<p>Simulating fluid behavior has proven to be a demanding challenge which requires complex computational models and highly efficient data structures. Smoothed Particle Hydrodynamics (SPH) is a particle based computational model used to simulate fluid behavior that has been found capable of producing convincing results. However, the SPH algorithm is computational heavy which makes it cumbersome to work with.</p><p>This master thesis describes how the SPH algorithm can be accelerated by utilizing the GPU’s computational resources. It describes a model for how to distribute the work load on the GPU and presents a suitable data structure. In addition, it proposes a method to represent and handle moving objects in the fluids surroundings. Finally, the performance gain due to the GPU is evaluated by comparing processing times with an identical implementation running solely on the CPU.</p>

<p>The Particle Level set method is a successful extension to Level set methods to improve thevolume preservation in fluid simulations. This thesis will analyze how sparse volume data structures can be used to store both the signed distance function and the particles in order to improve access speed and memory efficiency. This Particle Level set implementation will be evaluated against Digital Domains current Particle Level set implementation. Different degrees of quantization will be used to implement particle representations with varying accuracy. These particles will be tested and both visual results and error measurments will be presented. The sparse volume data structures DB-Grid and Field3D will be evaluated in terms of speed and memory efficiency.</p>

<p>The topic of this thesis concerns the LHC, the next large particle accelerator at CERN which will start operating in 2007. Being based on superconductivity, the LHC needs to operate at very low temperatures, which makes great demands on the cryogenic system of the accelerator. To cope with the heat loads induced by the particle beam, a beam screen cooled with forced flow of supercritical helium is used.</p><p>There is an interest in upgrading the energy and luminosity of the LHC in the future and this would require a higher heat load to be extracted by the beam screen cooling system. The objective of this thesis is to quantify different ways to upgrade this system by mainly studying the effects of different pressure and temperatures levels as well as a different cooling medium, neon.</p><p>For this a numerical program which simulates one-dimensional pipe flow was constructed. The frictional forces were accounted for by the empirical concept of friction factor. For the fluid properties, software using empirically made correlations was used. To validate the numerical program, a comparison with previous experimental work was done. The agreement with experimental data was good for certain flow configurations, worse for others. From this it was concluded that further comparisons with experimental data must be made in order to tell the accuracy of the mathematical model and the correlations for fluid properties used.</p><p>When using supercritical helium, thermo-hydraulic instabilities may arise in the cooling loop. It was of special interest to see how well a numerical program could simulate and predict this phenomenon. It was found that the numerical program did not function for such unstable conditions; in fact it was much more sensitive than what reality is.</p><p>For the beam screen cooling system we conclude that to cope with the increased heat loads of future upgrades, an increase in pressure level is needed regardless if the coolant remains helium, or is changed to neon. Increasing the pressure level also makes that the problems with thermo-hydraulic instabilities can be avoided. Of the two coolants, helium gave the best heat extraction capacity. Unlike neon, it is also possible to keep the present temperature level when using helium.</p>

<p>In this thesis a set of boundary conditions for multiphase flow is suggested.</p><p>Characteristic-based boundary conditions are reviewed for single-phase flow. The problem of open-boundary conditions is investigated, and to avoid drifting values, the use of control functions is proposed.</p><p>The use of control functions is also verified with a new test which assesses the quality of the boundary conditions. Particularly, P- and PI-control functions are examined. PI-controllers have the ability to specify a given variable exactly at the outlet as well as at the inlet, without causing spurious reflections which are amplified.</p><p>Averaged multiphase flow equations are reviewed, and a simplified model is established. This model is used for the boundary analysis and the computations. Due to the averaging procedure, signal speeds are reduced to the order of the flow speed. This leads to numerical challenges. For a horizontal channel flow, a splitting of the interface pressure model is suggested. This bypasses the numerical problems associated with separation by gravity, and a physical realistic model is used. In this case, the inviscid model is shown to possess complex eigenvalues, and still the characteristic boundary conditions give reasonable results.</p><p>The governing equations are solved with a Runge-Kutta scheme for the time integration. For the spatial discretisation, a finite-volume and a finite-difference method are used. Both implementations give equivalent results. In single-phase flow, the results improve significantly when a numerical filter is applied. For two-dimensional two-phase flow, the computations are unstable without a numerical filter.</p>

A multiple-point Arnoldi method is derived for model reduction of computational fluid dynamic systems. By choosing the number of frequency interpolation points and the number of Arnoldi vectors at each frequency point, the user can select the accuracy and range of validity of the resulting reduced-order model while balancing computational expense. The multiple-point Arnoldi approach is combined with a singular value decomposition approach similar to that used in the proper orthogonal decomposition method. This additional processing of the basis allows a further reduction in the number of states to be obtained, while retaining a significant computational cost advantage over the proper orthogonal decomposition. Results are presented for a supersonic diffuser subject to mass flow bleed at the wall and perturbations in the incoming flow. The resulting reduced-order models capture the required dynamics accurately while providing a significant reduction in the number of states. The reduced-order models are used to generate transfer function data, which are then used to design a simple feedforward controller. The controller is shown to work effectively at maintaining the average diffuser throat Mach number. / Singapore-MIT Alliance (SMA)

