Fault Diagnosis on Double-nut Ball Screw System Based on Preload Sensor

Abstract

A novel diagnosis strategy for double-nut ball screw system is proposed Most traditional diagnosis approaches are based on using accelerometers to measure ball pass frequency from ball screw-nut Severe signal contamination might occur because of noises from servo motor or other machineries in factories Therefore traditional approaches usually suffer from low accuracy The methodology employed by this thesis is to use a preload sensor developed by Industry Technology Research Institute (ITRI) for double-nut ball screw failure diagnosis To clarify how variation of preload affects the drive system this thesis at first builds up a dynamic model and discusses the relationship between preload and stiffness of each component in the test rig Secondly a test environment with constant temperature is set up to reduce the impact of heat expansion Thirdly different sizes of smaller steel balls are used to imitate the preload loss situation and the crack of recirculating mechanism is made by laser machinery The results of preload sensor show that: (i) the threshold of preload is 1981 N; (ii) additional variation of preload can be measured as the failure occurs in recirculating mechanism compared with normal recirculating mechanism characteristic frequency moves from 2 Hz to 1 8 Hz and the signal strength (by volts) of characteristic frequency is dropped by 25%~70%; (iii) more experimental test is undertaken on another test rig without lubrication and the results shows that the trend of preload variation is as same as the first test rig From the intensive experimental tests the diagnosis strategy in this thesis can be potentially applied to the real-world machine tools

Date of Award	2020
Original language	English
Supervisor	Nan-Chyuan Tsai (Supervisor)