Defects due to Improper Design or Job Preparation
Despite the continuous improvements of software tools for AM that support the designer's decisions and drive optimised build design and preparation settings, most design and process choices are still based on the operator's experience. Moreover, the high levels of complexity of additively produced parts often impose compromises or trial-and-error choices that can result in defects and poor part quality. One critical source of defects can be the improper design of the supports, which are needed to avoid geometrical distortion due to thermal stresses and to optimise cooling rates in overhangs or bridges (see Chapter 3). Another possible source of defect can be an improper part orientation within the build volume.
Process Setting–Induced Defects
Defects in PBF can often be due to the incorrect selection of process parameters and scanning strategies. Indeed, process parameters (for example, laser power, scan speed and layer thickness) and scan strategies (for example, hatching scanning path and contour scanning path) are not only material dependent but also geometry dependent, and complex products manufactured via PBF are usually characterised by critical geometrical features, such as overhang regions, thin walls, acute corners and other complex features, which would require locally variant process parameter settings. However, almost all commercial PBF systems allow the process parameters to be varied from one part to another, but not within the same part. This makes the selection of optimal parameters and scan strategies a compromise choice that typically does not guarantee first-time-right production. The scan strategy influences the temperature distribution over the slice, and inadequate strategies may increase residual stresses and porosity. In L-PBF, improper scan strategies may also inflate the generation of super-elevated edges and affect the microstructural properties of the part. Experimental 'process mapping' is usually carried out to identify the process- ability window, i.e. the set of appropriate process parameters and the scan strategy that corresponds to appropriate part density and quality for a given material. Such windows are usually quite narrow, which highlights the difficulty in keeping optimal process conditions for any possible geometry.
The energy density plays an important role in determining the residual stresses in the part and the powder wettability, together with the number and properties of process byproducts, i.e. spatter and plume emissions. Spatter is either powder particles blown away during the laser scan of the part or liquid material ejected from the melt pool as a result of unstable solid-liquid transitions (Liu et al. 2015, Khairallah et al. 2016, Ly et al. 2017). The plume is formed by the partial material vaporisation (King et al. 2014) and differs from the surrounding atmosphere in terms of chemical composition, temperature and pressure. When spatter falls on the powder bed, it can produce contamination, as it is characterised by larger size, more irregular shapes and different chemical compositions compared to loose powder particles. When the beam passes over the deposited spatter, the larger size of the spatter may prevent complete melting, which can result in formation of voids within the part (Khairallah et al. 2016). On the other hand, if the spatter is displaced by the recoating device, particle dragging may occur, with a consequent irregularity of the powder bed deposition (Foster et al. 2015). Besides spatter, the material vaporisation and plume formation can interfere with the optical properties of the processing beam path by altering the beam profile and the local energy density. EB-PBF differs from L-PBF both in terms of controllable parameters and process by-products. Also, in EB-PBF, non-optimal process parameters may lead to unstable process conditions and various kinds of defects in the final product.