Abstract:The wear debris present in a lubricating oil is an imporatnt basis for determining whether that equipment is operating properly and the causes of failure.A new method is developed to use direct laser imaging techniques and advanced neural network specifically to identify the wear debris.This method was combined the information of both particle counting and wear shape classification,it can determine the type,rate of production,and severity of mechanical faults by measuring the size,the size distribution and quantity of wear debris,and also can diagnose the possible root cause of the problem from wear particle and shape classifications.As compared with other methods,this method is fast in testing,which can be widely applied for offline laboratory and oil testing online.The method has already become the newest standard of ASTM D7596.