Particulate emissions from internal combustion engines is a well-known issue with direct implications on air quality and human health. Recently there is an increased concern about the high number of ultrafine particles emitted from modern engines. Here we explore a concept for grouping these particles, reducing their total number and shifting the relative size distribution towards fewer larger particles. Particles having a non-zero relaxation time may be manipulated to yield regions of high particle concentration, accommodating agglomeration, when introduced into an oscillating flow field. The oscillating flow field is given by simple periodic geometrical changes of the exhaust pipe itself. It is discussed how the shape of these geometrical changes and also the engine pulses effect the grouping behavior for different size particles, including when Brownian motion becomes relevant. Simulations are performed using a bespoke 1D-model sufficient for the basic parameter studies of the concept given here. 

In this study the impact of geometrical expansion on the intrinsic flow structures of corrugated-pipe like geometries is numerically investigated. Two axisymmetric geometrieswith wavy-walls are taken under consideration. The geometries have the same geometric wavelength but differ in their respective corrugation amplitude and mean hydraulic diameter. A Large Eddy Simulation (LES) approach is used to numerically simulate a 3D,incompressible, non pulsatile and turbulent gas flow. Firstly, a solver validation study is performed using experimental data for the periodic hill benchmark case (data set available at http://qnet-ercoftac.cfms.org.uk/w/index.php/Abstr:2D\_Periodic_Hill_Flow). This is followed by a grid resolution study in order to determine an appropriate mesh size. A comparison between the flow profiles and coherent structures as calculated for the two geometries is made. The coherent flow structures are exposedusing a Proper Orthogonal Decomposition (POD) approach and spectral analysis. The POD analysis revealed an asymmetric oscillatory mode for the geometry with the larger mean hydraulic diameter. A similar analysis highlighted axial modes for the geometry withthe smaller mean hydraulic diameter. Subsequent spectral analysis revealed that up to a certain frequency the strongest spectral content appeared in the expansion regions, en-compassing the shear-layer/recirculation bubble, of the geometry. Beyond that frequency the strongest spectral content appeared in a confined region within the shear-layer. For the geometry with larger mean hydraulic diameter the distinguishing frequency was determined to be 500Hz. In the case of the geometry with the smaller mean hydraulic diameter this frequency increased to 1500Hz. Frequency content lower than 100Hz is governed by the dynamics of recirculation bubble whereas higher frequency content is governed by the dynamics of the induced shear-layer.

In this study the influence of flow pulsations on the intrinsic flow structures found within corrugated-pipe like geometries is explored numerically. An axisymmetric geometry with wavy-walls is taken under consideration. Two different pulsatile inlet conditions are considered which consist of sinusoidal pulse form and a distinct pulsation frequency: 80Hzand 160Hz. The 3D, incompressible and turbulent gas flow is simulated using a Large Eddy Simulation (LES) methodology with a wall adapting local eddy viscosity (WALE) subgrid scale (SGS) model. A solver validation study is firstly presented based on the Direct Numerical Simulation (DNS) data of the wavy-wall benchmark case (data set available at http://cfd.mace.manchester.ac.uk/ercoftac/database/cases/case77/Case_data/). A subsequent grid resolution study is completed in in order to determinean appropriate mesh size. This is followed by a comparison between the flow profiles and coherent structures for the pulsatile flow under consideration. The coherent flow structures are exposed using a Proper Orthogonal Decomposition (POD) approach and spectral analysis. The appearance of an axial mode is highlighted. Spectral analysis indicates that this particular flow mode is driven by the flow pulsations. Furthermore, an asymmetric oscillatory mode is highlighted using the POD approach. Spectral analysis revealed that up to 500Hz the most dominant spectral content appeared in the expansion regions, enveloping the shear-layer/recirculation bubble, of the geometry. Increasing beyond 500Hz the strongest spectral content appeared inside a narrow, confined region within the shear-layer.

In this numerical study particle behavior inside a sinusoidal pipe geometry is analyzed. The 3D geometry consistsof three identical modules, with a periodic boundary condition applied to the flow in the stream wise direction. The incompressible, turbulent gas flow is modeled using a Large Eddy Simulation (LES) approach. Furthermore, the particle dynamics are simulated using a Lagrangian point force approach incorporating the Stokes drag and slip correction factor. Four different  size of particles are considered along with two different inflow conditions: continuous and pulsatile.The pulsatile inflow has an associated flow frequency of 80Hz . The solver has been validated using the backward facing step benchmark case and experimental data. The sensitivity of the predicted fluid flow solution by the CFD to the chosen computational grid resolution has been assessed through a grid resolution study for the sinusoidal pipe geometry. The first and second order streamwise velocity moments of the continuous inflow scenario are firstly highlighted. This is followed by the calculated phase locked first and second order streamwise velocity moments for the pulsatile inflow scenario. The fluid flow through the sinusoidal pipe is characterized by weak flow separation in the expansion zones of the sinusoidal pipe geometry, where induced shear layers and weak recirculation zones are identified. Particle behavior under the two inflow conditions is quantified using particle dispersion, particle residence time and average radial position of the particle. No discernible difference in the particle behavior is observed between the two inflow conditions. As the size of the particles considered is small, the particles follow the flow streamlines.The recirculation zones within the cavities of the expansion section of the geometry are weak. Thus, the particles do not reside in the computational domain for a long duration of time. This in turn reduces the likelihood of particles to agglomerate within the recirculation flow regions.

The hazardous impact of particulate emissions from internal combustion engines on human health and the environment is well known. One possible approach to curtail these emissions is to shift the particle number distribution by agglomeration. The agglomeration process will result in larger sized particles, in turn easier to filter out.The present study considers flow manipulation as a possible mean to achieve particle agglomeration. A corrugated pipe consisting of successive expansion and contraction regions based on a previously proposed 1-D model for periodic grouping is utilized under continuous and pulsatile inflow conditions. An experimental study on this pipe has already been reported, forming the base for and providing boundary conditions to the numerical simulations using the Large Eddy Simulations approach. The particle dynamics are modeled via Lagrangian particle tracking considering mono-dispersed case setups of three different particles of sizes: 1μm, 800nm and 400 nm. The residence time and radial distribution of theparticles under different inflow conditions are reported. Generally, the particles tend to reside within the computational domain for a short amount of time and thus less likely to interact with other particles. Furthermore, it is observed that particles are less prone to reach the recirculation zones formed in the cavities of the expansion region. Particle behavior under a pulsatile flow is slightly more favorable than a continuous flow scenario, a result due to the flow pulsations leading to formation of a back flow forcing the particles back into the observed recirculation zones.