This thesis presents a laser diagnostic investigation into the combustion, NOx formation and mixing processes occurring within the optically assessed combustion chamber of a methane-fired (10kW), fully premixed, self-aspirating, Helmholtz pulse combustor. The inlet geometry of the combustion chamber consisted of a step expansion and a bluff body obstacle formed by a stagnation plate. The focus of the investigation was the effects of the stream-wise position of the stagnation plate on the pulse combustion processes. A comprehensive parametric study of the performance of the pulse combustor is presented with stagnation plate position, air/fuel ratio and tailpipe length as the variables. The operating frequency and peak pressure amplitude trends were found to vary in accordance with the Rayleigh criterion. The operation of the combustor was more stable with the effective heat-release point preceding the resonant acoustic peak. Operation outside of this regime produce increased levels of CO. Time-resolved, laser-sheet flow visualisation images are presented of the flow structures within the combustion chamber. The inlet mixing - between the reactants and residual gases - was dominated by the formation of two counter-rotating toroidal vortices. In general, the inlet mixing was found to decrease as the stagnation plate was moved further into the combustion chamber. However, other mechanisms that tended to counter this trend were observed. Under certain conditions, significant flow reversals were imaged with gases penetrating the combustion chamber from the tailpipe. The combustion event was investigated using cycle-resolved chemiluminescence and laser induced fluorescence imaging of OH* radicals. Ignition of the fresh reactants by residual combustion/radical activity was found to occur along the interface between reactants and residual gases. The increase in reaction zone area generated by the action of the toroidal vortices provided the necessary mechanism for the rapid combustion of the reactants. The reduced mixing associated with moving the stagnation plate further into the combustion chamber produced a more compact combustion zone with less interaction between combusting reactants and cooler residual gases. This modification to the combustion zone was consistent with the measured trends of rising NOx tailpipe emissions and decreasing N02/NOx ratio. Under certain conditions, a reversal in the NOx and N02/NO, ratio trends was observed. This was explained by an augmentation of heat transfer rate out of the combustion chamber, characterised by increased flow reversal strength, which lead to cooler residual gases. Additional mechanisms, which modified the inlet mixing process, were also identified as contributing to the reversal of the NOx trends.

The presen present thesis spans a range of topics within thearea of turbulent flows, ranging from design of flow facilitiesto evaluation aluation of scaling laws and turbulence modelingdeling aspects through use of experimental data. A newwind-tunnel has been designed, constructed and evaluated at theDept. of Mechanics, KTH. Special attention was directed to thedesign of turning vanes that not only turn the flow but alsoallow for a large expansion without separation in the corners.The investigation of the flow quality confirmed that theconcept of expanding corners is feasible and may besuccessfully incorporated into low turbulence wind-tunnels. Theflow quality in the MTL wind-tunnel at the Dept. of Mechanics,KTH, was as also in investigated confirming that it still isvery good. The results are in general comparable to thosemeasured when the tunnel was as new, with the exception of thetemperature variation ariation that has decreased by a factorof 4 due to an improved cooling system. Experimental data from high Reynolds number zeropressure-gradient turbulent layers have been investigated.These studies have primarily focused on scaling laws withe.g.confirmation of an exponential velocity defect lawin a region, about half the size of the boundary layerthickness, located outside the logarithmic overlap region. Thestreamwise velocity probability density functions in theoverlap region was found to be self-similar when scaled withthe local rms value. Flow structures in the near-wall andbuffer regions were studied ande.g. the near-wall streak spacing was confirmed to beabout 100 viscous length units although the relative influenceof the near-wall streaks on the flow was as found to decreasewith increasing Reynolds number. The separated flow in an asymmetric plane diffuser wasdetermined using PIV and LDV. All three velocity componentswere measured in a plane along the centerline of the diffuser.Results for mean velocities, turbulence intensities andturbulence kinetic energy are presented, as well as forstreamlines and backflow coefficientcien describing theseparated region. Instantaneous velocity fields are alsopresented demonstrating the highly fluctuating flow. Resultsfor the above mentioned velocity quantities, together with theproduction of turbulence kinetic energy and the secondanisotropy inariant are also compared to data from simulationsbased on the k -wformulation with an EARSM model. The simulation datawere found to severely underestimate the size of the separationbubble. <b>Keywords:</b>Fluid mechanics, wind-tunnels, asymmetricdiffuser, turbulent boundary layer, flow structures, PDFs,modeling, symmetry methods.

