Combustion noise in the internal combustion engines

Competition in todays automative industry forces companies to satisfy customer expectations with day by day improvements. One of the most important topic to need improvement is powertrain systems. Performance, fuel economy and NVH are the basic characteristics to improve for both legal obligations and customer expectations. Noise and vibrations caused by internal combustion engine working effect vehicle comfort parameters negatively. Noise caused by internal combustion engine working can be seperated three sections. These are mechanical noise, accessory noise and combustion noise. The most dominant one of these is the combustion noise. Therefore improvement on engine noise should be started from combustion noise. Improvements on combustion noise provide efficient engine performance and comfort with reduced noise. This study aims to investigate the combustion noise correlation between experimental measurments and simulation analysis results. To satisfy this, structural and acoustical CAE models should be built. Correlation should be investigated on 500- 5000 Hz combustion noise dominant region. According to literature studies, combustion noise improvement studies consist two parts. First part is experimental development work and the second part is the FEM development work. Regression analysis, coherence analysis and banger rig test are the experimental development studies. Regression and coherence analysis measurements are applied on the anechoic chamber dynamometers. Data is collected on each engine speed and engine load condition of engine map. At the banger rig test basic engine components are used to calculate combustion noise transfer function. Thanks to controlled combustion on combustion chamber and collecting noise data via microphones aroun the engine, combustion transfer function of engine is measured with banger rig test. At the FEM development part of literature studies, noise calculations are done on only component level like air intake or exhaust system components. However this thesis contains FEM method applied on all engine components. Due to the differences between diesel and gasoline engine combustion occurence, noise caused by combustion is different on diesel and gasoline engines. At the gasoline engines, fuel and air mixture ignites with spark and controlled combustion occurs. At the diesel engines, fuel sprays into compressed air and occurs uncontrolled combustion. Due to uncontrolled combustion, higher firing pressure gradient is occured. Higher firing pressure gradient causes higher combustion noise. Combustion noise is composed of four parts. These are direct combustion noise, indirect combustion noise, mechanical noise and flow noise. Main parameter of direct combustion noise is firing pressure in combustion chamber. Vertical forces on piston, conrod and crankshaft are the main contributor of direct combustion noise. Indirect combustion noise occurs due to piston secondary motion. This noise is dominat on higher engine load regions. Main parameters of indirect combustion noise are piston-crankshaft axial offset and clearence between the piston and cylinder wall. Mechanical noise is caused by pumps or gears excited by crankshaft. Main parameters on mechanical noise are gear material and backlash of gears. Flow noise is caused by air intake and exhaust systems. Intake resonators and exhaust mufflers can be applied to reduce this noise. At the thesis study; 2.0 liter, 4 cylinders, 125 hp light commercial vehicle's engine is tested. Acceleration, cylinder pressure in combustion chamber and noise data collected. Accelerometers were located on cylinder top deck centers and engine mounts. Pressure sensors were located on each cylinders. Microphones were located on each side of engine. Microphones were located around 1 meter away the engine system. Engine system considered as rectengular parallellepiped shape. Top side microphone result is lower than front, right and left side microphone results. Because foam top cover located on the top of engine and absorbs the noise. Front side microphone has fead system and its belt noise. Right side microphone has turbocharger and exhaust sytem noise dominantly. Left side microphone has alternator, starter, steering pump and air intake system noise dominantly. Firing pressure in combustion chamber is the main contributor of engine noise thats why it should be investigated both time and frequency domain. There are three parameters which affect combustion noise. First one is peak pressure, second one is first gradient and the last one is the second gradient of firing pressure. Peak firing pressure in combustion chamber has contribution on the lower frequencies. First derivative of firing pressure has contribution on middle frequencies. Second derivative of firing pressure has contribution on higher frequencies. Engine sound power is calculated from noise data from each side of engine. According to survey method, A weighted sound power level is calculated for engine speed. After completed experimental measurement, FEM model was built. All 3D components are modelled as solid and assigned young's modulus, poison ratio and density. Bolts and welds are modeled as 1D and assigned geometric and material info. Sheet metals are modeled as 2D and assigned thickness and material info. Cylinder pressure input was applied on FEM model and acceleration results were taken from engine mounts and cylinder top dead center locations. Accelerometer comprasion between experimental measurement and FEM analysis has correlation on second engine order results. After structural CAE model correlation, acoustical CAE model preparation was started. BEM model was built according to appropriate eleman length. Eleman length selection is important for maximum reasonable frequency of CAE results. After BEM model built, acoustic transfer vector analysis is applied on model. On this stage microphone location on experimental measurment was defined. ATV analysis result is matched with surface velocity output from structural CAE model. This matching process provides noise data on specified microphone locations. Acoustical CAE method provides 3rd octave noise data for each engine speed and each mic loacation. These results can be compared with experimental microphone measurements. CAE overall noise result was compared with experimental noise measurement. On the experimental measurement, flow and mechanical noise contribution were seen slightly. With this assumption, acoustical CAE results were shown as expected. As a conclusion, all of these studies were showed that; • Firing pressure in combustion chamber is the main contributor of combustion noise. • Experimental studies on combustion noise are both costly and needs longer time. • Finite element methods are best solution to estimate and optimise combustion noise.