Detection of fuel type on a spark ignition engine from engine vibration behaviour

Abstract

An experimental investigation of the vibration behaviour of a spark ignition engine fuelled with gasoline, ethanol and methanol blends is presented in this paper. Three types of fuel were used, namely: unleaded gasoline as a base fuel for comparison and two alcohols, namely methanol and ethanol, blended with gasoline at volume percentages of 10%, 20% and 30%. The engine was tested at 1000, 1300, 1600 and 1900 rpm. The experimental results indicated that the vibration amplitude produced by the gasoline fuel was higher at low frequencies around 1500 Hz. On the contrary, the vibration amplitude generated by the gasoline–methanol blend had a high frequency content around 2400 Hz. The gasoline–ethanol blend gave intermediate results at both frequencies. An algorithm was constructed that reliably identified the type of fuel used irrespective of the engine rotational speed.

Introduction

Bio-fuels can be used for power generation in a number of ways. One possibility is to produce liquid or gaseous bio-fuels that can be burned in engines that normally burn fossil fuels. Bio-fuels also include alcohols, such as, ethanol and methanol. Sucrose and starch can be converted into ethanol by hydrolysis and fermentation. There is also the possibility of converting sugar from whey and starch, or sugar from wastes, into ethanol by fermentation [1], [2]. Ethanol has a lower calorific value and is both hygroscopic and corrosive. However, in practice, ethanol may do well as a fuel in spark ignition (SI) engines, either as pure ethanol, or when used as a blend with fossil fuels [3]. Methanol can be produced via synthesis gas from clued glycerine-derived syngas or biomass with acceptable energy cost [4]. In addition, methane can be directly converted into methanol [5]. Methanol, in turn, can be blended with gasoline up to 15% by volume and used as fuel in SI engines without major modifications required.

Gasoline–alcohols blends have been a subject of research in the 1980s and it has been proven that ethanol and methanol gasoline blends were technically acceptable for existing SI engine technology. There is a considerable amount of literature regarding various ethanol and methanol blends with gasoline. El-Kassaby [6] investigated the effect of gasoline–ethanol blends on spark ignition engine performance. Abdel-Rahman and Osman [7] carried out performance tests using different percentages of ethanol in gasoline fuel, up to 40% by volume, under variable compression ratios. Al-Hasan [8] showed that blending unleaded gasoline with ethanol increased the brake power, torque, volumetric and brake thermal efficiencies and fuel consumption, while it decreased the brake specific fuel consumption and equivalence air–fuel ratio. Wu et al. [9] tested ethanol–gasoline fuel in a conventional engine under various air–fuel equivalence ratios (λ) regarding engine performance and emissions. Liu et al. [10] showed that the engine power and torque decreased with increasing percentage of methanol in the fuel under wide open throttle conditions.

The effects of using ethanol and methanol unleaded gasoline blends on emissions have been investigated by a number of researchers. Palmer [11] showed that 10% of ethanol addition to gasoline could reduce the concentration of CO emissions up to 30%. Bata et al. [12] have tested different blend rates of ethanol gasoline fuels in engines, and found that ethanol could reduce the CO and UHC emissions. Taylor et al. [13] tested four different alcohol–gasoline fuels and concluded that ethanol addition can reduce CO, HC and NO emissions. Ceviz and Yüksel [14] investigated the effects of ethanol in unleaded gasoline blends regarding cyclic variability and emissions of a spark-ignited engine.

While the effects of using alcohols on spark ignition engine performance and emissions have been thoroughly investigated, very little research has been done on the engine vibrational behaviour. Othman et al. [15] investigated the vibrations produced by different types of fuel, such as, kerosene, gas, oil, methanol, and methanol–kerosene blend, using a simple turbo-shaft gas turbine engine with a free turbine driving dynamometer. The vibration patterns over the range of operating conditions were obtained with a real time spectrum analyser. It was found that the combustion driven oscillations increased with the carbon–hydrogen proportion in the fuel. The frequency and amplitude of the fundamental harmonics decreased with an increase in load. It was shown that the combustion-induced vibration levels in an engine could be predicted for off-design speeds and fuels; this could be useful for design and diagnostic purposes. Ajovalasit and Giacomin [16] studied the variations in diesel engine idle vibration caused by fuels of different composition and also their contribution to the steering wheel vibration. The time-varying covariance method and time–frequency continuous wavelet transform techniques were used to obtain the cyclic and instantaneous characteristics of the vibration data acquired from two turbocharged four-cylinder, four-stroke diesel engine vehicles at idle under 12 different fuel conditions. Flekiewicz et al. [17] investigated the effects of gasoline and liquefied petroleum gas (LPG) on combustion pressure and vibration of a spark ignition engine. Increase of engine load and speed caused an increase in engine block vibration acceleration, for gasoline in the range of 22.1–100.5 m/s², and for LPG operation from 4.1 to 95.5 m/s². Keskin [18] investigated the effects of ethanol–gasoline–oil blends on spark ignition engine vibration characteristics, as well as, its noise emissions. The experimental results indicated that when fuel blends were used, vibration amplitudes and noise emission of the engine showed an increasing trend especially at 1500 and 2500 rpm. The author concluded that these results are probably due to oxygen content and higher latent heat of evaporation of ethanol, in which the increasing rate of pressure and peak pressure values in the cylinder rise during the combustion processes.

The aim of this paper is to study the effects of ethanol and methanol gasoline blends on an SI engine vibrational behaviour and identify the type of fuel used from its vibration signature. For this task, the acceleration time-series for a number of operating conditions and different fuel types are analysed and compared. A special algorithm is then constructed that carries out the fuel identification process.