<p>The presen present thesis spans a range of topics within thearea of turbulent flows, ranging from design of flow facilitiesto evaluation aluation of scaling laws and turbulence modelingdeling aspects through use of experimental data. A newwind-tunnel has been designed, constructed and evaluated at theDept. of Mechanics, KTH. Special attention was directed to thedesign of turning vanes that not only turn the flow but alsoallow for a large expansion without separation in the corners.The investigation of the flow quality confirmed that theconcept of expanding corners is feasible and may besuccessfully incorporated into low turbulence wind-tunnels. Theflow quality in the MTL wind-tunnel at the Dept. of Mechanics,KTH, was as also in investigated confirming that it still isvery good. The results are in general comparable to thosemeasured when the tunnel was as new, with the exception of thetemperature variation ariation that has decreased by a factorof 4 due to an improved cooling system.</p><p>Experimental data from high Reynolds number zeropressure-gradient turbulent layers have been investigated.These studies have primarily focused on scaling laws with<i>e.g.</i>confirmation of an exponential velocity defect lawin a region, about half the size of the boundary layerthickness, located outside the logarithmic overlap region. Thestreamwise velocity probability density functions in theoverlap region was found to be self-similar when scaled withthe local rms value. Flow structures in the near-wall andbuffer regions were studied and<i>e.g</i>. the near-wall streak spacing was confirmed to beabout 100 viscous length units although the relative influenceof the near-wall streaks on the flow was as found to decreasewith increasing Reynolds number.</p><p>The separated flow in an asymmetric plane diffuser wasdetermined using PIV and LDV. All three velocity componentswere measured in a plane along the centerline of the diffuser.Results for mean velocities, turbulence intensities andturbulence kinetic energy are presented, as well as forstreamlines and backflow coefficientcien describing theseparated region. Instantaneous velocity fields are alsopresented demonstrating the highly fluctuating flow. Resultsfor the above mentioned velocity quantities, together with theproduction of turbulence kinetic energy and the secondanisotropy inariant are also compared to data from simulationsbased on the k -<i>w</i>formulation with an EARSM model. The simulation datawere found to severely underestimate the size of the separationbubble.</p><p><b>Keywords:</b>Fluid mechanics, wind-tunnels, asymmetricdiffuser, turbulent boundary layer, flow structures, PDFs,modeling, symmetry methods.</p>

The aim of this thesis is to increase engine efficiency by studying the flow structures created in an engine cylinder during the intake phase and the effect of the subsequent compression. The invention of the combustion engine has enabled three centuries of economic growth fueled by energy stored as hydrocarbons. However, during the latter part of the twentieth century negative consequences on health and environment of the combustion engine were observed. In order to reduce emissions without increasing fuel consumption, improved knowledge of all physical processes occurring in the engine are necessary. The aim of this thesis is to increase the understanding of the flow prior to combustion, which can lead to reduced engine emissions and fuel consumption. Intake flow structures are studied using large eddy simulations and experiments on a steady swirl test rig. Flow acceleration was observed to reduce the swirl coefficient, and higher swirl coefficient was found during valve closing as compared to during valve opening. This implies that the rotation is stronger during the later part of the intake then what has been previously assumed. In addition, the computations show that the volume above the valves has a profound effect on the swirl created during the intake. To take this into account a novel way of calculating the swirl number was suggested. This approach gives a lower swirl number as compared to the commonly used Thien methodology. The effects of compression are studied using simulations of predefined flow structures undergoing compression. The peak turbulence levels were found to be increasing with tumble number and decreasing with swirl. It was noted that compression increased the turbulent fluctuations in the cylinder axis leading to anisotropic turbulence and that a small tilt angle was observed to have a significant effect on swirl homogeneity at top dead center.  In this thesis, a new methodology was proposed and validated for calculation of in-cylinder turbulence for a flat piston. The results of the thesis enhance the understanding of the dynamic effects encountered during intake as well recognizing that a small tumble component has a strong effect on the flow structures prior to combustion. These results can be used to improve the simplified computational methods used to optimize the engine. / <p>QC 20150420</p>

