Sharpe, Patrick
(2023)
Characterising natural ventilation through open windows in the presence of wind: a theoretical and experimental investigation into the interaction between window geometry and environmental forces, and its application to envelope flow models of natural ventilation.
PhD thesis, University of Nottingham.
Abstract
Natural ventilation systems utilise pressure differentials that arise from wind and buoyancy forces to drive air through buildings. Natural ventilation is typically described using envelope flow models. These are used to size window openings at the design stage, and to predict the annual dynamic thermal performance of buildings. However, envelope flow models rely on highly idealised descriptions of flow through ventilation openings, which do not model realistic window geometries encountered in practice, and assume that ambient air is static. When envelope flow models are applied to building design, inadequate accounting for phenomena relating to wind and opening geometry can lead to under-sized ventilation openings and under-performing buildings.
This thesis develops empirical models that characterise the effect of wind and window geometry on ventilation rates through square orifices and square, hinged openings. To ensure that these models can be applied in the design case to an arbitrary building geometry, these models are characterised using conditions local to the opening. Here the effect of wind is accounted for by a simulated cross-flow parallel to the building façade, and dimensional scaling arguments are applied to develop empirical models of wind driven phenomena based on similarity theory. Three experimental conditions are modelled: still-air tests that measure the ventilation capacity of an opening in idealised conditions; local pressure tests that measure the wind-induced static pressure differential over an opening when no flow occurs through that opening; and dynamic flow tests that measure the ventilation capacity of an opening between these limits.
To validate the use of these local-scale models in building design, this work is then extended to predict the ventilation in a simple two-zone building. This requires the measurement of the speed and direction of the cross-flow on the building façades, and to that end a novel probe is developed that enables simultaneous measurements of these parameters. Wind-tunnel experiments are then used to measure the ventilation rate achieved in the model building, and the results are compared against the predictions of the local-scale window-characterisation models developed in this thesis. The results show an improvement over current models, which tend to overestimate ventilation rates.
This thesis shows that free area models, which are widely used to predict the ventilation capacity of windows, tend to systematically overestimate ventilation rate through simple hinged openings in still air. The Empirical Effective Area Model described in this thesis can be used to predict idealised discharge coefficients with a coefficient of determination of 0.98, compared to 0.57-0.74 for free area models.
A wind-driven cross-flow is shown to interact with window geometry to alter the local pressure field over the surface of an opening. This thesis develops experimental techniques to characterise this change in pressure using a local pressure coefficient. This is used to specify a local dimensionless pressure which is shown to describe the transition between inflow and outflow through an opening. Empirical equations are developed that characterise the local pressure coefficient for square hinged windows as a function of flow approach angle and opening angle, with a coefficient of determination of 0.98.
The generation of non-zero local pressure coefficients is shown to result in orifice discharge coefficients that tend to ±∞ as the dimensionless room pressure tends to zero. Dimensional analysis is used to suggest the total dimensionless volume flow rate as an alternative metric to characterise the ventilation capacity of an opening. This is shown to tend to the idealised discharge coefficient in still-air conditions, and to tend to zero as the local dimensionless pressure tends to zero. The total dimensionless volume flow rate is shown to be finite across the whole range of potential local dimensionless pressure values, and holds positive values for outflow and negative values for inflow. Empirical models are developed to predict the total dimensionless volume flow rate through a square orifice and a square, hinged window as a function of local dimensionless pressure, flow approach angle and opening angle. Coefficients of determination are between 0.975 and 0.984 for the hinged window and between 0.993 and 0.995 for the square orifice, depending on the opening direction and the direction of flow through the opening.
The speed and direction of the cross-flow on the façade of a model cube in a simulated atmospheric boundary layer were measured using the novel cross-flow probe developed in this thesis. Here, mean cross-flow speeds measured with the novel probe agree well with hot-wire measurements. The cross-flow measurements reveal a tetra-modal distribution in façade cross-flow direction, which interacts with a uni-modal variation in cross-flow speed to generate bi-modal distributions in the x and y velocity components at all measured wind angles. This data is used to generate profiles of mean cross-flow speed and mean façade cross-flow direction with wind angle, which are used as inputs to the empirical equations developed to describe the total dimensionless volume flow rate through isolated window openings.
The empirical models used to describe the total dimensionless volume flow rate through isolated windows is used to predict measured ventilation rates in a model building with an internal partition, with a coefficient of determination between 0.94 and 0.99, depending on the ventilation configuration. This compares with coefficients of determination between -0.4 and 0.08 found when applying a conventional orifice flow model. Conventional orifice flow models are predicted to provide good estimates of net volume flow rates through buildings with simple orifice-type openings when the internal resistance is lower than that of either of the external openings. As the internal resistance increases, the orifice flow equation is predicted to increasingly overestimate net volume flow rates.
This work contributes to knowledge by: quantifying the systematic errors arising from the use of free area models common to natural ventilation design; developing an empirical model that describes an idealised discharge coefficient of a family of hinged openings as a function of geometric parameters; identifying novel dimensionless parameters that characterise the change in static pressure across an opening that results from the interaction between wind and window geometry; developing experimental techniques that provide a simple and unambiguous measurement of the local pressure coefficient; experimentally and empirically describing the aerodynamic performance of a simple, square, hinged window in wind-driven conditions; developing a cross-flow probe that can measure the instantaneous speed and direction of wind-driven flow over the surface of a building; quantifying the systematic errors associated with the use of a range of calculation methodologies used to estimate the ventilation rate in a simple model building; and providing practical design guidance to minimise the effect of calculation errors that arise from the use of conventional envelope flow models in wind-driven conditions, when adequate data to describe the phenomena is unavailable.
| Item Type: |
Thesis (University of Nottingham only)
(PhD)
|
| Supervisors: |
Jones, Benjamin
Wilson, Robin |
| Keywords: |
Natural ventilation, Ventilation, PPOs, Windows, Wind, Wind tunnel, Characterisation, Modelling; Purpose provided openings |
| Subjects: |
T Technology > TH Building construction
T Technology > TH Building construction > TH7005 Heating and ventilation. Air conditioning |
| Faculties/Schools: |
UK Campuses > Faculty of Engineering
UK Campuses > Faculty of Engineering > Built Environment |
| Item ID: |
76633 |
| Depositing User: |
Sharpe, Patrick
|
| Date Deposited: |
05 Feb 2024 13:32 |
| Last Modified: |
05 Feb 2024 13:32 |
| URI: |
https://eprints.nottingham.ac.uk/id/eprint/76633 |
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
