English

ln the paper a numerical program package is described to calculate incompressible, unsteady, three-dimensional, viscous and turbulent flow fields around sharp edged obstacle. By this the velocity and pressure distributions in the flow field and on the surfaces of square-formed bodies in a plane channel can be determined, as well as the frequencies of periodic vortex separations The channel consists of two plates extended to infinity. On the lower plate the square-formed body, which is identical with the building model, is placed.

There is a need to calculate root mean square (r.m.s.) pressures from the output of steady-state computer programs. We know much less about calculating r.m.s. pressures than about calculating r.m.s. velocities. R.m.s. pressures can be quickly estimated from calculated mean pressures, mean velocities and r.m.s. velocities using the equations in this paper. The equations have been used in \Mind Engineering but can be applied in any turbulent flow where pressures are required.

This paper investigates the role of turbulence models in numerical calculations of flow over obstacles with second-moment closure models. Two models for the pressure-strain correlations are examined in the study. Computations of the main characteristics of the mean flow and the turbulent fields are compared against experimental data, and results obtained with the standard k-e model. All models give reasonable agreement with the data.

Flowfields around bluff bodies are characterized by complex distributions of the strain-rate tensor. Such flowfields can be analyzed with various turbulence models. The shortcoming of the eddy viscosity modelling in the k-e model is scrutinized in comparison with the results of ASM. The accuracy of the algebraic approximation adopted in ASM is examined using the numerical data given from LES. A new LES model with variable Smagorinsky constant is then presented.

The paper considers issues pertaining to the capabilities and limitations of computational methods for multidimensional turbulent flows of the type encountered in fluids engineering. It argues that CFD, whilst offering considerable predictive power and potential, is not yet sufficiently well established to be applied routinely to complex 3D flows, unless only a rough qualitative. statement is being sought.

Since the cost of energy is increasing sharply a trend to conserve energy in the indoor environment and in addition to improvements in thermal insulation, two possible solutions are adopted. The first one is to provide reduced air gaps and opening for newly constructed buildings to minimise the infiltration of outdoor air. The second one is to reduce the ventilation rate or the fresh air supplied in air conditioned buildings. These two solutions are the reason for some serious problems of indoor air quality.

In order to evaluate the accuracy of COMIS predictions for large openings,and to study its sensitivity, two tests have been performed. In the first test, COMIS is used together with four existing multizone air flow models to calculate natural ventilation in a building for various climatic and opening configurations. In the second test, COMIS predictions are compared with single-sided ventilation measurements taken in test cells. The results of the tests are reported.