Register now After registration you will be able to apply for this opportunity online.
Monitoring and Optimization of Grinding Process using Machine Learning
Gears are the backbone of the aircraft transmission systems, facilitating critical power transfer and speed adjustments for various components. Their flawless operation is the key to the seamless flight of lightning-fast fighter jets, commercial airliners, and agile helicopters.
As the demand for precise, robust, and dependable gears continues to rise, addressing production bottlenecks becomes increasingly crucial in meeting aviation's evolving needs. A significant contributor to these bottlenecks is the grinding operation, the final step in production. In the pursuit of achieving higher production efficiency by pushing the limits of process parameters, grinding burns often appear on the workpiece surface. These grinding burns are undesirable. They occur due to excess heat accumulation which impacts residual stresses and metallurgical structures, potentially leading to component cracks and failures. The rejection of such workpieces wastes time and resources, ultimately diminishing production efficiency.
The primary focus of this thesis is the early detection of grinding burns within the manufacturing process, achieved through the utilization of acoustic emission and current sensors in combination with critical process parameters. A mathematical model will be developed based on machine learning modeling techniques combining the sensors data, process parameters and evaluated surface quality. The objectives are prediction of grinding burn, and subsequently, optimization of the grinding process for higher productivity.
Not specified
Tasks:
• Literature review on grinding process, grinding burn & process optimization using ML
• Conduct experimental campaign for generation of grinding burns on our state-of-the-art grinding machine Studer S31
• Further development of in-house LabVIEW program for data acquisition.
• Detection of grinding burns using Barkhausen Noise Analysis and nital etching
• Development of data pipelines, ML model development and model optimization
• Final report and presentation in English
Tasks: • Literature review on grinding process, grinding burn & process optimization using ML • Conduct experimental campaign for generation of grinding burns on our state-of-the-art grinding machine Studer S31 • Further development of in-house LabVIEW program for data acquisition. • Detection of grinding burns using Barkhausen Noise Analysis and nital etching • Development of data pipelines, ML model development and model optimization • Final report and presentation in English