Lubricant induced pre-ignition in an optical spark-ignition engine

Abstract

This work focuses on the introduction of lubricant into the combustion chamber and the effect that this has on pre-ignition. Apparently for the first time, the work presented provides detailed full-bore optical data for lubricant induced pre-ignition and improves understanding of the super-knock phenomena that affects modern downsized gasoline engines. A new single-cylinder optical research engine was designed, fabricated and installed. The research engine featured poppet valves and a unique four-valve layout. At the same time, complete optical access to the piston crown and top-land crevice region was maintained in a disc-shaped combustion chamber. The engine featured a novel optical window clamping arrangement to withstand peak knocking pressures in excess of 120bar; a capability that is believed to be unique amongst full-bore optical engines at present. A system was designed for deliberately introducing a controlled sample of lubricant into the hottest region of the combustion chamber. It was found that, through injecting a small sample of lubricant (~3.2μl roughly equivalent to 0.1 air-to-fuel ratio) during the compression stroke, a strong pre-ignition response could be induced. A total of 18 lubricants were tested thermodynamically and their pre-ignition frequencies measured via analysis of in-cylinder pressure data. It was found that the occurrence of pre-ignition varied greatly depending on the lubricant injected, with pre-ignition frequencies ranging from 10% to 83%. The measured pre-ignition frequency of a lubricant was found to be inversely proportional to the lubricant density, with a coefficient of determination of 0.91. Calcium and magnesium based detergents were not found to have any measureable effect on pre-ignition frequency. Equally, anti-oxidant additives (including a novel, high-temperature anti-oxidant) were not found to have any effect on pre-ignition frequency in the concentrations considered. Image sets were captured for the deliberate introduction of two different lubricants. Images were captured via the natural light emission from combustion. It was found that pre-ignition most commonly initiated in the area surrounding the active exhaust valve head and resulted in a deflagration that caused the combustion phasing to be advanced. With 97RON gasoline, pre-ignition was rarely sufficiently advanced for end-gas auto-ignition and knock to occur. Image sets were also captured for the natural introduction of lubricant from the top-land crevice subsequent to high-intensity end-gas knock. High-intensity knock was achieved by fuelling with a low octane gasoline. Droplets of lubricant were observed moving within the main charge and causing pre-ignition in cycles subsequent to the natural release of lubricant. A total of 11 lubricant droplets were observed causing pre-ignition within the main charge. Naturally released lubricant droplets were found to survive within the combustion chamber for multiple cycles and lead to a corresponding “on-off” knocking combustion pattern that has been so widely associated with super-knock in real downsized spark ignition engines.