M. J. Davidson

Plume dispersion through large groups of obstacles—A field investigation

Abstract A set of field experiments has been conducted at the Cardington experimental site, in the United Kingdom, to investigate the dispersion of a neutrally buoyant plume released upwind of an array of cubes, each of which measured approximately 2 m × 2 m × 2 m. The plume was released below the height of the obstacle array., and a second control plume was released alongside the array. The behaviour of the plume was recorded using both flow visualisation and measurements of gas concentration. A limited number of velocity measurements was also made. It is found that, relative to the control plume, certain mean concentration statistics do not change materially as the plume passes through the obstacle array. These include the form of the cross-sectional profiles, the decay along the centre line and the lateral growth with downstream distance. However, the mean vertical extent of the plume increases by 40–50%. In addition the r.m.s. fluctuations of concentration within the the plume are reduced. A clipped normal distribution is shown to be an appropriate model for the concentration fluctuations within the array plume.

Wind tunnel simulations of plume dispersion through groups of obstacles

In this paper we present the results of two wind-tunnel simulations of dispersion from upwind point sources through a large group of obstacles, and compare these with an associated field study (Davidson et ai., 1995, Atmospheric Environment 29, 3245–3256). Detailed flow-field and plume concentration data were obtained from simulations at scales of 1:20 and 1:200. With these data we are able to provide experimental confirmation of many of the ideas developed during the field study and to confirm the experimental results obtained in the field. In doing so, we show that the upstream flow-field parameters are the most effective means of scaling the three data sets to achieve broad quantitative agreement. Measurements and flow visualisation of the flow-field confirm that there are a number of mechanisms influencing the behaviour of a plume as it passes through an obstacle array: in particular the divergence and convergence of streamlines and changes to the structure of the turbulence within the array. However, although the turbulence within the array is shown to be of greater strength and smaller scale than at corresponding locations outside the obstacle array, it is found that there is little change in the transverse diffusivity (and therefore in the lateral plume width). The concentration data confirm that the divergence of streamlines near the upstream end of the obstacle array has a significant effect on the vertical width of a plume (σz). Changes to the structure of the turbulence appear to have little effect, however, since the transverse diffusivities within the obstacle array are unchanged. Thus, the mean lateral spread and decay of mean concentration of the plume with downstream distance resemble that of a control plume; that is, a plume released under identical conditions where the obstacle array :.s not present. We also confirm that the mean structure of a plume has a Gaussian form as it passes through. an array of obstacles. By contrast, concentration measurements with a high-frequency-response detector confirm that the small-scale, high-strength turbulence rapidly mixes the plume internally, dramatically reducing the strength of concentration fluctuations within the plume. Since the wind tunnel is shown to be an effective means of modelling this type of field situation, with the appropriate sailing, these studies were extended to consider the effects on plume behaviour of changes in source position, array configuration and array height.

Concentration fluctuation measurements in tracer plumes using high and low frequency response detectors

A set of tracer experiments designed to compare two concentration fluctuation detectors and measure fluctuation statistics at high frequencies is described. A detector which has been used in several previous fluctuation experiments (the TIP photoionisation detector manufactured by Photovac of Canada) is compared with another with a much higher frequency response (the flame ionisation detector — FID — made by Cambustion of the UK). Good agreement is found and results show that the signal optimization system used in previous work with the TIP provides an accurate enhancement of the instrument output, thus improving confidence in the results of previous papers. They also confirm that the TIP detector is able to resolve most of the concentration variance in most situations of interest, but not at very short range. Measurements of the high frequency end of the fluctuation spectrum using the FID show inertial-convective subrange behaviour at frequencies not resolved by the TIP, supporting earlier work. Fluctuation spectra measured very close to the source are also shown to have a characteristic +2/3 power law behaviour (when nSc (n) is plotted against n) at lower frequencies, in agreement with theoretical predictions.

Analysis of transient signals in simple pipeline systems with an extended blockage

This paper investigates the effect that an extended blockage has on a transient signal by examining numerical and experimental results in the time and frequency domains. Extended blockages can develop in pipelines via the processes of tuberculation, scaling, bio film growth, sediment deposition or through designed reductions in pipe diameter. It is identified that consideration of the gradual formation of these faults is important in pipeline design as they can significantly change the fundamental period of a pipelines transient response, maximum and minimum transient pressure heads and the evolution of the transient signal, thus potentially increasing a systems susceptibility to failure. Analysis of transient signals affected by a blockage provides information on the property and location of the extended blockage. In this paper, comparisons between discrete and extended blockage models are made and the effects of changes in blockage diameter, length and wave speed on the transient response are investigated. Furthermore, the ability of existing models to represent extended pipeline faults under transient conditions is evaluated through comparisons of numerical transient responses with new experimental results from the laboratory. Results show that the transient behaviour is modelled with a good level of accuracy over the first few periods of oscillation and that the level of accuracy decays with time. A periodic Fourier analysis of the data demonstrates that the damping rates of the experimental response are higher than the numerical predictions. Applying this new method of signal analysis has shown that the transient signal exhibits larger damping rates in the higher frequency components of the response.

