Hao WenFollow


Doctor of Philosophy (PhD)


Mechanical Engineering

Document Type



Selective Laser Melting (SLM) is a laser powder bed fusion (L-PBF) based additive manufacturing (AM) method, which uses a laser beam to melt the selected areas of the metal powder bed. A customized SLM 3D printer that can handle a small quantity of metal powders was built in the lab to achieve versatile research purposes. The hardware design, electrical diagrams, and software functions are introduced in Chapter 2. Several laser surface engineering and SLM experiments were conducted using this customized machine which showed the functionality of the machine and some prospective fields that this machine can be utilized. Chapter 3 evaluated the effects of laser beam irradiation-based surface modifications of Ti-10Mo alloy samples under either Ar or N2 environment to the corrosion resistance and cell integration properties. The customized 3D printer was used to conduct the laser surface treatment. The electrochemical behaviors of the Ti-10Mo samples were evaluated in simulated body fluid maintained at 37 ± 0.5 ̊C, and a cell-material interaction test was conducted using the MLO-Y4 cells. Laser surface modification in the Ar environment was found to enhance corrosion behavior but did not affect the surface roughness, element distribution, or cell behavior, compared to the non-laser scanned samples. Processing the Ti-10Mo alloy in N2 formed a much rougher TiN surface that improved both the corrosion resistance and cell-material integration compared with the other two conditions. The mechanical behavior of spark plasma sintering (SPS) treated SLM Inconel 939 samples was evaluated in Chapter 4. Flake-like precipitates (η and σ phases) are observed on the 800-SPS sample surface which increased the hardness and tensile strength compared with the as-fabricated samples. However, the strain-to-failure value decreased due to the local stress concentration. γ’/ γ’’ phases were formed on the 1200-SPS sample. Although not fully formed due to the short holding time, the 1200-SPS sample still showed the highest hardness value and best tensile strength and deductibility. Apply machine learning to the materials science field was discussed in the fifth chapter. Firstly, a simple (Deep Neural Network) DNN model is created to predict the Anti-phase Boundary Energy (APBE) based on the limited training data. It achieves the best performance compared with Random Forest Regressor model and K Neighbors Regressor model. Secondly, the defects classification, the defects detection, and the defects image segmentation are successfully performed using a simple CNN model, YOLOv4 and Detectron2, respectively. Furthermore, defects detection is successfully applied on video by using a sequence of CT scan images. It demonstrates that Machine Learning (ML) can enable more efficient and economical materials science research.



Committee Chair

Guo, Shengmin