Performance of EHD assisted convective boiling heat exchangers utilizing dielectric fluids

Electrohydrodynamics in convective boiling heat exchangers has been studied since the early 1990’s and has been shown to result in a large variation in the average performance enhancement of these systems. The behaviour of EHD assisted convective boiling heat exchangers, is still largely unpredictable owing to a number of conflicting parameters which are rarely kept constant in empirical studies, i.e. flow pattern and heat flux. In this thesis, it is hypothesised that by reducing the number of confounding variables in the experimental test conditions, and understanding the behaviour of EHD in convective boiling systems from a flow pattern dependent point of view, this can allow for the development of flow pattern dependent experimental correlations & numerical models to develop a methodology for performance prediction, control strategies and system integration for an EHD assisted convective boiling heat exchange device. A 30 cm long, smooth, concentric, annular test section is used to analyse the effect of EHD on convective boiling performance under constant flow pattern, constant, low heat flux, and negligible free charge conditions. Saturated boiling conditions for flow-rates between 60 kg/m2s and 180 kg/m2s and thermodynamic quality range of 0.25 - 0.55 were tested. Heat transfer enhancement ranged from 0.95 to 2.3 fold and pressure drop penalty varied from 1.4-3 fold over these test conditions. The local EHD behaviour was found to be more consistent along the axial length of the test section compared to empirical data in the literature, which uses much longer test section lengths, where flow pattern can vary. An experimental database of EHD convective boiling data for horizontal annular electrode geometries was compiled to be used for analysis purposes. The performance of the heat exchanger in both free-field and high voltage conditions could be explained by looking at the flow patterns in each case. Electrostatic modelling was used to determine electric field strength distributions and interfacial stress due to the dielectrophoretic and electrostriction forces on the liquid vapour interface, which induce liquid extraction based flow pattern re-distribution in two phase dielectric flows. A fully coupled 2D, adiabatic numerical model for the effect of the electric body force on two phase flow pattern distribution was developed. Charge was neglected in this model. Two different models for the interface migration were used and compared; a moving mesh (MM) interface tracking model and a volume of fluid (VOF) interface tracking mode. Both were verified against published experimental data. For the liquid extraction verification case, the VOF model suffers interface stretching up to 300% resulting in a 42% slower extraction time and underestimated forces. However, it is useful to use the VOF model when simulating complex flow patterns which are subject to topological changes like bubble detachment or droplet coalescence as these cannot be simulated with the moving mesh model. The moving mesh model can be used to determine the error in forces and phase velocities when using the VOF model. A methodology for generating two-phase EHD flow pattern maps was developed by incorporating the electric Froude number into each of the flow pattern transition equations. A semi-analytical model was developed to determine the maximum interfacial stress due to EHD for stratified flows to reduce the requirement of numerical modeling, and thus the flow pattern map generation methodology is fully equation based. Although transition equations developed by multiple researchers were used and compared, it is recommended that the Steiner transitions equations be used for EHD two-phase flow pattern mapping, until more fundamental experimental data can be gathered to modify the semi-empirical transition equations used in more state-of-the-art maps. EHD was found to significantly affect the “stratified-stratified wavy (SSW)” and “stratified wavy – intermittent/annular (SWIA)” transitions for concentric horizontal geometries, with minimal effect on the transition to dryout and no effect on the “intermittent dispersed bubbly (IB)” transition. The EHD flow pattern maps were generated and compared against data from the present study and a database of experimental EHD convective boiling studies. The regions where maximum enhancement were seen in the literature correlate well with those regions predicted by the maps. Performance correlations for the EHD convective boiling heat transfer and pressure drop were developed. They are based on the free-field Kandlikar correlation [1] for two-phase heat transfer and the Chisholm-Laird [2] correlations for two-phase pressure drop, respectively. The EHD flow pattern map is used to determine what the flow pattern for a given applied voltage will be, and flow pattern based enhancement linear multipliers are then used to determine the EHD performance above the free-field case. EHD is a form of active enhancement, i.e. it requires power. Thus, it would be used in systems that require performance control or regulation, in addition to some niche applications like space where it can be used instead of gravity. A method for EHD controller design was established and an EHD control algorithm was designed and implemented on the test section for the flow pattern and applied waveforms that were determined to be optimal to maximize enhancement in this geometry. System identification was performed empirically to determine the transfer function between EHD voltage and heat load to be controlled for. This resulted in a 1st order plus dead-time model to which proportional-integral controller constants were tuned. Two controllers were developed; a PID control system and a Smith model predictive control system and these were compared based on their ability to regulate the output quality of the heat exchanger when subject to dynamic heat loading. Regulation was achieved for a dynamic heat load within ±25% bound from the designed steady state load. These controllers operate on one flow pattern as the test section is 30 cm long. Flow pattern dependent controller design would be required for a full length convective boiling heat exchanger.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.