Sonia Woudberg

The performance of drag models on flow behaviour in the CFD simulation of a fluidized bed

The aim of this study is to verify the use of a newly developed drag model in the simulation of fluidized beds. The drag model is based on a geometric description of the geometry found in a fluidised bed, treating it as a spatially and temporally variable inhomogeneous, locally isotropic, porous medium. Account is taken of the fact that flow conditions in low porosity parts of a bed can be viewed as flow between particles. At high porosities the bed resembles flow past the particles of a dilute assemblage and for that the current model is complemented with results from other models. The new drag model, as well as other models found in literature, was tested in the numerical simulations. Computational results are compared mutually, as well as to experimental data, and the differences and discrepancies discussed.

Characterization of microstructure and fatigue life of CP titanium grade 2 specimens subject to various bending processes.

This study focussed on an investigation into the fatigue behaviour of commercially pure (CP) titanium (Ti) grade 2 specimens produced by various bending (deformation) processes, i.e. mechanical bending, laser bending and laser-mechanical bending. Although published literature on the fatigue behaviour of components produced by laser forming is relatively limited, it was found that laser bending could improve the mechanical behaviour of such components. At high fatigue load settings the microstructure was the dominant contributor to crack initiation and crack growth. After the specimens were bent, residual stress analysis as well as a characterization of the microstructure was performed on the three bent samples as well as the parent plate. Mathematical equations were formulated and presented for the laser forming process for predicting the fatigue life of CP Ti grade 2. This study revealed that the laser forming process can be successfully used as a manufacturing method in the forming of CP Ti grade 2.

A geometric pore-scale model for predicting the permeability of 3D flow through fibrous porous media

A geometric pore-scale model, based on rectangular geometry, is used to quantify the fluid-solid interaction in fibrous porous media in order to predict the permeability. The analytical modeling procedure is based on sound physical principles. Permeability predictions are presented for flow parallel and perpendicular to the axes of unidirectional fibres. In the latter case maximum possible staggering is introduced. A weighted average is performed to obtain the permeability prediction for 3D flow through fibrous porous media. Effects such as pore blockage at very low porosities and developing flow are incorporated into the predictive equations for the permeability to provide a model that is applicable over the entire porosity range. The resulting 3D model leads to satisfactory agreement with other three-dimensional models and data from the literature.

Modelling And Analysis Of Permeability Of Anisotropic Compressed Non‐Woven Filters

An existing geometrical pore‐scale model for flow through isotropic spongelike media is adapted to predict flow through anisotropic non‐woven glass fibre filters. Model predictions are compared to experimental results for the permeability obtained for a filter under different stages of compression to demonstrate the capability of the model to adjust to changes in porosity. The experimental data used are for a glass fibre paper with a uniform fibre diameter. The input parameters of the pore‐scale model are the porosity, fibre diameter and some measure of the anisotropy between the in‐plane and normal directions to the paper. Correlation between the predictions and the experimental results is satisfactory and provides confidence in the modelling procedure. It is shown that the permeability is very sensitive to changes in the level of anisotropy, i.e. the level of compression of the nonwoven material.

Modelling of Diffusion in Porous Structures.

An existing pore-scale model is used to predict the effective diffusivity of staggered two-dimensional rectangular unconsolidated arrays through the use of a Representative Unit Cell concept. A tri-diagonal matrix algorithm is used to solve the diffusive flux field and to compute the effective diffusion coefficient for concentration gradients of staggered arrays. The numerical results and analytical model are compared critically with theoretical and numerical studies, as well as experimental data reported in literature. The good correlations obtained for the effective diffusivity provide confidence in both the computational and analytical work.

A Study Of Microstructural Fatigue And Residual Stress Evolution In Titanium Plates Deformed By Mechanical And Laser Bending

This manuscript details the investigation of residual stresses and microstructural changes in commercially pure titanium plates deformed by three bending methods, i.e. (i) mechanically, (ii) using a laser beam and (iii) by a combination of laser and mechanical bending to a final radius of curvature of 120 mm. The residual strains were measured using the hole-drilling method and the analyses indicate that higher tensile residual stresses reside in the laser bent plate samples compared to that assessed from the other two bending methods. However, an important finding was that the residual stress state of laser bent samples was significantly reduced after application of mechanical bending for samples subjected to the combined bending application. This aspect coupled with the increased hardness due to microstructural changes makes the laser/mechanical bending process more favourable when designing against fatigue failure.

Investigation of the effect of compression on a soft fibrous porous medium.

An adaptable geometric pore-scale model is proposed to determine the effect of compressionof a fibrousporousmedium on the permeability.The model is applied to a soft compressible polyester material typically found in pillows. Seven stages of compression of the porous medium are considered. The effect of compression on the micro-structural parameters is determined, because it directly affects the permeability and pressure drop predictions. The model is also adapted to account for the combined effects of compression and developing flow on permeability. A sensitivity analysis is performed to determine the effect of experimental errors on

An Adaptable Analytical Ergun‐Type Equation For High Porosity Spongelike Porous Media

An analytical Ergun‐type equation for spongelike media is introduced in which developing flow in the short ducts of high porosity metallic foams are accounted for. Instead of the customary procedure of adjusting the empirical coefficients of the Ergun equation to apply to consolidated spongelike media, a pore scale model is introduced and the physical flow conditions remodelled. The pore‐scale linear dimensions are expressed as a function of porosity and the dependence of the form drag coefficient on porosity is incorporated into the model which leads to satisfactory predictions for the inertial coefficient. The model predictions are compared to experimental data from the literature and the satisfactory correspondence provides confidence in the physical adaptability of the model.

Measurement and Prediction for Air Flow Drag in Different Packing Materials.

Packed bed reactors are widely used in industry to improve the total contact area between two substances in a multiphase process. In some cases, like for the packing elements of some CO2 absorption towers, the packing material can be of such geometric nature that during discharge through them different flow conditions can be present in different parts of the packing. This renders prediction of pressure drops quite difficult. This paper concerns experimental and modelling activities to improve predictive equations for pressure drops over a packed bed of Raschig rings. It is shown that the application of some corrective measures can dramatically improve the correlation between theory and experiment, but that more research is needed in this field, regarding both carefully controlled experiments and mathematical modelling.