**2.1 Defects in PBF build parts**

Several common defects are encountered in metal PBF processes, which lead to weakening of mechanical properties of the build part. These defects occur during fabrication and/or post -processing operation. The presence of defects limits the industry-wide spread of this technology as a result of insufficient repeatability, reliability, and precision. Such defects can be classified and analyzed based on how they affect the printed part, how they generate, how common and significant they are for the overall quality of the build part etc. In addition to that, different process

*Multiscale Modeling Framework for Defect Generation in Metal Powder Bed Fusion Process… DOI: http://dx.doi.org/10.5772/intechopen.104493*

parameters influence the generation and propagation of different anomalies in the print. Many studies have been conducted to identify the defects in AM processes [3–5].

In a recent work [6], the common defects in additive manufacturing have been classified and reviewed on the basis of geometry, surface quality, microstructure, and mechanical properties. For example, on the basis of geometry and dimension, there can be (a) geometry inaccuracy (form deviation) and (b) dimension inaccuracy (size deviation). Again, common surface quality related defects are surface roughness, balling, surface oxidation etc., while anisotropy, heterogeneity, porosity etc. are microstructure related defects found in 3D printed builds. A comprehensive classification is shown in **Figure 1** [6].

### **2.2 How do process parameters affect printing defects**

The significant process parameters in the metal powder bed fusion (also known as selective laser sintering process) such as laser power, scanning speed, hatching distance, scanning strategy etc. affect the generation of printing defects. It is crucial to identify the combination of these parameters to obtain the required level of quality of the product. For instance, higher laser power increases the thermal shrinkage, and higher scanning speed hatch spacing lowers the thermal shrinkage [7]. The laser power and scanning strategy contributed to the temperature variation, that leads to non-uniform shrinkage in a particular layer [8]. Part weight, build chamber temperature, cooling rate, layer thickness and material can affect thermal shrinkage in a way that shrinkage decreases with increasing layer thickness, part bed temperature, and interval time [9–11].

Surface roughness depends on numerous parameters, some of which can be controlled, while few others are not controllable from the designer and operator's perspective. For lower scan speeds, the average roughness decreases with increase in speed, while in the higher speed range, it remains constant [11–16]. Moreover, warping and distortion that impact the surface quality, are highly dependent on thermal phenomena during the printing process. In addition to laser power and scanning speed, the thermal gradient between scanning zones can impact the quality of each fabricated layer. Surface geometry and fundamental geometric features of orientation, thickness etc. also have impact of geometric errors and surface quality [17]. Smaller hatch spacing seems to be beneficial in this context as it induces gradual temperature increase of powder bed and slower cooling rate. In addition to this, the initial powder spreading in LPBF also influences the layer quality, and thus consequentially impacts the porosity of produced parts [18].

**Figure 1.** *Defects in metal PBF.*

Typical manufactured components using traditional manufacturing methods (milling, drilling, surface grinding etc.) comes from a solid building block of material and they are not porous unless porosity is induced by design. But in additive manufacturing, porosity of the build parts is a very common occurrence. Poor wetting, powder packing density, gas flow condition, entrapped gas etc. play vital roles behind the unwanted porosity of printed parts.

Among the numerous process parameters that contribute to defect generation, most vital ones are laser power, laser scanning speed, laser spot size, powder size, layer thickness, external pressure and material's absorptivity. It is important to understand how the process parameters impact the generation of defects and what are the signatures that relate to the defects. The process-structure–property (PSP) relationships have been under discussion and research works are published earlier as well [19–21]. In recent years, use of machine learning and smart manufacturing has started to be used in AM field [22–31], which also provides inspiration to use that for our research goal.
