往復動式圧縮機の故障予兆検出技術に関する研究

抄録

Reciprocating compressors are parts of the most critical systems in many production facilities, and the long-term continuous operation must be required. However, their occurring troubles during operation are not simple but such various phenomena as degradation, wear, etc., are complicatedly intertwined. We focused on vibration characteristics of reciprocating compressors, which could be dominant in the dynamics responses in operation, and constructed the mathematical model to calculate the natural frequencies and modes for crank angles and proposed a method to estimate the degree and the suspicious portion of failure by difference of temporal parameter values, which are identified using measuring data and the mathematical model. In this paper, we apply the proposed method to the field data, where the wear of the connecting part might occur, and examine whether the abnormality detection could be evaluated. For this result, we could show the possibility of monitoring technique observing deterioration with the temporal model parameters determined by measuring data during operation