Metal additive manufacturing (AM) is gaining increasing attention from academia and industry due to its unique advantages compared to the traditional manufacturing process. Parts quality inspection is playing a crucial role in the AM industry, which can be adopted for product improvement. However, the traditional inspection process has relied on manual recognition, which could suffer from low efficiency and potential bias. This study presented a convolutional neural network (CNN) approach toward robust AM quality inspection, such as good quality, crack, gas porosity, and lack of fusion. To obtain the appropriate model, experiments were performed on a series of architectures. Moreover, data augmentation was adopted to deal with data scarcity. L2 regularization (weight decay) and dropout were applied to avoid overfitting. The impact of each strategy was evaluated. The final CNN model achieved an accuracy of 92.1%, and it took 8.01 milliseconds to recognize one image. The CNN model presented here can help in automatic defect recognition in the AM industry.

Multilayer direct laser deposition (DLD) is a fabrication process through which parts are fabricated by creating a molten pool into which metal powder is injected as particles. During fabrication, complex thermal activity occurs in different regions of the build; for example, newly deposited layers will reheat previously deposited layers. The objective of this study was to provide insight into the thermal activity that occurs during the DLD process. This work focused on the effect of the laser parameters of newly deposited layers on the microstructure and mechanical properties of the previously deposited layers in order to characterize these effects to inform proper parameter selection in future DLD fabrication. Varying the parameters showed to produce different effects on the microstructure morphology and property values, leading to some tempering and aging of the steels. The microstructure of the top layer was equiaxed, while the near substrate region was fine dendritic. Typically, both the travel speed and laser power significantly affect the microstructure and hardness. Using the commercial ABAQUS/CAE software, a thermo-mechanical 3D finite element model was developed. This work presents a 3D heat transfer model that considers the continuous addition of powder particles in front of a moving laser beam using ABAQUS/CAE software. The model assumes the deposit geometry appropriate to each experimental condition and calculates the temperature distribution, cooling rates and re-melted layer depth, which can affect the final microstructure. Model simulations were qualitatively compared with experimental results acquired in situ using a K-type thermocouple