Vision-based experimental investigation of two-phase flow in a rising pipe

Abstract

Multiphase pipeline flow is very important to oil, gas and process industry. This is a class of problems where comprehensive simulations, although desirable on physical grounds, are not practical due to the large aspect ratio between the pipe length and diameter. In order to enhance the understanding of the pipeline flow, as well as to validate and identify the critical aspects ofthese models, intensive experimental data are needed. This will defme how these models can be improved. In this paper, we report our recent work in this aspect. A transparent pipeline, consisting of horizontal and rising sections, was established for conducting the experiments. The purpose of the experiments was to investigate the transition between the basic flow patterns. The flow states were recorded into image sequences using digital camera. Since the required viewing area was much larger than that the required magnification could be satisfied, Image sequence analysis algorithms were developed based on optical flow detection theory to overcome this issue. Apart from the algorithm and experimental results, we also report our further planed research.