Date of Award
Electrical and Computer Engineering
In industry fiber degradation during processing of biocomposite in the extruder is a problem that requires a reliable solution to save time and money wasted on producing damaged material. In this thesis, We try to focus on a practical solution that can monitor the change in temperature that causes fiber degradation and material damage to stop it when it occurs. Ultrasound can be used to detect the temperature change inside the material during the process of material extrusion. A monitoring approach for the extruder process has been developed using ultrasound system and the techniques of machine learning algorithms. A measurement cell was built to form a dataset of ultrasound signals at different temperatures for analysis. Machine learning algorithms were applied through machine-learning algorithm’s platform to classify the dataset based on the temperature. The dataset was classified with accuracy 97% into two categories representing over and below damage temperature (190oc) ultrasound signal. This approach could be used in industry to send an alarm or a temperature control signal when material damage is detected. Biocomposite is at the core of automotive industry material research and development concentration. Melt mixing process was used to mix biocomposite material with multi-walled carbon nanotubes (MWCNTs) for the purpose of enhancing mechanical and thermal properties of biocomposite. The resulting composite nano-bio- composite was tested via different types of thermal and mechanical tests to evaluate its performance relative to biocomposite. The developed material showed enhancement in mechanical and thermal properties that considered a high potential for applications in the future.
Elseddawy, Ahmed, "Developing of Ultrasound Experimental Methods using Machine Learning Algorithms for Application of Temperature Monitoring of Nano-Bio-Composites Extrusion" (2018). Electronic Theses and Dissertations. 7429.