Welding Distortion Calculator

Because the relationship this calculator uses puts Plate Thickness SQUARED in the denominator alongside Weld Size squared in the numerator, so doubling thickness quarters that ratio. Two caveats matter for reading the number: the 4x only holds while the result stays below the 15-degree reporting ceiling -- above it the output is flat, and doubling thickness changes nothing until you drop back under the cap -- and the squared shape is this calculator's own calibration against the 1-to-5-degree band published for unrestrained fillet welds, not a coefficient any standard publishes.

Not unconditionally. Rigid fixturing (this calculator's "Heavy" restraint setting) does reduce visible angular distortion, which is what this calculator models. But real welding literature also documents that over-restraining a joint can trap residual stresses that release unpredictably after cooling, sometimes causing cracking or distortion elsewhere in the structure -- a tradeoff this calculator's single restraint multiplier doesn't capture, since it only tracks distortion, not residual stress.

AWS D1.1 bounds fillet weld size from both directions. Table 7.7 sets a minimum by the thickness of the thinner part joined -- 1/8 in up to 1/4 in material, 3/16 in through 1/2 in, 1/4 in through 3/4 in, and 5/16 in above -- so that the weld cools slowly enough to avoid cracking. Separately, the maximum fillet size along an edge is the material thickness for material under 1/4 in, and the thickness minus 1/16 in at or above 1/4 in, so the plate edge is not melted away. A 1 in leg on 1/16 in plate is not a weld you can make, which is why the shrinkage figures for it are an extrapolation rather than a fit-up allowance to work from.

Because it cannot be computed from these inputs. Camber is the bending deflection that longitudinal shrinkage produces when the weld sits off the assembly's neutral axis, so it depends on the cross-section's second moment of area, the weld's eccentricity from the neutral axis and the span -- none of which are inputs here, and none of which follow from plate thickness and weld length alone. An earlier version of this calculator printed a camber figure derived from weld length over plate thickness, which returned 0.000 in across the entire default configuration and had no mechanical basis; it was removed rather than left in place. For real camber, use the section properties of the actual member.

Treat them as a planning-stage estimate, not a substitute for procedure qualification. The shrinkage figures follow TWI's published allowances, but TWI's own guidance stresses that it is almost impossible to predict shrinkage accurately, and real distortion also depends on welding process, sequence, preheat, number of runs and material properties that are not inputs here -- for a fit-up tolerance that matters structurally, verify against AWS D1.1 procedure qualification or your own shop's measured distortion data.