Integrated Solution for Cutting Tool Health Monitoring in CNC Mills
Guardat en:
| Publicat a: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Autor principal: | |
| Publicat: |
ProQuest Dissertations & Theses
|
| Matèries: | |
| Accés en línia: | Citation/Abstract Full Text - PDF |
| Etiquetes: |
Sense etiquetes, Sigues el primer a etiquetar aquest registre!
|
| Resum: | In the realm of modern manufacturing, Computer Numerical Control (CNC) machines are known for high precision, efficiency and versatility for manufacturing. Despite their great accuracy and versatility, CNC machines frequently require extensive maintenance due to the wear of their cutting tools. If tool wear is not handled immediately, it can result in inaccurate mill components and expensive machine downtime. Conventional tool maintenance approaches, which rely on planned replacements and recurrent inspections by experts, are unsustainable because they either result in premature replacement of components or worn tools that go unreported between examinations. Improper cutting tool maintenance will raise machine operational costs and lead to reduced productivity.The goal of this work is to develop an integrated solution for monitoring and predicting cutting tool health in multi-axis CNC milling machines. The proposed solution leverages vibration data obtained from accelerometers within the machine enclosure to develop a sensor-based method for CNC machine tool health prediction. A modular sensing unit is developed as part of this work, which can be integrated onto the vise of milling machines holding the machined part, and capture vibration and tool movement data. A complete framework is also proposed to process the collected data and to extract important information on tool wear and the machine’s operational condition. In order to categorize the cutting tool health and predict associated maintenance work, the collected vibration data is processed using Principal Component Analysis (PCA) and then given as input to Support Vector Machines (SVM) to classify the tool condition. With the help of this, we aim to reduce unplanned machine downtime and achieve more effective maintenance programs by anticipating tool wear.The integrated cutting tool health monitoring solution proposed in this thesis addresses limitations in traditional tool monitoring by introducing accessible and automated monitoring capabilities of milling conditions, and predictive classification of cutting tool health/wear. The integration of machine learning algorithms enables the prediction of tool wear conditions from the collected measurement data with high accuracy, minimizing false and missed detections. The proposed system achieved a classification accuracy of 89.28% using an SVM with RBF kernel, with minimal false positives and negatives, confirming its robustness in distinguishing between good and worn tools in practical industrial applications. The proposed end-to-end solution may then offer CNC machine operators a significant reduction in unscheduled machine interruptions, while extending the life of their tools and improving overall machine efficiency. |
|---|---|
| ISBN: | 9798286491629 |
| Font: | ProQuest Dissertations & Theses Global |