Mechanical Modeling of Thermally Induced Residual Stresses and Distortion of Ti-6Al-4V Double Track Thin Walls Fabricated With Different Scan Strategies

Research has been performed to connect additive manufacturing (AM) process parameters including laser power and scanning strategy to different aspects of part quality, such as porosity, mechanical properties, fatigue life, microstructure, residual stresses, and distortion. The lack of predictive capabilities to fully estimate residual stresses and distortion within parts produced via AM have hindered part qualification; however, modeling the AM process can aide in process and geometry optimization compared to traditional trial-and-error methods. The presence of unwanted thermally induced residual stresses and distortion can lead to tolerancing issues, reduced fatigue life, and decreased mechanical performance compared to similar components fabricated with traditional manufacturing methods such as casting and machining. This work focuses on examining the effect of varying scanning strategies on the temperature history of a double track thin wall, and thus, the resultant thermally induced distortion and residual stresses.

Ti-6Al-4V double track thin specimens were fabricated via Laser Engineered Net Shaping (LENS), a directed energy deposition (DED) process characterized by blown powder with a coaxial laser heat source. Each double track thin wall was fabricated using the same process parameters (except for scanning strategy) and measure 10 mm in height for a total of 20 layers, 25 mm in length, and 1 mm wide. Two different scanning strategies were used in this work: (i) unidirectional and (ii) bidirectional. The differing scanning strategies result in a different temperature history for each part, particularly the heating and cooling rates. The difference in heating and cooling rates found in each part results in a difference in the thermally induced distortion and residual stresses.

Dual-wave pyrometer images taken of the melt pool during the build process were used to calibrate the three-dimensional finite element thermal model. The thermal model was constructed in the Abaqus framework, and the simulated temperature history was then read in to the sequentially coupled mechanical model to examine the differences in residual stresses and distortions between the different scan strategies. The simulated temperature history was then read in to the sequentially coupled mechanical model in Abaqus to examine the differences in residual stresses and distortion between the two builds. The mechanical model used in this work is the Evolving Microstructural Model of Inelasticity (EMMI), which is a rate- and temperature-dependent dislocation mechanics based internal state variable plasticity model that has been calibrated using stress-strain data taken at various temperatures and strain rates. Results were examined at different areas of interest for each part, specifically at the interfaces between the substrate and the base of the thin wall along with the middle of the part.