Flow mechanisms of compound meandering channels are recognised to be far more complicated than compound straight channels. The compound meandering channels are mainly characterised by the continuous variation of mean and turbulent flow parameters along a meander wavelength; the existence of horizontal shear layer at the bankfull level and the presence of strong helical secondary flow circulations in the streamwise direction. The secondary flow circulations are very important as they govern the advection of flow momentum, distort isovels, and influence bed shear stress, thus producing a complicated and fully three-dimensional turbulent flow structures. A great deal of experiments has been conducted in the past, which explains flow mechanisms, mixing patterns and the behaviour of secondary flow circulations. However, a complete understanding of secondary flow structures still remains far from conclusive mainly because the secondary flow structures are influenced by the host of geometrical and flow parameters, which are yet to be investigated in detail. The three-dimensional Reynolds-averaged Navier-Stokes and continuity equations were solved using a standard Computational Fluid Dynamics solver to predict mean velocity, secondary flow and turbulent kinetic energy. Five different flow cases of various model scales and relative depths were considered. Detailed analyses of the measured and predicted flow variables were carried out to understand mean flow mechanisms and turbulent secondary flow structures in compound meandering channels. The streamwise vorticity equation was used to quantify the complex and three-dimensional behaviour of secondary flow circulations in terms of their generation, development and decay along the half-meander wavelength. The turbulent kinetic energy equation was used to understand energy expense mechanisms of secondary flow circulations. The strengths of secondary flow circulations were calculated and compared for different flow cases considered. The main findings from this research are as follows. The shearing of the main channel flow as the floodplain flow plunges into and over the main channel influences the mean and turbulent flow structures particularly in the crossover region. The horizontal shear layer at the inner bankfull level generates secondary flow circulations. As the depth of flow increases, the point of generation of secondary flow circulations moves downstream. The secondary shear stress significantly contributes towards the generation of streamwise vorticity and the production of turbulent kinetic energy. The rate of turbulence kinetic energy production was found to be higher than the rate of its dissipation in the crossover region, which demonstrates that the turbulence extracts more energy from the mean flu\\' than what is actually dissipated. This also implies that, in the crossover region, the turbulence is always advected downstream by the mean and secondary flows, The strength of geometry induced secondary flow circulation increases with the increase in the relative depth.

Surface mean pressures, oil flow visualization, and 3-velocity-component laser-Doppler velocimeter measurements are presented for a turbulent boundary layer of momentum thickness Reynolds number, 7300 and thickness delta over two circular based axisymmetric bumps of height H = delta and 2delta and one rectangular based symmetric bump of H = 2delta. LDV data were obtained at one plane x/H ¥ 3.26 for each case. Complex vortical separations occur on the leeside and merge into large stream-wise mean vortices downstream for the 2 axisymmetric cases. The near-wall flow (y+ < 90) is dominated by the wall. For the axisymmetric cases, the vortices in the outer region produce large turbulence levels near the centerline and appear to have low frequency motions that contribute to turbulent diffusion. For the case with a narrower span-wise shape, there are sharper separation lines and lower turbulence intensities in the vortical downstream flow. Fine-spatial-resolution LDV measurements were also obtained on half of the leeside of an axisymmetric bump (H/delta = 2) in a turbulent boundary layer. Three-dimensional (3-D) separations occur on the leeside with one saddle separation on the centerline that is connected by a separation line to one focus separation on each side of the centerline. Downstream of the saddle point the mean backflow converges to the focal separation points in a thin region confined within about 0.15delta from the local bump surface. The mean backflow zone is supplied by the intermittent large eddies as well as by the near surface flow from the side of the bump. The separated flow has a higher turbulent kinetic energy and shows bimodal histograms in local and U and W, which appear to be due to highly unsteady turbulent motions. By the mode-averaged analysis of bimodal histograms, highly unsteady flow structures are estimated and unsteady 3-D separations seem to be occurring over a wide region on the bump leeside. The process of these separations has very complex dynamics having a large intermittent attached and detached flow region which is varying in time. These bimodal features with highly correlated local u and w fluctuating motions are the major source of large Reynolds stresses local u2, w2 and -uw. Because of the variation of the mean flow angle in the separation zones, the turbulent flow from different directions is non-correlated, resulting in lower shearing stresses. Farther from the wall, large stream-wise vortices form from flow around the sides of the bump. / Ph. D.

