A DESKTOP STUDY OF MILL ROLLER LIFE PREDICTION USING CUMULATIVE FATIGUE DAMAGE THEORY

摘要

Failed roller shafts cause major problems for sugar mills in terms of downtime and lost production. Due to the typically catastrophic nature of these failures, associated equipment such as brasses, mill cheeks, pinions and tail-bars can also be damaged as a result of the primary roller failure. Therefore, it is logical to determine which rollers are most likely to fail so that failure can be mitigated. The tracking of in-service mill rollers to manage non-destructive testing (NDT) programs and failures is by no means a new concept. In this case, a desktop study has been conducted that is based on predicting the relative life of in-service roller shafts using the Miner's rule cumulative fatigue damage theory. Using SRI literature and mill operating parameters, peak alternating bending stress levels were determined for each roller position within a candidate 5-mill train. Tonnage and mill speed data were utilised to identify the rates of 'stress cycling' at each mill roller position over the historyof existing in-service mill roller shafts. For each mill roller shaft currently in use, the service history (i.e. durations spent at different positions in the milling train) was then used to determine the 'usage' of theoretical fatigue life. The resultwas a relative measure of failure likelihood for use in maintenance strategy development into the future. The focus of this paper is on describing the development of the numerical tool and comparing its life predictions with those made purely upon length of service. In addition, the concept of extending the tool to develop location strategies for optimising mill roller service life is discussed.