Hot isostatic pressing in metal additive manufacturing: X-ray tomography reveals details of pore closure

Abstract

Hot isostatic pressing (HIP) of additively manufactured metals is a widely adopted and effective method to improve the density and microstructure homogeneity within geometrically-complex metal structures fabricated with laser powder bed fusion (LPBF). The role of pores in the fatigue performance of additively manufactured metal parts is increasingly being recognized as a critical factor and HIP post-processing is now heralded as a method to eliminate pores, especially for high-criticality applications such as in the aerospace industry. Despite the widely reported positive influence on fatigue performance and high efficiency of pore closure, examples have been reported in which pores have not been entirely closed or have subsequently re-opened upon heat treatment. A variety of porosity distributions and types of pores may be present in parts produced by LBPF and the effectiveness of pore closure may differ depending on these pore characteristics. In this work, X-ray tomography was employed to provide insights into pore closure efficiency by HIP for an intentional and artificially-induced cavity as well as for a range of typical process-induced pores (lack of fusion, keyhole, contour pores, etc.) in coupon samples of Ti6Al4V. The same samples were imaged non-destructively before and after HIP and aligned carefully for side-by-side viewing. High pore closure efficiency is demonstrated for all types of cavities and pores investigated, but near-surface pores of all types are shown to be problematic to varying degrees, in some cases perforating the superficial surface and creating new external notches. Subsequent heat treatments (annealing after HIP) in some cases resulted in internal pore reopening for previously closed internal pores as well as a new “blistering” effect observed for some near-surface pores, which the authors believe is reported for the first time. Implications of these results for quality control and HIP processing of LPBF parts are discussed. Finally, the utility of using HIP to consolidate intentionally-unmelted powder in order to improve production rates of powder bed fusion has great potential and is preliminarily demonstrated.