Loading...

Monitoring of Valve Leakage in Internal Combustion Engine Using Acoustic Emission Method

Jafari, Mohammad | 2015

917 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 46949 (08)
  4. University: Sharif University of Technology
  5. Advisor(s): Behzad, Mehdi; Mehdigholi, Hamid; Ahmadi, Mohammad Mehdi
  6. Abstract:
  7. Valve leak detection is very important due to bad direct effects on parts life and engine power. Acoustic emission (AE) has a good known ability on leak detection between fault detection methods. The aim of this thesis is presenting the ability of AE method for valve leak detection in internal combustion engines. The research method in a general framework based on the idea that firstly some evaluations were done in static tests for leakage due to three faults including clearance, quasi-crack and notch. Then, the evaluations were continued with dynamic analysis of rotational speeds and different loads on the engine. Finally, the neural network approach was used to automatically detect all kinds of errors and distinguish between normal mode and faulty engines. Static tests were performed on the cylinder head removed from the engine. Static analysis showed: Valve faults were distinguishable using AE and neural network. Also, in the static analysis, a theoretical relationship was extracted between AE and valve leakage characteristics. Dynamic tests were done on the operating engine in the test bench. In dynamic analysis, the three sources of “valve flow”, “valve seating” and “valve leakage” were identified as the main sources of AE signal from the valve situations. Continue to valve fault detection, the AE signals were separated by engine combustion cycle. It was identified in the crank angle and frequency analysis that opening and closing valves have the largest contribution to the formation of AE signals. Results of neural network in the dynamic tests proved the theory of AE usage for valve leak detection. The results showed that both of engine speed and load increasing improved separation between healthy and faulty valve modes
  8. Keywords:
  9. Internal Combustion Engines ; Acoustic Emission ; Artificial Neural Network ; Leak Detection ; Valve Leakage ; Valve Faults

 Digital Object List

 Bookmark