PSI - Issue 19

Lloyd Hackel et al. / Procedia Structural Integrity 19 (2019) 346–361 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

353

8

the cut. Finite element analysis of the strained surfaces was used to determine the two-dimensional residual stress of confinement. Numerous panels were fabricated. Guided by the FEA predictions, panels were laser peened with variations of process parameters. Contour measurements taken and an optimum process for canister peening determined. The lower diagram of Figure 6 shows the stress profile results in false-color representation for an un-peened panel. The false color depictions indicate tensile stresses in the 10 ksi (70 MPa) to 20 ksi (140 MPa) range as expected in the weld and heat affected zones. These would be of concern for stress corrosion cracking. In contrast, the upper panel of Figure 6 shows measurement results after laser peening of the same area of an identically welded panel. Three specific areas of importance are indicated with arrows where lineouts of the measured stress are graphically displayed in Figure 7. Comparing the lineouts it can be seen that the laser peening transformed these tensile areas to compressive (blue) residual stress. More specifically the line outs indicate that the laser peening converted the tensile stress in the weld area from approximately 25 ksi (175 MPa) tensile to 20 to 25 ksi (140 MPa to 175 MPa) compressive. In the heat affected areas the transformation was from20 ksi tensile (140 MPa) both to 20 to 25 ksi (140 MPa to 175 MPa) compressive. Depth of compression in the weld zone extended to 0.16 inches (4 mm) to a deeper 0.225 inches (5.6 mm) in the heat affected zones. As will be discussed below, the intensity and depth of compressive stress in a canister will exceed that measured in the free-state panels with depth of compressive stress in the canister being 2 mm greater. The high tensile stress regions at the ends of the panel are “artifacts” of the torch cutting of the panel panels from the larger plate. The residual stress transformation by the laser peening from tensile to compressive between un-peened and peen areas is consistent with the ability of the laser peening to prevent stress corrosion cracking.

Figure 7. Lineouts of stress profiles from Contour measurements of unpeened and laser peened panels. Laser peening converted tensile stress of up to 20 ksi (140 MPa) to 20 ksi (140 MPa) compressive stress and generated deep compression to 0.16 inch (4 mm) to .25 inch (6.2 mm) depth.

Computing stress state differences between cut panels and equivalent area in canister: In measuring residual stress in a component or test sample it is very important to consider geometry constraints because stress and strain are fundamentally coupled by the material’s modulus of elasticity. Peening of a geometrically constrained component will give more intense and deeper residual stress compared to identical peening of a test sample with edges free to strain and thus free to relax some of the residual stress. Such is the case in measuring the residual stress imparted by laser peening to a 316L test sample intended to represent the stress in a treated canister; The test sample will under report the stress in the canister. However using our FE analysis we can calculate corrections that show deeper and more intense residual compressive stress that will actually be retained within the canister. Figure 8 illustrates the methodology used to relate measurements made in isolated panels to the actual stress computed by our FEA in the canisters walls. Blocks 2 x 2 x 0.650 inches (50 x 50 x 16 mm) of material, for example 316L stainless steel, are fabricated and individually laser peened with 1, 2 or 3 layers of peening. Stress per layer of peening is then derived from slitting measurements. An FEA model of the canister is built and the laser peening applied in the model to specific canister areas of the canister.