The objective of this thesis is to investigate the flow conditions that lead to coarse grain entrainment at near incipient motion conditions. Herein, a new conceptual approach is proposed, which in addition to the magnitude of hydrodynamic force or flow power, takes into account the duration of the flow event. Two criteria for inception of grain entrainment, namely the critical impulse and critical energy concepts, are proposed and compared. These frameworks adopt a force or energy perspective, considering the momentum or energy transfer from each flow event to the particle respectively, to describe the phenomenon. A series of conducted mobile particle experiments, are analyzed to examine the validity of the proposed approaches. First a set of bench-top experiments incorporates an electromagnet which applies pulses of known magnitude and duration to a steel spherical particle in a controlled fashion, so as to identify the critical level for entrainment. The utility of the above criteria is also demonstrated for the case of entrainment by the action of turbulent flow, via analysis of a series of flume experiments, where both the history of hydrodynamic forces exerted on the particle as well as its response are recorded simultaneously. Statistical modeling of the distribution of impulses, as well as conditional excess impulses, is performed using distributions from Extreme Value Theory to effectively model the episodic nature of the occurrence of these events. For the examined uniform and low mobility flow conditions, a power law relationship is proposed for describing the magnitude and frequency of occurrence of the impulse events. The Weibull and exponential distributions provide a good fit for the time between particle entrainments. In addition to these statistical tools, a number of Adaptive Neuro-Fuzzy Inference Systems employing different input representations are used to learn the nonlinear dynamics of the system and perform statistical prediction. The performance of these models is assessed in terms of their broad validity, efficiency and forecast accuracy. Even though the impulse and energy criteria are deeply interrelated, the latter is shown to be advantageous with regard to its performance, applicability and extension ability. The effect of single or multiple highly energetic events carried by certain coherent flow structures (mainly strong sweep events) with regard to the particle response is also investigated. / Ph. D.

The thesis describes state of the art volumetric measurement techniques and applies a 3D measurement technique, 3D Scanning Particle Tracking Velocimetry, to the transitional backward facing step flow. The measurement technique allows the spatial and temporal analysis of coherent structures apparent at the backward facing step. The thesis focusses on the extraction and interaction of coherent flow structures like shear layers or vortical structures.

