What causes welding distortion?

Distortion comes from uneven heating and cooling. The weld and nearby metal expand when heated, but when cooling they shrink against the cold, restrained surrounding plate. Because the shrinkage is greater near the weld face than the back, the plate bends, which is angular distortion.

What is the difference between angular and longitudinal distortion?

Angular distortion is the rotation of the plate about the weld line, measured in degrees, caused by the shrinkage gradient through the thickness. Longitudinal shrinkage is the shortening of the part along the weld length as the deposited metal contracts, measured in millimetres per metre.

How does heat input affect distortion?

More heat input means a larger heated zone and more shrinkage, so distortion rises roughly with heat input and falls with thickness. Using lower heat input with smaller, faster beads is one of the most effective ways to keep a part flat.

What are the best ways to control distortion?

Preset the parts at the opposite angle so they spring back to flat, use a backstep welding sequence, balance passes about the neutral axis or weld from both sides, clamp the work in stiff fixtures, and minimise heat input. Combining several of these works best.

How accurate is this estimate?

It is a planning-level rule of thumb. Real distortion depends strongly on restraint, weld sequence, fixturing, and material, so treat the numbers as relative guidance for comparing options rather than precise predictions. Trial welds on representative assemblies give the real figure.

Welding Distortion Estimator

Distortion is the unavoidable side effect of pouring concentrated heat into a restrained plate. As the weld and its surroundings cool, they shrink against the cold metal around them, bending the part and shortening it along the joint. This estimator gives planning-level figures for angular and longitudinal distortion so you can compare joint designs and choose countermeasures before striking an arc.

How it works

Distortion scales with the heat poured in and shrinks with thickness. The estimator uses calibrated empirical relations of the form:

angular (deg)        ~ C x HI / t^1.3   summed with diminishing per-pass effect
longitudinal (mm/m)  ~ 0.9 x HI x passes / t   (capped)

where HI is heat input per pass in kJ/mm, t is thickness in mm, and C is larger for a single-V butt than a fillet because its weld metal sits more eccentrically across the thickness. Spreading the same fill over more passes reduces each pass’s contribution, which is why the per-pass effect diminishes.

Tips and countermeasures

Preset: tilt the parts to the opposite angle before welding so shrinkage pulls them flat.
Backstep: lay short weld increments in the reverse direction of overall progress to cancel longitudinal pull.
Balance: alternate passes either side of the neutral axis, or use a double-sided prep, so opposing shrinkage offsets.
Restraint: strongbacks and stiff fixtures hold geometry while welding, though they raise residual stress.
Less heat: smaller, faster beads with lower heat input shrink less; this is usually the highest-leverage change.

Treat the output as relative guidance for comparing options. A trial weld on a representative assembly is the only way to get a reliable number for production.