Determination of Residual Stress Profiles in A Friction Stir Welded Stringer-to-Skin Lap Joint

Riveting has been for many years the conventional method of joining aluminum components such as aircraft stringer-to-skin panels in the lap joint configuration. Due to a demand for a more economical and improved joining method, friction stir welding has been proposed as an alternative welding process. Residual stresses left behind from the friction stir welding process caused the stringer-to-skin lap joint to twist after the clamps are removed. A post-welding process of hammer peening was able to remove the distortion in the plate. In this thesis, the crack compliance method is used to calculate the residual stresses in the friction stir welded lap joints in the through thickness, transverse, and longitudinal direction. A non-destructive residual stress measurement technique known as neutron diffraction was used on the lap joint in both as-welded and hammer peened conditions. The residual stress profiles in the through thickness, transverse and longitudinal directions were successfully measured and compared with the residual stress from the compliance method. It was determined that the compliance method was successful in calculating the residual stresses in the through thickness and transverse direction but was not able to calculate the residual stress in the longitudinal direction. Distortion in the plate was a result of a combination of non-uniform residual stress profile in the three principal directions. Fatigue testing was performed on the as-welded and hammer peened configuration and it was determined that the hammer peened condition outlived that of the as-welded condition under the same loading frequency. An examination of the failure locations in the friction stir welded lap joints concluded that the as-welded specimens failed at a location in the weld region with a combination of weld defects, lowest hardness value and highest tensile residual stresses.