Increasing the productivity of laser powder bed fusion: Influence of the hull-bulk strategy on part quality, microstructure and mechanical performance of Ti-6Al-4V

Abstract

To increase the productivity of Laser Powder Bed Fusion (LPBF), a hull-bulk strategy can be implemented. This approach consists in using a high layer thickness in the core of the part, hence reducing the build time, and a low layer thickness in the skin, to maintain a high accuracy and good surface finish. The present study investigated to what extent this strategy affected the surface roughness, relative density, microstructure and mechanical properties of Ti-6Al-4 V parts. Ti-6Al-4 V specimens were built using two distinct sets of process parameters, one optimized for a 90 μm-layer thickness in the bulk and the other for a 30 μm-layer thickness in the hull. In addition to surface roughness and relative density measurements, a thorough microstructure analysis was done using both optical microscopy and SEM. Additionally, EBSD measurements and numerical reconstruction of the parent β grains were performed to evaluate the mesostructure and texture evolution from hull to bulk. Microhardness measurements and tensile tests were done to assess the effect of the hull-bulk strategy on the mechanical properties. This analysis was completed on both as-built and stress-relieved specimens. The present study demonstrated the possibility of using the hull-bulk strategy to build high-quality Ti-6Al-4 V parts, without impacting their tensile properties, hence increasing the productivity of the process by a geometry-dependent factor, typically ranging between 25 % and 100 %.

Introduction

Laser Powder Bed Fusion (LPBF), also known as Selective Laser Melting (SLM) or direct metal printing (DMP), offers many advantages, in particular a higher design freedom than conventional manufacturing. The layer-by-layer approach inherent to additive manufacturing (AM) allows the building of complex geometries and custom-made designs. In comparison with other AM processes for metals, such as Electron Beam Melting (EBM) or Directed Energy Deposition (DED), LPBF offers a high accuracy. This is related to the fine layer thickness and relatively small powder size distribution typically used in the LPBF process. However, the ability of LPBF to manufacture parts with a high resolution is counterbalanced by the low productivity of the process. The build rate of LPBF is indeed typically lower than that of other powder-bed AM processes, such as EBM. To illustrate this, Table 1 provides a comparison of the productivity of an LPBF ProX320 machine and an EBM Arcam A1 machine, considering a build job of 256 tensile bars, i.e. the capacity limit of an Arcam A1 EBM machine.

The build rate is directly proportional to the layer thickness, scanning velocity and hatch spacing. Increasing the layer thickness is therefore a straightforward approach to improve the productivity [1] but directly impacts the surface roughness and geometrical accuracy of the part [2] and requires a thorough process parameters optimization in order to avoid the formation of lack-of-fusion porosity [3]. The trade-off between part quality and productivity remains one of the major obstacles to a wider industrialization of the LPBF process. To address this issue, various approaches have been developed and implemented, such as increasing the machine’s build volume to reduce operating time, or using multiple laser beam sources and multiple laser-scanning systems in one machine (“twin-laser” or “quad-laser” [4]) so that the build area can either be multiplied or that one build space can be processed by several lasers and scanning-systems at the same time [5]. However, these approaches involve significant research and development and typically lead to more expensive LPBF machines.

In order to combine geometrical accuracy and process productivity, while avoiding the use of larger or more expensive machines, a “hull-bulk” strategy (also known as “hull-core” or “skin-core” strategy) can be implemented [4,6]. This approach consists of applying a high precision parameter set for the hull (or skin) of the part, and a high productivity parameter set for its bulk (or core). From a practical point of view, this can be achieved by varying the spot size and laser power. As mentioned by Poprawe et al. [5], a large beam setting with high laser power combined with a large hatch spacing can be used for the bulk, whereas a fine beam is used to build the outer shape of the parts (skin). This approach was proposed by Schleifenbaum et al. [6], who used a duo-laser set-up and by Metelkova et al. [7], who increased the spot size by defocusing the laser beam. Scanning with varying spot sizes, however, was out of scope for this study, as it requires significant hardware modifications.

A simpler method, which can be applied as such in state-of-the-art industrial LPBF machines, without hardware adaptation, consists in using two different sets of process parameters in the bulk and in the hull. The hull is built with a conventional layer thickness, whereas a larger layer thickness and higher power are used for the bulk. This requires a preliminary process parameter optimization step, in order to define the optimal set of power, scanning speed and hatch spacing for this layer thickness.

Although the hull-bulk strategy is not new to the AM community [[6], [7], [8]], its impact on the material properties remains poorly understood for Ti-6Al-4 V, one of the most widely used alloys in the AM industry. Koutny et al. [8] recently investigated the effect of this strategy on Al parts processed with a skin-core strategy, reporting the presence of a graded microstructure, but to the best of the authors knowledge, no such study was performed on Ti-6Al-4 V.

In Ti-6Al-4 V manufactured by LPBF, the meso- and microstructure are known to be closely related to the processing conditions [9,10]. The mesostructure of AM Ti-6Al-4 V parts typically consists of columnar parent β grains resulting from rapid solidification and epitaxial growth along the thermal gradient [[11], [12], [13]]. In addition, the high cooling rate of the LPBF process determines the final microstructure within the prior β grains, which is martensitic in as-built condition [11,14]. Given the close correlation between process parameters and microstructure, the use of different processing conditions in the bulk and in the hull might result in the genesis of distinctive microstructures, hence impacting the mechanical properties. In order to validate the use of the hull-bulk strategy in critical industrial applications, it is essential to determine to what extent this strategy impacts the microstructure and mechanical properties of LPBF Ti-6Al-4 V parts. Hence, the present study investigated the meso- and microstructure in the skin, core and skin-core interface and evaluated how the presence of a gradient in mesostructure impacted the quasi-static mechanical properties of the material. In addition to a thorough microstructural characterization, microhardness and uniaxial tensile properties were evaluated in order to establish structure-property relationships. The effect of a subsequent stress-relief post-treatment was also investigated.