In this study, cross-section analysis was performed on a novel rotating direct-metal deposition method capable of preliminary surface treatment and damage repair of cylindrical inner walls. The cross-sectional shape, microstructure, and metallurgical composition were analyzed to verify feasibility. No defects such as porosity or cracks were found in the cross section, but asymmetric dilution was observed because of the non-coaxial powder nozzle. Microstructural coarsening was confirmed over a higher dilution area by high-magnification optical microscope images. As the dilution ratio was increased, hard carbides in the dendrite were bulk-diffused into inter-dendrite spaces, and the toughness was lowered by Fe penetration into the deposit. Therefore, the increased laser heat input can be modulated to the typical dilution by decreasing the laser scanning velocity.

Poppet valves made from high-frequency heat-treated SUH3 steel have insufficient durability, and scratches appear on the valve face in prolonged use. It is necessary to develop surface treatment technology with excellent durability to prevent the deterioration of engine performance. Therefore, a surface treatment technology with higher abrasion resistance than existing processes was developed by direct metal deposition to the face where the cylinder and valve are closed. In this study, heat pretreatment and deposition tests were performed on three materials to find suitable powders. In the performance evaluation, the hardness, friction coefficient, and wear rate were measured. Direct metal deposition using Inconel 738 and Stellite 6 powders without heat pretreatment were experimentally verified to have excellent durability.

Deposition defects like porosity, crack and lack of fusion in additive manufacturing process is a major obstacle to commercialization of the process. Thus, metallurgical microscopy analysis has been mainly conducted to optimize process conditions by detecting and investigating the defects. However, these defect detection methods indicate a deviation from the operator’s experience. In this study, artificial intelligence based YOLOv3 of object detection algorithm was applied to avoid the human dependency. The algorithm aims to automatically find and label the defects. To enable the aim, 80 training images and 20 verification images were prepared, and they were amplified into 640 training images and 160 verification images using augmentation algorithm of rotation, movement and scale down, randomly. To evaluate the performance of the algorithm, total loss was derived as the sum of localization loss, confidence loss, and classification loss. In the training process, the total loss was 8.672 for the initial 100 sample images. However, the total loss was reduced to 5.841 after training with additional 800 images. For the verification of the proposed method, new defect images were input and then the mean Average Precision (mAP) in terms of precision and recall was 0.3795. Therefore, the detection performance with high accuracy can be applied to industry for avoiding human errors.