|Thin-wall part fabrication is significant for aerospace, automotive, and thermal applications, which require complex metallic features hard to manufacture using conventional manufacturing processes. Laser powder bed fusion (LPBF) additive manufacturing (AM) technique enables the construction of such components with added design flexibility. However, build challenges, such as micro-cracking and in-process failure, inherent in LPBF, limit processability and affect final part quality. Micro-cracking is especially important for high γ’ Ni-based superalloys, which are used in serpentine-shaped components such as fuel nozzles and turbine vanes in gas turbine jet engines. To better understand the underlying mechanisms for thin-wall failure and determine if micro-cracking affects in-process failure, metallurgical and mechanistic factors affecting part processing must be considered. In this thesis, micro-cracking statistics, in-process stresses, part displacements, and stress triaxiality states are evaluated for thin-wall parts made with different build conditions, including different wall thicknesses, scan strategies, and alloy compositions. Experimentally-obtained and numerically-simulated results are used to explain in-process failure and determine optimal thin-wall build conditions to reduce in-process micro-cracking.
Firstly, a design of experiment (DOE1) including two different alloy compositions, RENÉ 65 (R65) and RENÉ 108 (R108), and three different part thicknesses, 5.00 mm, 1.00 mm, and 0.25 mm, is developed to evaluate the in-process failure and micro-cracking trends for thin-wall parts built with LPBF. The materials chosen in this study represent high γ’ Ni-based superalloys with different γ’ contents and solidification ranges, beneficial to observe different types of micro-cracking behavior. As-processed thin-wall parts demonstrate different limiting build heights (LBHs) with respect to wall thickness. Builds with thinner walls fail to achieve the designed part height, exhibiting lower LBHs compared to thicker parts for both R65 and R108. Microstructure characterization shows that micro-cracking is independent of the LBH effect as R108 thin-wall components exhibit larger crack densities than R65. The stress distributions along the build height of a thin-wall part also do not contribute to LBH as the stresses increase with wall thickness. Ultimately, the reduced LBH in thinner structures is correlated to increased part distortion, resulting from increased compressive stresses along the length and height directions. The slenderness ratio is proposed as a valuable tool to consider during design to achieve the desired height and avoid premature failure during LPBF thin-wall part fabrication.
The following study explores the effect of wall thickness on the micro-cracking tendencies and elucidates the micro-cracking mechanism in LPBF-processed thin-wall parts using the highly crack-susceptible R108. Experiment-based statistical analysis confirms increased micro-cracking with larger wall thickness, agreeing well with the previous work. Electron microscopy shows that all micro-cracks exhibit inter-dendritic morphologies indicating the solidification cracking mechanism. Contrary to other studies, micro-cracking is not caused by carbides or borides, as the Hf-,Ta-,Ti-rich carbide phases are fine and homogeneously distributed within the microstructure. All wall thicknesses have higher tendency to form straight micro-cracks along the build direction (BD). Numerical simulations performed at the layer scale show that thicker parts generate higher in-process tensile stresses and exhibit positive stress triaxialities during build progression, supporting the higher micro-cracking propensities observed in thicker parts. Beam-scale simulations show that the stress triaxiality state becomes positive at a higher (super-solidus) temperature within the melt pool, which along with high in-process tensile stresses supports larger number of straight micro-cracks.
The effect of in-process stresses on the micro-cracking densities is supported by a short study on the influence of build position on LPBF thin-wall component micro-cracking. Experimental results show that micro-cracking is lower at the base of the build. Numerical simulations predict compressive stresses at lower build positions, supporting the experimental finding and indicating that in-process stresses contribute to micro-cracking in thin-wall parts.
Subsequently, the effect of laser scan strategy on thermally-influenced stresses and cracking behavior is studied by generating a large DOE consisting of 32 R108 thin-wall parts including four different wall thicknesses (1.00 mm, 0.75 mm, 0.50 mm, and 0.25 mm) and ten different scan strategies. The effects of vector length and scan rotation are examined separately for different wall thicknesses. Thicker parts always demonstrate higher crack densities due to earlier transition to positive stress triaxialities and larger stress magnitudes. As the vector length increases, the micro-cracking propensity increases for all wall thicknesses. This observation is explained using two separate finite element model for parts processed with short and long vector lengths, respectively. Longer vector lengths produce larger in-process stresses perpendicular to BD and transition earlier to positive stress triaxialities, supporting the larger number of micro-cracks determined experimentally. The study of inter-layer scan rotations during processing indicates that the alternating short scan strategy is preferred over the continuous 67° scan rotation strategy for thicker parts. These results open up simple viable alternatives to mitigating micro-cracking in LPBF thin-wall parts by changing the wall thickness or laser scan strategy.
Finally, the effect of alloy composition on micro-cracking is investigated to better understand the internal micro-cracking mechanisms in high-γ’ Ni-based superalloys. Two LPBF-printed thin-wall components made of R65 and R108 with identical thicknesses of 1.00 mm are studied. The micro-cracking propensity of R108 is higher than R65 and the cracking behavior is interdendritic, which suggests the solidification cracking mechanism. Higher cracking in R108 is supported by finite element simulations which show that the stress triaxiality is positive at higher temperatures in R108 compared to R65. Thermodynamic simulations show that R108 has higher liquid fractions in the terminal stages of solidification which makes the material weaker at high temperatures under positive stress triaxiality conditions.