Parameter Estimation of the Thermal Network Model of a Machine Tool Spindle by Self-made Bluetooth Temperature Sensor...

Area of Science:

• Mechanical Engineering
• Thermal Analysis
• Manufacturing Technology

Background:

• Machine tool spindle failures can arise from thermal issues.
• Accurate thermal characteristic analysis is crucial for spindle reliability.
• Predicting spindle thermal behavior under varying conditions is challenging.

Purpose of the Study:

• To develop a robust Thermal Network Model (TNM) for machine tool spindles.
• To characterize both steady-state and transient thermal behavior.
• To enable real-time temperature prediction for smart manufacturing.

Main Methods:

• Development of a Bluetooth Temperature Sensor Module (BTSM) for data acquisition.
• Derivation of heat transfer characteristics based on heat transfer theory and empirical formulas.
• Grey-box estimation combining theoretical models with experimental data to build the TNM.
• Application of Model Order Reduction (MOR) for creating a reduced-order TNM for edge computing.

Main Results:

• The TNM accurately predicts spindle temperature with 99.5% agreement.
• The developed BTSM offers high precision (±(0.1 + 0.0029|t|) °C) and low power consumption (7 mW).
• The modeling approach is validated as robust, reliable, and transferable to similar spindle structures.

Conclusions:

• The presented TNM and parameter estimation scheme effectively characterize machine tool spindle thermal behavior.
• The methodology provides a reliable tool for temperature prediction, enhancing operational safety.
• The reduced-order TNM is suitable for real-time implementation in edge computing for smart manufacturing.