Anthony J. H. Goddard

AEROSOL DEPOSITION IN TURBULENT CHANNEL FLOW ON A REGULAR ARRAY OF THREE-DIMENSIONAL ROUGHNESS ELEMENTS

Abstract Understanding particle deposition onto rough surfaces is important for many engineering and environmental applications. An experimental system was designed for the study of aerosol deposition on regular arrays of uniform elements (in the form of discrete protrusions) in a turbulent ventilation duct flow using monodisperse tracer small particles, in the range 0.7– 7.1 μ m . The Reynolds number for the test conditions was 44,000 in the 150 mm square duct. The roughness elements were arranged at two different orientations with respect to the airflow direction and the aerosol deposition velocity and pressure drop were measured for both orientations. Compared to earlier measurements in the same duct system involving smooth or ribbed surfaces, a significant increase in deposition velocity onto the regular roughness elements is observed. In addition, the associated pressure loss penalty is lower than in the presence of the roughness elements than in the presence of the ribbed surfaces. This may be attributable to the small dimensionless roughness height of the elements, which results only in a moderate distortion of the flow structure near the surfaces.

The Inner-Element Subgrid Scale Finite Element Method for the Boltzmann Transport Equation

Abstract This paper presents a new multiscale radiation transport method based on a Galerkin finite element spatial discretization of the Boltzmann transport equation. The approach incorporates a discontinuous subgrid scale (SGS) solution within the continuous finite element representation of the spatial variables. While the conventional discontinuous Galerkin (DG) method provides accurate and numerically stable solutions that suppress unphysical oscillations, the number of unknowns is relatively high. The key advantage of the proposed SGS approach is that the solutions are represented within the continuous finite element space, and therefore, the number of unknowns compared with DG is relatively low. The applications of this method are explored using linear finite elements, and some of the advantages of this new discretization over standard Petrov-Galerkin methods are demonstrated. The numerical examples are chosen to be demanding steady-state mono-energetic radiation transport problems that are likely to form unphysical oscillations within numerical scalar flux solutions. The numerical examples also provide evidence that the SGS method has a thick diffusion limit. This method is designed to work under arbitrary angular discretizations, so solutions using both spherical harmonics and discrete ordinates are presented.

Adjoint A Posteriori Error Measures for Anisotropic Mesh Optimisation

In this paper an adjoint- (or sensitivity-) based error measure is formulated which measures the error contribution of each solution variable to an overall goal The goal is typically embodied in an integral functional, e.g., the solution in a small region of the domain of interest. The resulting a posteriori error measures involve the solution of both primal and adjoint problems. A comparison of a number of important a posteriori error measures is made in this work. There is a focus on developing relatively simple methods that refer to information from the discretised equation sets (often readily accessible in simulation codes) and do not explicitly use equation residuals. This method is subsequently used to guide anisotropic mesh adaptivity of tetrahedral finite elements. Mesh adaptivity is achieved here with a series of optimisation heuristics of the landscape defined by mesh quality. Mesh quality is gauged with respect to a Riemann metric tensor embodying an a posteriori error measure, such that an ideal element has sides of unit length when measured with respect to this metric tensor. This results in meshes in which each finite-element node has approximately equal (subject to certain boundary-conforming constraints and the performance of the mesh optimisation heuristics) error contribution to the functional (goal).

An automatic adaptive meshing technique for Delaunay triangulations

Abstract A new algorithm is described for adaptivity of Delaunay triangulations. It distinguishes between regions of mesh according to whether they are to be refined, unaltered or coarsened. The method automatically identifies nodes of the mesh which are candidates for deletion. For a region to be coarsened, a node deletion process is proposed and it is proved that the resulting triangulation is Delaunay. After coarsening, the mesh retains a high quality for two reasons: the first is because a set of nodes have been used to obtain the coarse mesh, which are equi-distant apart, in some sense defined by the original mesh; the second is because the triangulation is Delaunay. The mesh refinement procedure sub-divides elements into a number of similar elements in the regions to be refined. Two-dimensional examples demonstrate the quality of the mesh after coarsening/refinement.

Particle deposition in ventilation duct onto three-dimensional roughness elements

Gaining insights on particle deposition onto ventilation duct has many important applications. One key pathway by which outdoor polluted air enters the indoor environment is through mechanical ventilation ducts. An experimental system was designed for the study of particle deposition on regular arrays of uniform elements (in the form of discrete protrusions) in a turbulent ventilation duct 6ow using monodisperse tracer small particles, in the range 0.7–7:1 � m. The Reynolds number for the test conditions was 44,000 in the 150 mm square duct. Four di:erent types of uniform roughness elements were tested. Compared to earlier measurements in the same duct system involving smooth or ribbed surfaces, a signi;cant increase (up to 74 times) in deposition velocity onto the regular roughness elements is observed. ? 2002 Elsevier Science Ltd. All rights reserved.

On an improved design of a fluidized bed nuclear reactor-I: Design modifications and steady-state features

Abstract This paper describes several modifications to the design of a fluidized bed nuclear reactor in order to improve its performance. The goal of these modifications is to achieve a higher power output, requiring an excess reactivity of 4% at maximum expansion of the bed. The modifications are also intended to obtain a larger safety margin when the reactor does not operate; a shutdown margin of 4% is required when the bed is in a packed state. The modifications include installing an embedded side absorber, changing the reactor cross-section area, and modifying the moderator-to-fuel ratio. The new design based on the modifications related to the aforementioned parameters achieves the desired shutdown margin and the excess reactivity. A model describing the coupling of neutronics and thermal/fluid dynamics is developed, and it is used to study the behavior of the reactor at steady conditions. The results show that the reactor can achieve a high output temperature of 1163 K and produce a thermal power of ~120 MW. Further, the results indicate that the power level of the reactor can be controlled easily by adjusting the flow of helium into the core without any further use of control rods or other active control mechanisms.

An automatic mesh coarsening technique for Delaunay triangulations

A new algorithm is described for automatic coarsening of Delaunay triangulations. The method automatically identifies the nodes of the mesh which are candidates for deletion in each region to be coarsened. The mesh coarsening procedure is to regenerate a Delaunay triangulation for all the remaining nodes in the solution domain. The adapted mesh retains a high quality for two reasons: (i) because a set of nodes have been used to obtain the coarse mesh, which are equidistant apart in some sense defined by the original mesh; (ii) because the triangulation is Delaunay. A two-dimensional example demonstrates the quality of the mesh after coarsening.

Nuclear Reactor Reactivity Prediction Using Feed Forward Artificial Neural Networks

In this paper, a feed forward artificial neural network (ANN) is used to predict the effective multiplication factor (k eff ), an indication of the reactivity of a nuclear reactor, given a fuel Loading Pattern (LP). In nuclear engineering, the k eff is normally calculated by running computer models, e.g. Monte Carlo model and finite element model, which can be very computationally expensive. In case that a large number of reactor simulations is required, e.g. searching for the optimal LP that maximizes the k eff in a solution space of 1010to 10100, the computational time may not be practical. A feed forward ANN is then trained to perform fast and accurate k eff prediction, by using the known LPs and corresponding k eff s. The experiments results show that the proposed ANN provides accurate, fast and robust k eff predictions.