Interlaboratory Reproducibility of Contour Method Data in High Strength Aluminum Alloy
Interlaboratory Reproducibility of Contour Method Data in High Strength Aluminum Alloy
Wednesday, October 22, 2025: 9:40 AM
Background: The contour method for residual stress measurement has seen significant development, and a first experimental reproducibility study has recently been published. Objective: A double-blind reproducibly study, having scope beginning with EDM cutting and ending with residual stress calculation, is reported. Methods: A reinforced I-beam sample geometry is identified for its unique residual stress profile when extracted from residual stress bearing quenched aluminum bar (7050-T74). Contour measurements are prescribed on a midplane of symmetry with dimensions 24mm by 50mm. Fourteen identically prepared samples are fabricated from a single long bar with well characterized and uniform residual stress. Three samples throughout the bar are identified for benchmark contour measurements to validate the sample uniformity. Three samples are identified for measurement by neutron diffraction and other techniques. Eight samples are distributed to an international group of participants to execute their standard practice measurement. A double-blind process is followed to protect the participants privacy. Benchmark contour measurements of normal residual stress are reported alongside neutron diffraction results. Normal residual stress on the plane of interest is requested from each of the eight participants having received distributed samples. Results: Results were provided by eight participants: six were self-similar, and two were quite different. An average residual stress field is established from the available data, and the spatial distribution of reproducibility standard deviation is reported. The largest reproducibility standard deviations are observed around the exterior of the measurement region, within 1-2mm of the sample extents. Within the interior of the measurement region, the deviations are significantly lower. Analysis excluding these results indicates significantly lower reproducibility standard deviation, but similar trends regarding the exterior and interior regions.