Detecting Piston Ring Instability with Engine Vibration Analysis, 08-R8342
Craig M. Wall
Inclusive Dates: 10/01/12 – 04/01/14
Background — Engine vibration analysis is a potentially lucrative capability that SwRI is well-positioned to exploit, given the amount of engine testing done annually. In this project, a difficult task was chosen to highlight the sophistication of SwRI sensor technology. Specifically, piston ring instability is a low-energy event that has profound effects on engine performance and emissions. Detecting ring instability without using invasive sensors means subtle mechanical processes can be monitored and recorded without disassembling a client’s engine. That is, vibration recording and analysis can be offered as a value-added service at any point during testing, and offers a significant form of analytical insurance when unexpected events require post-test analysis. Other potential benefits include the possibility of developing ring performance monitors that can advise an engine control unit in real time to adjust engine operation and avoid high emission states. This is particularly welcome as engines age, as this capability may well reduce both lifetime emissions and the after-treatment packages required to suppress pollution.
Approach — The initial sensor suite consisted of a combustion pressure transducer in the cylinder head, a crankcase pressure transducer for blow-by gas pulses, and an accelerometer attached to the engine block with an optical encoder referencing crankshaft angle. These were installed on a 500cc single cylinder spark ignition gasoline research engine. Piston ring instability was first sensed by monitoring blow-by gas flow from the crankcase, which rises when the piston rings fail to seal against the combustion gases. This project was not an engine test per se, but was conducted to develop sensing methodology. We used an engine running with a modified piston as a “signal generator” to provide mechanical vibrations for optimizing a non-invasive sensor suite. The initial suite was eventually reduced to only the accelerometer and crank angle encoder. A simple analog device was also developed, which easily displays the vibration information in real time: the accelerometer on the engine block was used to drive an LED light source that illuminated a rotating disk on the camshaft. A mark on the disk was “strobed” similar to the effect of an ignition timing light, but in this case the vibrations are substituted for the ignition event and appear at the characteristic crank angles shown in the accompanying plot.
Accomplishments — The technique we evolved used a piston with excessive clearance in the top ring groove. We have managed to arrive at a convenient experimental method that allows us to “turn on and turn off” piston ring instability by independently adjusting rpm and power. Increased rpm decreases ring stability by “throwing” the ring off the lower ring groove seating surface, while increasing power increases ring stability by increasing the gas pressure that holds the ring down against the inertial forces unseating it. Within very narrow limits, we can easily produce instability or stability by a simple turn of a control potentiometer. This is our key ability that provides an opportunity to develop real-time ring stability sensors and eventually an onboard engine control capability to avoid high emission states, even in worn engines. The sensor suite may ultimately consist of no more than the accelerometer attached to the engine block and a crankshaft angle encoder, which can be installed on a client engine in minutes. The following illustrates the shift in energy signature when a transition from stable to unstable ring operation occurs in the middle of a five-second accelerometer recording. These energies are plotted as “G” forces versus crankshaft angle and are the result of repeatedly interrupted high pressure gas escaping past the piston rings as increased blow-by.