is struck.

Heat Input as the Primary WPS Lever

Heat input is the most direct handle on distortion at the WPS level. AWS D1.1:2025 uses heat input in the essential variables framework (Table 6.6) primarily for mechanical property purposes—but the same formula applies to distortion control:

HI = (V × A × 60) / (S × 1000)

Where V = arc voltage, A = welding current, S = travel speed in inches per minute. Result is in kJ/in.

A WPS that specifies only nominal voltage and amperage without a travel speed range gives the welder an effective blank check on heat input. At nominal 24V / 280A, a welder running 8 ipm delivers 12.6 kJ/in; running 6 ipm delivers 16.8 kJ/in—a 33% increase in heat input that the WPS technically permitted.

Better practice: specify a heat input range as a supplemental parameter, or specify a minimum travel speed. On distortion-sensitive work, stating "maximum heat input 14 kJ/in" gives the CWI a verifiable limit and gives the welder a clear constraint.

Rule library based on AWS D1.1:2025; verify against your governing edition.

Weld Sequence: What the WPS Should Specify

Sequence requirements are among the most underspecified elements in structural fab WPS packages. Most WPS documents say nothing about sequence, leaving it entirely to the fitter and welder. On simple fillet welds to non-critical geometry, that's acceptable. On distortion-sensitive built-up members, it's a gap.

Balanced welding about the neutral axis: For I-shaped built-up members (plate girders, columns), alternating flange-to-web welds between top and bottom flanges distributes heat symmetrically. A welder who completes all four passes on the top-flange-to-web before starting the bottom will induce a consistent camber—which may or may not be the design intent.

Backstep sequence for long joints: In a backstep sequence, the overall welding direction is left-to-right (for example), but each individual weld segment runs right-to-left. As each segment completes and cools, it opposes the shrinkage of adjacent segments. This technique reduces longitudinal shrinkage on long groove welds and can cut angular distortion on long fillet welds.

Block sequence vs. continuous pass: For multi-pass groove welds in thick plate, a block sequence (completing a section to full depth before moving along the length) concentrates distortion at completion points. A continuous layer-by-layer sequence distributes distortion more evenly. The choice depends on joint geometry and fit-up rigidity.

When distortion is a design constraint—as it often is for crane runway girders, press platen weldments, or precision machine bases—the WPS should explicitly state the required sequence. "Weld in balanced sequence; alternate passes on opposite sides of neutral axis" is a verifiable instruction. "Weld as required" is not.

Interpass Temperature: The Distortion Tradeoff

Interpass temperature control is specified in the WPS for metallurgical reasons (preventing HAZ grain coarsening, maintaining toughness), but it also affects distortion:

Lower interpass maximum = lower average thermal mass during welding. Each pass starts on cooler base metal, creating a higher temperature gradient and faster cooling rate. This tends to reduce total shrinkage per pass but may increase residual stress.

Higher interpass = part stays warmer longer. The cumulative thermal cycle is extended. This can be useful on highly restrained joints (reduces cracking risk) but may allow more total distortion.

For distortion control on lightly restrained members, keeping the interpass temperature moderate—below 400°F on carbon steel unless metallurgy requires higher—and returning the part to ambient between multi-pass sequences can help manage cumulative distortion.

The WPS must specify both minimum preheat (for metallurgical reasons; tied to Table 3.2 in D1.1:2025) and maximum interpass temperature. On distortion-sensitive work, the maximum is as important as the minimum.

Joint Preparation and Fit-Up as a Distortion Variable

The WPS can't fully address distortion without considering joint geometry. Groove angle affects joint volume, which directly affects shrinkage. A 45° included angle groove in 1-inch plate has roughly twice the weld volume of a 30° groove in the same plate—meaning roughly twice the shrinkage force.

Where distortion is critical, reducing groove angle (within the prequalified or qualified limits) reduces joint volume. For prequalified joints, AWS D1.1:2025 Clause 3.9 sets minimum groove angles; you can tighten them but not widen them beyond limits without affecting prequalification status.

Pre-set (pre-cambering in the opposite direction of expected distortion) is a common fab practice but belongs in the fabrication procedure, not usually in the WPS itself. However, the WPS can reference a fabrication procedure that specifies pre-set targets.

What the CWI Verifies in the Field

Distortion control language in a WPS is only valuable if the CWI can and does verify compliance. Practical verification points:

Heat input spot-checks: Measure voltage and amperage with a calibrated clamp meter during welding; verify travel speed by timing a measured weld length. Calculate actual HI and compare to the WPS maximum. Do this during the first setup on a new joint type, not just at the start of a project.

Interpass temperature checks: Infrared thermometer or contact pyrometer before each pass, especially on multi-pass joints where interpass temperature tends to climb as the part heats up. Document at least the first few passes; many CWIs require continuous documentation on CVN-qualified or seismic welds.

Sequence observation: On built-up members with a specified sequence, observe and document the sequence during the first production unit. If the sequence is correct and the result is within distortion tolerance, you have a verified baseline. If the sequence is correct and the result is out of tolerance, the issue may be pre-set, restraint, or heat input—not the sequence itself.

Distortion measurement: On precision members, a straight-edge or laser check of camber, sweep, and angular distortion after each major weld sequence (before the part moves to the next station) catches problems early. Once a member is fully welded and has cooled, correcting distortion often requires flame straightening—expensive and time-consuming.

For related reading, see Arc Energy and Heat Input in AWS D1.1 for the full heat input formula with units, and Built-Up Plate Girder WPS: Controlling Camber and Sweep for a specific member-type application. The WPS Revision Control Best Practices article covers how to document and control sequence changes without creating unqualified WPS variants.

Building Distortion Controls Into Your WPS Library

Distortion control isn't a one-size-fits-all spec addition. A single-pass fillet weld on a non-critical connection doesn't need a sequence requirement. A 72-inch built-up plate girder for a crane runway absolutely does.

The practical approach: identify the distortion-sensitive work in your shop's typical scope (built-up members, large weldments, precision base plates), and develop specific WPS variants for those configurations that include heat input limits, sequence requirements, and interpass maxima. Standard WPS variants for non-sensitive work can remain simpler.

When those WPS variants are developed alongside a qualified PQR (or confirmed prequalified conditions), they give the quality system a defensible, verifiable foundation—not just a note in the job traveler that says "watch for warping."

For shops building out structured WPS libraries that include distortion-sensitive variants, wpswelding.com/pricing outlines the documentation support available.