MILLING CUTTER TEMPERATURE PREDICTION WITH MACHINE LEARNING MODEL TO SUBSTITUTE TEMPERATURE SENSOR

• Main Author: Dwi Nugraha, Wahyu
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/68491
• Institution: Institut Teknologi Bandung

Mechanical machines are very important things in industry, especially manufacture. These machines have function to do machining operation to workpiece which is target of production. To do machining process on workpiece, is needed an important tool called cutter. In every machining process, cutters have different characteristics based on operation goal. In machining process, cutter temperature is one of important factor need to be observed. The reason is because high cutting temperature has effect on cutter quality decreasing such wear, crack, and chips be sticked to cutter. These effects can decrease cutting accuracy and which able to lead decreasing product quality. To prevent high temperature rising on cutter, it’s necessary to do observation periodically during machining operation. In this condition, temperature sensor has several problems, such can’t use to observe temperature while cutter in the middle of workpiece or on any side of workpiece which can’t reachable by sensor. Based on that several problems, this final project has goal to create machine learning model which able to substitute temperature sensor by predict cutter temperature while machining operation based on it’s relationship with several machining parameters. There are several models have been created. Based on that models, is selected one best model with low errors tested with test data and also low errors tested by validation data with different material dimension.