我的新闻

Aiming to predict the occurrence of cracks in additively manufactured (AM) Ni-based superalloys, this study proposed a novel integrated multi-cracking susceptibility model and index (CSI). The model comprehensively accounts for key thermophysical and microstructural factors, including the viscosity of the alloy melt, the presence of carbides and γ/γ΄ eutectic, as well as the influence of stacking fault energy and grain boundary (GB) energy on crack formation during rapid solidification of AM. To experimentally validate the susceptibility model, CM247LC superalloys with varied (Hf+C) contents were fabricated via laser-based powder bed fusion (LB-PBF), with their cracking behavior systematically quantified. Predictions revealed that crack susceptibility initially decreased and then increased with rising (Hf+C) content, fully corroborated by experimental results. Notably, the addition of 1 wt% (Hf+C) resulted in a significant reduction in cracks, yielding minimum crack length and area densities of 0.004 mm/mm² and 0.017 %, respectively. The healing effect on cracks is attributed to two synergistic mechanisms: the backfilling effect of the Hf-rich eutectics into cracks and the dispersion-strengthening effect of more fine carbides along the GBs. The predictive capability and generalizability of the proposed CSIs were validated via literature-derived machine learning, demonstrating that CSI_SC achieves 20 % higher accuracy than Kou’s classical criterion. The developed model shows significant potential for capturing the complex cracking behavior and guiding the design of crack-free AM superalloys.