• Problems Existing in the Fault Diagnosis Technology for Valves

Services

Problems Existing in the Fault Diagnosis Technology for Valves

  • Mar 30, 2022
  • Hits: 5
Detection accuracy for the valve state parameter is not high.
Measuring the state parameters of the valve is the basis for valve fault diagnosis, but some detection methods and equipment currently have the problem of insufficient detection accuracy due to the limitation of the detection principle or detection environment. For example, when using a clamp-type thrust sensor (C-Clamp) to measure the thrust of the gate valve stem, the maximum thrust measurement error can reach 20%, and this error often exceeds the margin reserved for the output force of the valve actuator.  Obviously, the reliability of judging the future opening and closing performance of the valve by the measurement results of the valve stem thrust is poor. Another example is when using acoustic emission equipment to detect the internal leakage of the valve, once the pressure difference on the valve is small and the medium impacting the valve body cannot generate a stress wave signal, much internal leakage of the valve cannot be detected.
 
The diagnosis for the weak early failure of the valve is difficult.
The diagnosis of valve faults must first obtain various state signals when the valve is running, and gets various characteristic information from the signals. The sign of failure is obtained by analyzing the characteristic information. However, the diagnosis and analysis methods of the valve are all used for the situation that the valve has an obvious failure during the operation at present.  For some early, weak and complex faults, there is still no effective diagnostic analysis method. There is still a lack of relevant research on the reliability of the valve fault diagnosis and analysis method, which is also the reason why the valve fault diagnosis and analysis method is rarely used in practice. In addition, some valve failures are not caused by a single cause, but are caused by the coupling of multiple causes. Moreover, the existing diagnostic methods are almost all carried out for a single failure mode. Therefore, for mixed and systematic valve faults, the existing diagnostic methods cannot effectively and reliably trace the cause of the fault.
 
The research on the failure mechanism of the valve is not profound
The failure mechanism of the valve refers to the physical, chemical, mechanical, electrical or human causes and causal relationships that cause valve failures. At present, there are many engineering examples for the analysis of valve failure modes, but there is little research on the causes of these symptoms. The study of valve failure mechanisms needs a lot of basic theoretical research and practical engineering data support. Generally, it is necessary to establish a basic mathematical model of the failure mechanism, and to modify and improve it through the actual failure data. However, because there are many types of valves, and the working conditions are very different, it is almost impossible to obtain comprehensive fault data for all valves. Therefore, it is an effective method to analyze the failure mechanism of valves through simulation.

The intelligence of the diagnostic and decision-making system needs to be improved.
The accumulation of actual valve failure data is relatively small due to the lack of research on valve failure mechanisms. Therefore, the basic knowledge used by the valve failure diagnosis system for reasoning and judgment relatively lacks, which also leads to the need for human participation in the fault diagnosis of the valve. The current expert systems used for diagnosis and decision-making lack sufficient knowledge representation, and the reasoning efficiency is low. However, the artificial neural network system cannot obtain enough training samples and needs to set some conditions and parameters, resulting in the practical application being limited. Therefore, the underlying fault mechanism and the lack of fault data restrict the intelligence of the above-mentioned diagnostic methods.