The behaviour of a single, horizontally discharged, buoyant flow in a non-turbulent coflowing ambient fluid

An integral model of the behaviour of a single horizontally discharged buoyant flow in a coflowing ambient fluid is presented. The usual integral equations for the conservation of mass, momentum and buoyancy flux are used. However, the model differs from those presented previously by using the spread assumption rather than the more usual entrainment assumption and by explicitly forcing the initially gaussian velocity distribution to become a thermal distribution. The position and the form of this transition is determined from laboratory data (Brown [1984], Knudsen [1988]) and field data (Lee and Neville-Jones [1987]). The performance of the model is compared with both laboratory and field data and it reproduces the data for a wide range of cases. A model of this type is essential in gaining an understanding of the range of behaviour of merging plumes (Cheng, Davidson and Wood) and is a preliminary to the understanding of the more general case where the buoyant discharge is ejected at an angle to the flow.

A study of a buoyant axisymmetric jet in a small co-flow

This paper deals with measurements of and an integral model for a buoyant axisymmetric jet in a very small co-flow. The integral model is based on the insights gained from the measurements presented here and other recent experiments, which suggest that the turbulent flow is simply advected and the entrainment is that normal for a jet or plume in a still ambient fluid. General equations for the turbulent flow trajectories are also developed. The predictions of the theory are verilied for the no cross-flow case and the cases where the jet or plume is ejected vertically or horizontally into a very small cross-flow. The results of experiments in which a buoyant jet is released in the same direction as the horizontal ambient flow, show that outside the turbulent region the entrainment velocities can be represented with uniform flow and the appropriate sink. Direct measurement of the strength of the sink allows the transition from weakly- to strongly-advected behaviour to be determined. The departure of the trajectory measurements from the theory also provides information about the transition. The transition location is then compared with recent measurements in which the jets and plumes are ejected vertically into a very small cross-flow.

The front condition for intrusive gravity currents

A mathematical model is developed for a high Reynolds number, quasi-steady, intrusive gravity current propagating into a two-layer ambient fluid. The model is based on inviscid, irrotational flow theory, and is used to predict the rate of advance of the current front as a function of its head height. The key underlying assumptions of the model are that the local conditions at the head determine its speed of propagation, and that the upstream portion of the head is well approximated by an inviscid, irrotational flow. In order to provide a complete set of boundary conditions it is assumed that the ambient flow over the top of the head is horizontal, but not uniform. An upper bound solution is derived and is compared with a full solution generated with an optimisation based Boundary Element Method technique. The upper bound solution is shown to be entirely adequate for relative current heights greater than 0.3. The results from two experimental programmes provide support for the model predictions, and particle tracking velocimetry measurements indicate that the assumption of horizontal flow over the head is reasonable.

On the behaviour of advected plumes and thermals

The results of an experimental investigation into the behaviour of advected plumes are presented. Measurements of tracer concentration are made using laser-induced fluorescence and image processing techniques. These measurements provide the basis for studying the bulk properties and fluctuation statistics of the flows. The experimental results show that the appropriate characteristic velocity for predicting the centreline location of an advected plume is the cross-sectional average velocity and not the centerline velocity (which has been utilised for this purpose in the past). The reason for this is discussed in the context of the large scale turbulent structures evident in the flow. In addition, it is shown that reasonably accurate predictions of the bulk properties of advected plumes can be made by assuming the behaviour of an advected plume is similar to that of a plume discharged in a still ambient fluid (a still plume), which is simply advected by the ambient current. Notable differences can be found...

Atmospheric Flow through Groups of Buildings and Dispersion from Localised Sources

We present experimental and theoretical investigations into the problems of wind flow and scalar dispersion within and around a group of buildings.

Removing the boundary influence on negatively buoyant jets

A comprehensive laboratory study of negatively buoyant discharges is presented. Unlike previous studies, here the focus is on generating data sets where influences of the bottom boundary have been eliminated. There are significant discrepancies in the published dilution data for these flows and a contributing factor is the large variation in the bottom boundary condition. A Laser-induced Fluorescence system is employed to gather flow spread, peak concentration (minimum dilution) and trajectory data for a wide range of densimetric Froude numbers and initial discharge angles. Data from these experiments are compared with previously published data, along with predictions from integral models and a revised form of the previously published semi-analytical solutions. The new data sets are not distorted by mixing processes associated with the bottom boundary and therefore provide the basis for more meaningful assessments of the predictive capabilities of existing models, given that the influences of the bottom boundary on contaminant mixing are not incorporated into these models. In general the models assessed are able to predict key geometric quantities with reasonable accuracy, but their minimum dilution predictions are conservative. Importantly dilution at the return point shows a strong dependence on the initial discharge angle and this could have important implications for the design of discharge systems.