Publikationen

Lorentz force velocimetry (LFV) is a non-contact electromagnetic flow measurement technique for electrically conducting liquids. It is based on measuring the flow-induced force acting on an externally arranged permanent magnet. Motivated by extending LFV to liquid metal two-phase flow measurement, in a previous test we considered the free rising of non-conductive bubbles/particles in a thin tube of liquid metal (GaInSn) initially at rest. We observed that the Lorentz force signals strongly depend on the size of the bubble/particle and on the position, where it is released. Moreover, the force signals cannot be reproduced in detail, which necessitates a statistical analysis. This is caused by chaotic trajectories due to the rising velocities of about 200 mm/s. Therefore, in this paper, we use an improved setup for controlled particle motions in liquid metal. In this experiment, the particle is attached to a straight fishing line, which suppresses any lateral motion, and is pulled by a linear driver at a controllable velocity (0-200 mm/s). For comparison, we solve the induction problem numerically using Oseen's analytical solution of the flow around a translating sphere that is valid for small but finite Reynolds numbers. This simplification is made since the precise hydrodynamic flow is difficult to measure or to compute. The aim of the present work is to check if our simple numerical model can provide Lorentz forces comparable to the experiments. Although Oseen's solution becomes inaccurate near the sphere for finite Reynolds numbers, it provides a fore-aft asymmetry of the flow and is globally well-behaved. It provides an upper limit to the measurement results. We recover the peak-delay of the Lorentz force signals as well.

A theoretical investigation of the thermal performance (coefficient of performance COP and specific cooling power SCP) of a two bed Adsorption Ice Production AIP system based on the Silica gel-methanol as adsorbent- refrigerant in the first bed and activated carbon-methanol in the second bed is presented in this paper. Two fined-tube heat exchangers were designed (named SG-bed and AC-bed) in order to generate the same desorbed refrigerant amount of 1 kgmeth and to contain two different adsorbents. The mass transfer limitations from both the two beds and the heat transfer ability between the particles of adsorbents and heat exchanger fins are taken into account in the simulated model based on the linear driving force LDF model. To desorb 1 kgmeth from the SG-bed and AC-bed a cycle simulation computer program of the AIP system was developed to investigate the effect of desorption temperature Tdes, adsorption temperature Tads and the effect of difference of the required desorption/adsorption time on the system performance and on the amount of the ice produced per cycle mice. In the present simulations, the variation of the heat source temperature from 65 to 100 oC and chilled water temperature from 15 oC to 25 oC are taken. The results showed, that the AIP system attains a coefficient of performance COP of 66 % when the AC-bed is working and attains of 44 % when the SG-bed is working. The amount of the ice produced from the system estimated to 6kg per cycle (3 kg is produced from each of bed), but the Qin input energy required to activate the AC-bed has been saved by 46 % compared with that required to activate the SG-bed. Although each of the adsorbent beds was filled with different amount of the sorption material, it is found that the mass of the sorption materials inside the both beds has no effect on the cycle time but has important effect on the specific cooling power SCP. The cycle time is strongly dependent on driven temperature of heat exchange fluid, the design of the heat exchanger and the mass transfer coefficient of sorption material Dso. An experimental set up is planned to be built to make validation of the simulation results.

This paper demonstrates the possibility of capturing full-resolution high-speed image sequences using a regular 3CCD color camera in conjunction with high-power light emitting diodes of three different colors. This is achieved using a novel approach that is referred to as spectral-shuttering where a high-speed image sequence is captured using short duration light pulses of different colors that are sent consecutively in very close succession. The work presented in this paper demonstrates the feasibility of configuring a high-speed camera system using low cost and readily available off-the-shelf components. This camera can be used for recording six frame sequences at frame rates up to 20kHz or three frame sequences at even higher frame rates. Both color crosstalk and spatial matching between the different channels of the camera are found to be within acceptable limits. A small amount of magnification difference between the different channels is found and a simple calibration procedure for correcting the images is introduced. The images captured using the approach described here are of good quality to be used for obtaining full-field quantitative information using techniques such as digital image correlation and particle image velocimetry. A sequence of six high-speed images of a bubble splash recorded at 400Hz is presented as a demonstration.