The complex multiphase flow physics of spray-swirl interaction in both reacting and non-reacting environment is of fundamental and applied significance for a wide variety of applications ranging from gas turbine combustors to pharmaceutical drug nebulizers. Understanding the intricate dynamics between this two phase flow field is pivotal for enhancing mixing characteristics, reducing pollutant emissions and increasing the combustion efficiency in next generation combustors. The present work experimentally investigates the near and far-field break-up, dispersion and coalescence characteristics of a hollow cone spray in an unconfined, co¬annular isothermal swirling air jet environment. The experiments were conducted using an axial-flow hollow cone spray nozzle having a 0.5 mm orifice. Nozzle injection pressure (PN = 1 bar) corresponding to a Reynolds number at nozzle exit ReN = 7900 used as the test setting. At this setting, the operating Reynolds number of the co-annular swirling air stream number (Res) was varied in four distinct steps, i.e. Res = 1600, 3200, 4800 and 5600. Swirl was imparted to the co¬axial flow using a guided vane swirler with blade angle of Ф=45° (corresponding geometric swirl number SG = 0.8). Two types of laser diagnostic techniques were utilized: Particle / Droplet imaging analysis (PDIA) and shadowgraph to study the underlying physical mechanisms involved in the primary breakup, dispersion and coalescence dynamics of the spray. Measurements were made in the spray in both axial and radial directions and they indicate that Sauter Mean Diameter (SMD) in radial direction is highly reliant on the intensity of swirl imparted to the spray. The spray is subdivided into two zones as function of swirl in axial and radial direction: (1) near field of the nozzle (ligament regime) where variation in SMD arises predominantly due to primary breakup of liquid films (2) far-field of the nozzle where dispersion and collision induced coalescence of droplets is dominant. Each regime has been analyzed meticulously, by computing probability of primary break-up, probability of coalescence and spatio-temporal distribution of droplets which gives probabilistic estimate of aforementioned governing phenomena. In addition to this, spray global length scale parameters such as spray cone angle, break-up length, wavelength of liquid film has been characterized by varying Res while maintaining constant ReN.

In the last decades, the focus of internal combustion engine development has moved towards more efficient and less pollutant engines. In a Diesel engine, approximately 30-40% of the energy provided by combustion is lost through the exhaust gases. The exhaust gases are hot and therefore rich of energy. Some of this energy can be recovered by recycling the exhaust gases into turbocharger. However, the energy losses in the exhaust port are highly undesired and the mechanisms driving the total pressure losses in the exhaust manifold not fully understood. Moreover, the efficiency of the turbine is highly dependent on the upstream flow conditions. Thus, a numerical study of the flow in the exhaust port geometry of a Scania heavy-duty Diesel engine is carried out mainly by using the Large Eddy Simulation (LES) approach. The purpose is to characterize the flow in the exhaust port, analyze and identify the sources of the total pressure losses. Unsteady Reynolds Averaged Navier-Stokes (URANS) simulation results are included for comparison purposes. The calculations are performed with fixed valve and stationary boundary conditions for which experimental data are available. The simulations include a verification study of the solver using different grid resolutions and different valve lift states. The calculated numerical data are compared to existent measured pressure loss data. The results show that even global parameters like total pressure losses are predicted better by LES than by URANS. The complex three-dimensional flow structures generated in the flow field are qualitatively assessed through visualization and analyzed by statistical means. The near valve region is a major source of losses. Due to the presence of the valve, an annular, jet-like flow structure is formed where the high-velocity flow follows the valve stem into the port. Flow separation occurs immediately downstream of the valve seat on the walls of the port and also on the surface of the valve body. Strong longitudinal, non-stationary secondary flow structures (i.e. in the plane normal to the main flow direction) are observed in the exhaust manifold. Such structures can degrade the efficiency of a possible turbine of a turbocharger located downstream on the exhaust manifold. The effect of the valve and piston motion has also been studied by the Large Eddy Simulation (LES) approach. Within the exhaust process, the valves open while the piston continues moving in the combustion chamber. This process is often analyzed modeling the piston and valves at fixed locations, but conserving the total mass flow. Using advanced methods, this process can be simulated numerically in a more accurate manner. Based on LES data, the discharge coefficients are calculated following the strict definition. The results show that the discharge coefficient can be overestimated (about 20 %) when using simplified experiments, e. g. flow bench. Simple cases using fixed positions for valve and piston are contrasted with cases which consider the motion of piston and/or valves. The overall flow characteristics are compared within the cases. The comparison shows it is impossible to rebuild the dynamic flow field with the simplification with fixed valves. It is better to employ LES to simulate the dynamic flow and associated losses with valve and piston motion. / <p>QC 20131204</p>