Spatiotemporal variations of residual stresses during multi-track and multi-layer deposition for laser powder bed fusion of Ti-6Al-4V

Abstract

Laser powder bed fusion (LPBF) is an additive manufacturing technology to fabricate parts with complex geometries. Residual stresses induced by the rapid heating and cooling processes may cause defects such as cracks, distortion and delamination. This work presents an in-depth study for the understanding of the evolution of residual stresses during LPBF of Ti-6Al-4V. A 3D finite element model was developed to study the spatiotemporal variations of the temperature and the stresses during multi-track and multi-layer LPBF. The results show that the scanning strategies significantly affected the temperature gradients and the resultant stress distributions. The induced residual stresses were higher along the laser scanning direction than that along the perpendicular direction. Monitored at the longitudinal central plane, stresses changed from compressive to tensile along the vertical build direction from the substrate to the top of the build. The Z-component of stress was smaller than the X- and Y-components stresses. The peak tensile Z-component of stress was at the interface between the part and the substrate, which may lead to local warping and cracking. The findings from this work provide insights for the understanding of stress related issues such as the massive longitudinal and transverse solidification cracking during LPBF.

Introduction

Laser powder bed fusion (LPBF) is a fast-growing additive manufacturing technique, during which process, layer-wise deposition is realized through the repetitive heating, melting, solidification, and cooling processes of the feedstock material [1], [2]. Despite its numerous advantages, LPBF has many issues such as residual stresses, cracking, and geometric inaccuracies [3], [4], [5], [6]. These issues are interdependent and are affected by the cyclic thermal cycles generated by the rapidly moving laser beam during LPBF. A deep understanding of the formation mechanisms and the mitigation strategies of these printing issues is important to the production of a sound LPBF component.

A comprehensive understanding of the spatiotemporal variations of the temperature and the stresses during LPBF is a prerequisite for the solution of the above printing issues. The development of residual stresses during LPBF depends on several factors including the thermo-mechanical properties of the feedstock material, the input energy densities, and the scanning strategies of the heat source [6], [7]. Experimental approaches can be used to characterize the stresses including non-destructive methods such as X-ray diffraction (XRD) and neutron diffraction (ND), and destructive methods such as contour and hole drilling methods. For instance, Chen et al. [8] proposed a residual stresses measurement via XRD technique with pre-treatment using mechanical polishing and chemical etching. Uhlmann et al. [9] presented a system for the realization of in situ synchrotron XRD during LPBF of multi-layer deposition. Observations of arbitrary positions in the sample enabled the measurement of the dynamic evolution of strains and stress during LPBF. Staub et al. [10] measured residual stresses via XRD during LPBF of SS316L. Wang et al. [11] presented a method using ND to validate thermo-mechanical models developed to predict the residual stresses in Inconel 625 walls fabricated by laser directed energy deposition. ND and thermomechanical modelling were stated as complementary techniques for the determination of residual stresses. Wu et al. [12] measured residual stresses using digital image correlation in conjunction with build plate removal and sectioning. The obtained results were compared to the non-destructive volumetric ND technique. It was found that residual stresses were reduced by increasing the island to wall rotation to 45 degrees, increasing energy input, and decreasing the island size [12].

Destructive monitoring approaches such as contour and hole drilling techniques were reported for the measurement of residual stresses [13], [14], [15], [16]. The contour method involves cutting a sample into two parts with subsequent measurement of the resultant deformations [13]. The hole drilling method involves applying strain gauges near the surface of a deposited sample, and the computation of original stresses from the deformations occurred upon the drilling of the hole [14]. Cao et al. [15] measured residual stresses using the hole drilling method in the electron beam PBF of Ti-6Al-4V builds. For the hole drilling method in the sixth layer, only the in-plane stresses at or near the free surface could be obtained. In contrast, ND was able to measure the elastic strain along a transverse line 4 mm below the top surface in the cross-section of the substrate [15]. Barros et al. [16] used hole drilling strain gauge method to measure the residual stresses and direction depth profiles during LPBF of Inconel 718. It was found that in the as-built condition, the residual stresses were approximately half of the yield strength on the top surface and near yield strength on the vertically oriented lateral surface [16]. Note that both non-destructive and destructive methods for stress measurement focus on limited monitoring locations and are time consuming. It is thus enormously challenging to obtain the spatiotemporal variations of the stresses in the entire build via experimental approaches, which necessitate help from numerical simulations.

The finite element analysis (FEA) plays an important role in the study of thermal–mechanical behavior during LPBF. Several studies have been reported on the optimization of process variables for the mitigation of residual stresses [17], [18], [19], [20]. For instance, Mukherjee et al. [17] used an improved numerical approach considering both heat transfer and fluid flow to explore the evolution of residual stresses and distortion during laser DED. It was found that smaller layer thickness resulted in lower residual stresses. Xiao et al. [18] investigated the influences of laser power, scanning speed, and hatch spacing on the evolution of temperature and residual stresses based on the software of Abaqus. Cheng et al. [19] investigated the effects of eight scanning strategies on residual stresses and deformation. It was found that the out-in scanning pattern had the maximum stresses and large directional stresses were generated by using the horizontal line scanning strategy. Zou et al. [20] investigated the effect of scanning strategy on residual stresses using single and multiple lasers. It was observed that the residual stresses were significantly higher as the laser beams increased from one to four with a resultant higher heat input. FEA models have also been used to explore mitigation approaches for the residual stresses using pre-process or post-process treatments. For example, Sharma et al. [21], investigated the development and mitigation of residual stresses in LBPF of Ti-6Al-4V. It was concluded that the maximum value of residual stresses was reduced by 41% with a substrate preheating of 673 K comparing with that without preheating. Ali et al. [22] investigated the effect of re-scanning during LBPF of Ti-6Al-4V, and observed that re-scanning with 150% energy density resulted in 33.6% reduction in residual stresses.

Although many researches have been reported on the study of residual stresses during LPBF, a deeper understanding of the evolution of residual stresses during LPBF is still needed. For instance, the temporal and spatial variations of the temperature and stress fields during the LPBF process need systematic explorations. Thus, the dynamic variations of the stresses during the deposition processes can be revealed, which can provide guidance for the correlated researches such as solidification cracking. To this end, in this study, a 3D finite element model was developed to study the spatiotemporal variations of the temperature and the stresses during multi-track and multi-layer LPBF. Various scanning strategies were used for different layers. The dependences of the residual stresses on the layer-wise variable temperature distributions would be explored. The variations of X-, Y-, and Z-components of stress along different monitoring paths during multi-layer LPBF were systematically examined. The findings of this work provide a better understanding of the temporal development and the spatial distributions of residual stresses during LPBF of Ti-6Al-4V.