Comprehensive analysis of thermoelectric generation systems for automotive applications

Autor: Zhijia Yang, M.A. Wijewardane, Richard Stobart
Rok vydání: 2017
Předmět:
Zdroj: Applied Thermal Engineering. 112:1433-1444
ISSN: 1359-4311
DOI: 10.1016/j.applthermaleng.2016.09.121
Popis: With the introduction of carbon dioxide emissions legislation for vehicles, the pressure on fuel efficiency in vehicle propulsion systems has grown significantly. Cost-effective efficiency improvements have become the topic of intensive research efforts. Amongst such efficiency measures, waste heat recovery (WHR) has attracted a deep interest both in the industrial and academic sectors. As a potentially low maintenance solid-state implementation, the thermo-electric generator (TEG) represents a promising candidate technology. Thermoelectric (TE) solutions, compared with other WHR methods have an appealing simplicity that could translate rapidly into robust engineering solutions. Achieving competitive efficiencies and low manufacturing cost however remains a substantial research challenge. Critical to progress are modelling processes that allow solutions to be formulated and assessed. The work reported in this paper demonstrates that a modelling process that makes use of mainstream computational fluid dynamics (CFD) codes is feasible. A benchmark TEG design first simulated and then run on an engine test bed showed agreement between simulation and experiment to within 10%. Closed form results for the optimised performance of module and overall TEG design have been reported in the literature and lend important insights into implementation methods. However practical implementation must take account of varying conditions and spatial variations in the heat exchange process. A CFD code that would permit a detailed evaluation TEG parameters and material properties has been demonstrated with a plate-fin design of heat exchanger. Meanwhile a simpler model has achieved agreement to within 12% with the CFD code indicating that a rapid modelling process is feasible and could support new “in the loop” test techniques.
Databáze: OpenAIRE