Inline Monitoring
Wednesday, June 30, 2021 2:00 PM to 2:20 PM · 20 min. (Africa/Abidjan)
Stream#3
Information
Contributors:
Abstract:
The thesis 'Inline monitoring of titanium AM based on image processing of meltpool and powderbed data' aims to analyze the suitability of monitoring images from the additive manufacturing of Ti-6Al-4V for a defect detection through image based processing. The detection of defects during the powder bed fusion process optimizes the process chain by a reduction of time and costs that are required to meet and validate the high quality standards in aerospace industry. Several image processing methods are applied for the analysis of the generated meltpool as well as powder bed images. The data is prepared and subsequently validated through the data of a reference part as well as based on a mathematical definition of the defects. The generated reference defects can be detected by the applied methods from a certain size upwards, where the monitoring images are differently suited. Beside the unsuitability of the powder bed images a deviation of the laser power is mainly reflected in the meltpool images of the photodiode and a deviation of the amount of molten powder is mainly reflected in the meltpool images of the camera. Therefore, a further development of the applied methods and the consideration of additional reference defects are reasonable and required for the qualification of a quality control through monitoring images.
Abstract:
The thesis 'Inline monitoring of titanium AM based on image processing of meltpool and powderbed data' aims to analyze the suitability of monitoring images from the additive manufacturing of Ti-6Al-4V for a defect detection through image based processing. The detection of defects during the powder bed fusion process optimizes the process chain by a reduction of time and costs that are required to meet and validate the high quality standards in aerospace industry. Several image processing methods are applied for the analysis of the generated meltpool as well as powder bed images. The data is prepared and subsequently validated through the data of a reference part as well as based on a mathematical definition of the defects. The generated reference defects can be detected by the applied methods from a certain size upwards, where the monitoring images are differently suited. Beside the unsuitability of the powder bed images a deviation of the laser power is mainly reflected in the meltpool images of the photodiode and a deviation of the amount of molten powder is mainly reflected in the meltpool images of the camera. Therefore, a further development of the applied methods and the consideration of additional reference defects are reasonable and required for the qualification of a quality control through monitoring images.
