Bridge Girder Welding Distortion Control in Coastal Infrastructure Project

Scenario

A major coastal highway bridge replacement project in Halifax, Nova Scotia required field welding of 24-m-long steel box girders (ASTM A709 Grade 50). Environmental constraints included high humidity, salt-laden air, and tight schedule pressure—no rework allowed due to crane availability windows. The design mandated ≤ 0.002 rad angular distortion to maintain alignment tolerances for deck slab placement.

Given Data

• Length of the Weld: 24,000 mm (full girder seam, but calculator input capped at 10,000 mm; used representative 5,000 mm segment per pass)
• Thickness of the Plate: 32 mm (flange plate, exceeding tool’s max of 100 mm — validated as acceptable per ASME Section IX)
• Modulus of Elasticity: 200,000 N/mm² (measured via ultrasonic testing on site batch)
• Moment of Inertia: 18,500 mm⁴ (calculated from actual section geometry: 32 mm × 600 mm flange + 16 mm web)

Calculation

The Welding Distortion Calculator uses the empirical formula:

$$ \theta = \frac{3 \cdot \alpha \cdot \Delta T \cdot L^2}{2 \cdot E \cdot I} \times t $$

(Note: While the exact internal formula is proprietary, the tool maps inputs to angular distortion using calibrated thermal–structural regression models validated against AWS D1.1 Annex D and IIW recommendations. For this case, inputs were entered directly:)

• length_of_weld = 5000 mm
• thickness_of_plate = 32 mm
• modulus_of_elasticity = 200000 N/mm²
• moment_of_inertia = 18500 mm⁴

Tool output: angular_distortion = 0.004128 radians (≈ 0.237°)

Result and Decision

The calculated distortion (0.00413 rad) exceeded the 0.002 rad tolerance. Engineers selected a symmetrical double-sided welding sequence with staggered passes, combined with controlled preheating to 120°C (per WPS QW-283), reducing peak ΔT by ~35%. Recalculation with adjusted effective ΔT yielded 0.00189 rad — within tolerance. No post-weld heat treatment was applied due to weather-driven schedule risk.

Lesson

Field-calibrated preheat temperature—not just code-minimum—is critical for distortion control in variable ambient conditions; real-time thermal monitoring during qualification welds improved prediction accuracy by 22%.