Fully morphing wings allow the active change of the wing surface contours/wing configuration in flight enabling the optimum wing design for various flight regimes. These wing shape deformations are obtained by using smart actuators, which requires that the wing structure be flexible enough to morph under applied actuator loads and at the same time be fully capable of holding the aerodynamic loads. The study of such wing surface deformation requires an aeroelastic analysis since there is an active structural deformation under an applied aerodynamic field. Herein, a 2-D wing section, that is, an airfoil is considered. Modeling a variable geometry airfoil is performed using B-spline expansions. B-spline representation is also favorable towards optimization and provides a methodology to design curves based on discrete polygon points. The energy required for deforming the airfoil contour needs to be minimized. One of the methodologies adopted to minimize this actuation energy is to use the aerodynamic load itself for wing deformation. Another approach is to treat the airfoil deformation as a Multi Disciplinary Optimization (MDO) problem wherein the actuation energy needs to be minimized subject to certain constraints. The structural analysis is performed using commercial finite element software. The aerodynamic model is initiated from viscous-inviscid interaction codes and later developed from commercial Computational Fluid Dynamics (CFD) codes. Various modeling levels are investigated to determine the design requirements on morphing airfoils for enhanced aircraft maneuverability. / Singapore-MIT Alliance (SMA)

A trajectory piecewise-linear (TPWL) approach is developed for a computational fluid dynamics (CFD) model of the two-dimensional Euler equations. The approach uses a weighted combination of linearized models to represent the nonlinear CFD system. The proper orthogonal decomposition (POD) is then used to create a reduced-space basis, onto which the TPWL model is projected. This projection yields an efficient reduced-order model of the nonlinear system, which does not require the evaluation of any full-order system residuals. The method is applied to the case of flow through an actively controlled supersonic diffuser. With an appropriate choice of linearization points and POD basis vectors, the method is found to yield accurate results, including cases with significant shock motion. / Singapore-MIT Alliance (SMA)

The airflow field and particle trajectories inside hard disk drive (HDD) are investigated in this study using commercial software Fluent and Gambit. Three-dimensional grids inside the HDD configuration are built using Gambit taking into account all the components and their geometric details. The airflow field inside HDD is simulated using three incompressible Navier-Stokes equations for various disk rotational speeds. The effects of using the various turbulence models inside the Fluent software such as the standard k - ε , RNG k - ε and Reynolds Stress Method on the computed airflow characteristics are also assessed. Steady flow fields and the effects of rotational speeds are assessed. Based on the computed steady airflow patterns, particle trajectories are computed using routines available in Fluent as well as special particle trajectory functions defined by the user via the user-defined functions. Particles of different sizes and materials are injected at various locations in the computed flow field and the corresponding particle trajectories are studied. Based on the investigation, the trajectory tends to be different according to sizes and materials. The present work forms a basis for further investigation of heat transfer processes inside the HDD to address thermal management issues and also the computation of unsteady flow fields in the HDD due to the movement of the actuator arm during data storage and retrieval / Singapore-MIT Alliance (SMA)

This work is focused on a state-of-the-art tubular micro-solid oxide fuel cell (TSOFC), ~3 millimeters in diameter and ~300 microns thick, with Ni/YSZ and LSM/YSZ composite electrodes and a YSZ electrolyte. A 2D axi-symmetric, multi-scale CFD model is developed which includes the fluid flow, mass transfer, and heat transfer within the gas channels and the porous electrodes. The electrochemical reactions are modeled within the volume of the electrodes, enabling the model to account for the extent of the reaction zone. Thermodynamic expressions are developed to estimate the single-electrode reversible heat generation and the single-electrode electromotive force of a non-isothermal electrochemical cell. The isothermal, non-isothermal, and transient models are each validated against the experimental results, and consistent with the physical reality of the TSOFC. A novel approach is used to estimate the kinetic parameters, enabling the simulations to be used as a diagnostic tool. The model is used to gain a thorough insight about the TSOFC. The cathode electrochemical activity and the anode support ohmic loss are identified as the two major performance bottlenecks for this cell. Including radiation is found to be essential for a physically meaningful heat transfer model. The thermoelectric effects on the cell overall electromotive force is found to be negligible. It is found that the anode reaction is always endothermic, while the cathode reaction is always exothermic, and that the temperature gradients across the cell layers are less than 0.05C The cell transient response is found to be fast, and dominated by the thermal transients. Several physical properties used in the model are measured experimentally, indicating that that the correlations used in the literature are not always suitable, especially when new fabrication techniques are used. The conductivity of the anode support was measured to be several orders of magnitude lower than expected and very sensitive to temperature, which explains the lower than expected and occasionally degrading cell performance. A hypothesis is proposed to explain this phenomenon based on the thermal expansion effects which result in the formation and disruption of particle to particle contacts within the composite electrode. / Chemical Engineering

We present methods for incorporating bubbles into a photorealistc fluid simulation. Previous methods of fluid simulation in computer graphics do not include bubbles. Our system automatically creates bubbles, which are simulated on top of the fluid simulation. These bubbles are approximated by spheres and are rendered with the fluid to appear as one continuous surface. This enhances the overall realism of the appearance of a splashing fluid for computer graphics. Our methods leverage the particle level set representation of the fluid surface. We create bubbles from escaped marker particles from the outside to the inside. These marker particles might represent air that has been trapped within the fluid surface. Further, we detect when air is trapped in the fluid and create bubbles within this space. This gives the impression that the air pocket has become bubbles and is an inexpensive way to simulate the air trapped in air pockets. The results of the simulation are rendered with a raytracer that includes caustics. This allows the creation of photorealistic images. These images support our position that the simple addition of bubbles included in a fluid simulation creates results that are much more true to life.