Abstract

The inverse estimation of a source location of pollutant released into a turbulent flow is a probability problem instead of a deterministic one, as the turbulent flow is chaotic and irreversible. However, researches can be conducted to provide helpful instructions to the possible source location with corresponding uncertainty. This study aims to propose a method of inverse estimation of a passive-scalar source location. Experimental investigation of the dye plume characteristics released into a fully-developed turbulent flow is performed in a water channel. A planar laser-induced fluorescence (PLIF) technique is used to obtain two-dimensional images of spreading dye plumes at a bulk Reynolds number of 20,000. The distributions of high concentration areas in the PLIF images are chosen as features that characterize the traveling (diffusion) distance or time from the dye source. Graphical analysis is used to extract these high concentration areas. The procedure of graphical analysis has three steps: 1) binarization using a threshold to extract high concentration dye patches; 2) labeling individual high-concentration dye patches in the binarized images; and 3) pixel-counting to measure the area and perimeter of each dye patch. We examine the variations of fractal dimension of patches, and the fractal dimension is observed to be almost constant irrespective of the distance from the source. The kurtosis of the probability density function curve of the logarithm dimensionless dye patch areas is found to be related with the downstream diffusion distance, based on which an inverse estimation method to locate a passive-scalar point source is proposed and evaluated.

Highlights

  • A problem commonly encountered in emergent chemical leakage is the quick localization of the leaking source using the data observed at monitoring sites

  • This study proposes a possible inverse estimation method for the source location of a dye plume in a turbulent flow

  • A water-channel experiment was carried out simulating the dispersion process of injected dye solution into a quasi-homogeneous turbulence in the channel central region, and the dye plume images are captured through a planar laser-induced fluorescence (PLIF) technique

Read more

Summary

Introduction

A problem commonly encountered in emergent chemical leakage is the quick localization of the leaking source using the data observed at monitoring sites. In order to describe the intermittency characteristics of scalar field in turbulent flow, the intermittency was defined and studied [8] [9] [10] Based on those previous researches [8]-[16], many quantities are found to be related with the diffusion distance of dye plumes: these quantities includes time-average concentration, plume width, concentration fluctuation quantities, intermittency, concentration burst quantities, spatial correlation, and integral length. Page et al [21] declared that the high-intensity spike encounters dominate the tracking motion of a crab, and high frequency of odorant spike encounters terminates its upstream movement These high-intensity pulses are called dye patches in this paper, which contain much information about source location. A possible sourcelocating estimation method would be proposed in this study

Experiment Setup
Graphical Analysis
Binarization to Highlight Dye Patches
Connected Components Labeling to Separate Dye Patches Apart
Pixel-Counting to Calculate Areas and Perimeters of Dye Patches
Fractal Dimension
Dye Patch Area Distribution
Proposal of a Source Locating Method by Kurtosis
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call