M.Tech (Department of Electronic Engineering, Faculty of Engineering and Technology), Vaal University of Technology / In the world of research and development, the ability to rapidly manufacture a prototype or part has become a significant part of the manufacturing process. This requirement has given rise to some unique manufacturing technologies. One of these technologies is Additive Manufacturing (AM), or also more commonly known as 3D printing. There are several AM technologies available and can be divided into three major AM categories namely: liquid, powder and solid sheet based. For this research study, the primary focus will be on powder-based technologies. Powder-based technologies make use of materials in powder form and use different fusion techniques to fuse the powder particles together. All the powder bed fusion technologies consist of the same basic components, namely a powder chamber, build chamber, re-coater and a powder fusion system. For each layer of the build, the re-coater applies a new layer of powder from the powder chamber to the build chamber, and then the specific type of powder fusion system will fuse the powder particles together. This process will then be repeated until the entire build has completed.
Currently, powder bed fusion AM platforms do not have re-coating quality feedback into the printing system. Thus, when errors or defects occur on the powder bed surface during the re-coating process, they can affect the structural integrity of the parts. Parts must then be reprinted, which becomes costly due to wasted raw materials, electricity and time. Raw material and sundry wastage was some of the key factors that reduces the overall efficiency of the identified AM technology. Due to the increased problem with wasted materials, the need arose to develop a re-coater monitoring system, which could be used to increase the overall efficiency of a powder-based system.
For the development of a re-coater monitoring system, a review of three different types of monitoring technologies such as computer vision, laser scanning and a time-offlight camera was conducted. Based upon the relatively low cost, low computer resource requirements and high accuracy, computer vision was considered as the best suited technology for development of the monitoring system. To select the correct camera to capture images of the powder bed, the required specifications for the camera, lens and mounting position were determined mathematically. A software program was then developed to autonomously detect re-coating errors on the captured image after each re-coating cycle using image processing techniques. Each of the captured powder bed images were divided into 16 equal sized quadrants, where each quadrant was processed individually. Each of the quadrants was examined using an edge detection algorithm to detect any changes in contrast that would indicate a defect or re-coating error. The probability of a possible re-coating error or defect was calculated for each quadrant and displayed as a percentage value.
The active re-coater monitoring system was also integrated into the Voxeljet VX500 to validate the system’s operation. The system was used to monitor a total of seven build jobs on the Voxeljet VX500. However, the first three build jobs could not be successfully monitored as some parameters of the system had to be re-adjusted to ensure proper operation. The last four build jobs were monitored successfully and recorded results that proved that the active re-coater monitoring system could indeed detect defects and re-coating errors when they occurred.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:vut/oai:digiresearch.vut.ac.za:10352/496 |
| Date | 05 1900 |
| Creators | Du Rand, Francois |
| Contributors | Van Tonder, P. J. M., Dr., Pienaar, H. C. v. Z., Prof. |
| Publisher | Vaal University of Technology |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0016 seconds