Michael Teubner

Heat exchange in an attic space

Abstract Thermal effects in building attics are extremely important in regions where temperature extremes occur regularly. In particular, in the northern regions of North America, cold winters can lead to a dangerous build-up of ice on roofs, as well as preventing heat being distributed efficiently through the building. This paper examines heat transfer within attics and looks at an inexpensive solution technique involving ceiling fans. The computer model developed accurately reproduces the results of previous investigations, presents solutions for realistic Grashof numbers, and demonstrates the effects of installing ceiling fans in the ceilings of top storey rooms in affected buildings.

Numerical Studies on the Approximation of Horizontal Diffusion using Sigma Coordinate System

In this paper, horizontal diffusion using sigma coordinates was examined numerically. To transform the horizontal diffusion equation, the chain rule was applied twice resulting in extra terms that are cumbersome to treat computationally. Two publisheddirect modifications to the transformed equation were chosen, and numerical solutions using these modified equations are compared to those using the fully transformed equation. Our numerical studies have shown that only the fully transformed equation produces realistic solutions.

A RANS-based Statistical Noise Model for Trailing Edge Noise

ow velocities. The method uses a Green’s function for a semi-innite half-plane to generate a far-eld acoustic autospectrum using a statistical model for the turbulence in the boundary layer in the vicinity of the trailing edge. This statistical model is an assumed turbulent velocity cross spectrum, and is dened in terms of the mean ow data provided by a RANS simulation. Reasonably good agreement is observed between acoustic predictions and experiments for the NACA 0012 cases. For the DU 96-180, the model over-predicts the experimental acoustic spectra.

Source Localisation in a Real Human Head

Summary:Neural activity within the human brain produces electrical potentials that are transmitted through the various tissues of the head to the scalp. A three-dimensional finite difference model has been applied to simulate this process and used as the basis for an inverse model, wherein known potentials on the scalp are used to locate sources of neural activity within the brain. The inverse model uses linear and nonlinear response functions, together with nonlinear regression to determine the source location. The model has been applied to three different simulations, and in each case was able to locate the source using a combination of linear and nonlinear response functions.

Minimization of trailing edge noise by parametric airfoil shape modications

reduce turbulent boundary layer trailing-edge (TBL-TE) noise compared with a reference NACA-0012 airfoil. The semi-empirical model used in the investigation is a combination of XFOIL aerodynamic prediction and the BPM TBL-TE noise model. An airfoil shape parameter investigation is conducted to identify links between airfoil shape and TBL-TE noise reduction. Transition behaviour, in particular control over the laminar separation bubble is shown to be critical to the downstream development of displacement thickness and hence predicted TBL-TE noise levels. Shape optimization results are shown for a modied NACA-0012 airfoil at a Reynolds number of 1:5 10 6 and zero angle of attack. It is found that the 12% thickness NACA-0012 airfoil can be modied such that predicted TBL-TE OASPL noise is reduced by approximately 2:9 dB while concurrently reducing drag by 21:5%.