Dynamic tracking of structures in multiphase flows by ultrafast X-ray tomography and image based scanning steering

点击翻译按钮获取中文摘要

Ultrafast imaging techniques are indispensable tools for the analysis of highly dynamic processes such as multiphase flows. The latter occur in many technical processes, e.g. in chemical multiphase reactors, in manifold ways. Modelling of multiphase flows using numerical calculation schemes (CFD) is especially demanding because of the multitude of physical effects to be considered for impulse, heat and mass transfer along with the multiple scales of the problem. For the development and validation of models for the numerical calculation of multiphase flows temporal and spatial highresolution measurement techniques, such as the ultrafast X-ray tomography system (ROFEX), which was co-developed by the applicant, are essential. The dimensioning of bubble columns, for example, relies significantly on the knowledge about the residence time behavior of the liquid as well as the gaseous phase. However, there are only few studies on the tracking of single bubbles in bubble columns, which are furthermore mainly limited to bubble columns with low gas fractions, where the single bubbles are not affected by swarm effects. Fluid mechanic investigations for determining the above mentioned parameters within a bubble swarm would strongly benefit from an online measurement technique for simultaneous measurement of bubble structure, bubble movement and surrounding phase distribution within realistic bubble columns.Therefore, the aim of the project is the advancement of the ultrafast X-ray tomography for tracking bubbles in a two-phase flow. The principle is based on a fast image steered positioning of the to-mography plane. Thus, gas bubbles can be tracked throughout their way through the flow region. During their passage, structural changes as well as interactions with other gas bubbles (coales-cence, fragmentation, and collision) can be observed. Furthermore, the complete flow structure within the imaging plane can be recorded. For the realization of dynamic experiments, efficient measurement, analysis and controlling techniques are essential. They are supposed to be obtained by interdisciplinary cooperation with the group of A. Kopmann (KIT), which possesses proved expertise in online data acquisition of large data amounts, 3D image reconstruction and image based controlling. Based on this knowledge, GPU based image reconstruction and image analysis algorithms in combination with a fast data transfer solution should be designed, developed and implemented at the ROFEX system. Concrete aim is the demonstration of the feasibility of tracking a single bubble of a bubble swarm within a bubble column. The main challenge lies in the verification of the realtime capability of data transfer and the whole data analysis chain.