PSI - Issue 13

5

Richard J. Williams et al. / Procedia Structural Integrity 13 (2018) 1353–1358 Richard J. Williams et al./ Structural Integrity Procedia 00 (2018) 000–000

1357

(MPa)

Fig. 6. Stress in the build direction, or σ zz , stress contours for an isometric view of XZ plane of the vertical cylinder, at the end of the build

The vertical sample can be seen to have very high magnitude stresses in the build direction, driving the deformation to take place when slit. The large height of the part in this orientation also means it undergoes more stress generating thermal cycles, which superimpose. Compared to the vertical sample, the horizontal cylinder has lower magnitude peak stresses in the build direction, as it is shorter in this direction undergoes fewer thermal cycles. Although the absolute peak magnitude of tensile stress is higher in the horizontal component, this is in a small region and the majority of the component has lower tensile stress than the vertical. While it has a long aspect ratio in the horizontal, or scanning, direction the largest thermal gradient still exists in the vertical direction as the heat is conducted down and away through the build plate and hence this is the principal stress direction. The horizontal cylinder did not deflect upon slitting as much of the stress was relaxed when the component was cut from the build plate. The curling up upon removal from the build plate can be attributed to its low second moment of area about this axis. Conversely, more of the stress was retained in the vertical sample when this was cut from the build plate. It can be seen that in both build orientations the components exhibit generally compressive stress at the center and tensile stress at the surface, with stress magnitudes exceeding yield by almost two times predicted. The same trends are seen in the other principal stress directions, albeit with reduced magnitudes. This is likely to be detrimental to fatigue and fracture performance as any cracks initiating at the part surface will grow rapidly. 5. Conclusion A computationally efficient FE modelling methodology has been applied to investigate the effect of build orientation on the stress field of a component. A cylindrical component was built in both the horizontal and vertical build orientations and sectioned down its diameter. The resulting deflections were compared with those predicted by the model and good agreement to within 5% of the experimental measurement was found. This provided further validation that the pragmatic FE modelling methodology developed is able to accurately predict residual stresses and distortions in LPBF components. Examination of the predicted stress fields also supported an understanding of the influence of component geometry and build orientation on residual stress profiles. In particular, component build height and aspect ratio were found to impact stress profile and resultant behavior upon sectioning and slitting.