Development, manufacture and full experimental validation of an exhaust heat thermoelectrical generator with temperature control.

Universidade do Minho Escola de Engenharia Rui Miguel da Silva Carvalho Development, Manufacture and Full Experimental Validation of an Exhaust Heat Thermoelectrical Generator with Temperature Control december 2021 Universidade do Minho Escola de Engenharia Rui Miguel da Silva Carvalho Development, Manufacture and Full Experimental Validation of an Exhaust Heat Thermoelectrical Generator with Temperature Control. Dissertação de Mestrado Mestrado Integrado em Engenharia Mecânica Trabalho efetuado sob a orientação do Professor Doutor Francisco Carrusca Pimenta Brito e coorientação do: Professor Doutor Jorge José Gomes Martins december 2021 ii DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Licença concedida aos utilizadores deste trabalho Atribuição CC BY https://creativecommons.org/licenses/by-nc-nd/4.0/ iii ACKNOWLEDGMENTS Quero neste pequeno capítulo exprimir os meus agradecimentos a todos aqueles que, de alguma forma, me acompanharam e fizeram parte deste trabalho. Em primeiro lugar e sempre em primeiro, à minha família, em especial aos meus pais e à minha avó, por me terem proporcionado todas as condições possíveis para que todo este percurso fosse alcançado e por me terem dado a educação e as ferramentas que me tornaram quem sou hoje. Como acredito no valor e importância do legado da família, esta conquista é também vossa. Em segundo lugar gostava de expressar o meu profundo agradecimento ao Professor Francisco Pimenta Brito, que foi muito mais além do que o significado e a responsabilidade que a palavra Orientador carrega. Obrigado pela transmissão de conhecimento, cultura, valores e pela amizade que levo comigo após um ano de convivência quase diária. Em terceiro lugar aos meus amigos Carlos e o Joaquim, pelas manhãs e tardes passadas no laboratório de motores, pelas risadas, palhaçadas, pelo stress constante e por todas as peças que quebravam diariamente ou falharam durantes os ensaios experimentais, pelo companheirismo e pela constante motivação, valor e confiança que em mim depositaram ao longo deste ano. Gostava também de deixar uma palavra de agradecimento a todos aqueles que de algum modo ajudaram a que este trabalho fosse concluído. Em especial ao Engenheiro João Carvalho por todo o acompanhamento atento e conselhos incisivos, ao Professor Doutor Jorge Martins pela constante motivação, ao Engenheiro Filipe Marques pelas horas perdidas em reparações de última hora. Não fica esquecido o agradecimento à Energest® e à BorgWarner® por serem os nossos parceiros na construção destes protótipos inovadores de permutadores de calor. Por fim, a todos os meus colegas com quem partilhei os últimos cinco anos da minha vida, em especial à amiga Mariana que esteve sempre ao meu lado nos melhores e nos piores momentos, que me acompanhou até à Europa de leste para um semestre peculiar e com quem viajei por esse mundo fora. Entregaste a tese primeiro que eu já viste? Desejo aos nossos filhos e aos filhos deles terem a sorte de poderem encontrar amizades como a nossa. iv STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho. v Ao meu avô. vi ABSTRACT The automotive industry is pressed with increasingly stringent efficiency and emissions goals. Vehicles are having increasing degrees of hybridization. Therefore, on-board electricity production seems highly attractive. Thermoelectric (TE) generators can directly convert the exhaust heat into electricity with no moving parts and unlike competing technologies, need little to no maintenance if carefully designed. However, they are temperature limited, and their electric output is very sensible to the thermal level. So, it is a challenge to design a system with optimized operation under the wide range of driving conditions. A novel Temperature-Controlled Thermoelectric Generator (TCTG) concept incorporating thermal control through a phase change heat spreading device incorporated into the exhaust of a vehicle was previously proposed and modelled by the group. It incorporates corrugated pipes for exhaust flow, which are embedded in a cast aluminium matrix along with variable conductance heat pipes (VCHPs) working as excess heat spreaders. This concept is designed to maximize the absorption of exhaust heat under highly variable thermal load applications while still avoiding both thermal dilution at low loads and overheating at high loads. Wherever the temperature limit is exceeded at the hotter upstream regions of the heat exchanger (HX), heat is absorbed by vaporization and spreads to the colder regions of the heat exchanger by condensation. Therefore, in high thermal load events the excess heat is spread along the HX instead of being wasted through by-pass systems, as done in existing systems. The present work performed the first full experimental validation of this concept and of the corresponding numerical model. Firstly, a previously built proof-of-concept prototype was instrumented and tested at ambient pressure with a resulting heat spreading temperature around 100ºC. Subsequently, a new TCTG prototype capable of withstanding the operating pressures needed for optimal heat spreading temperature (∼40 bar / 250ºC) was built, attached to the exhaust of an Internal Combustion Engine (ICE) and tested. Evidence for the temperature control and heat spreading previously claimed for the concept was thus obtained for the first time. Results indicate that the system is able to maximize heat absorption under highly variable thermal load and achieve TE conversion efficiencies close to the theoretical maximum, with an electrical output level that seems to be a breakthrough for road vehicle TE generators, also showing promising results in terms of fuel consumption and 𝐶𝑂2 emissions reduction. Key-Words: Thermal control; Thermoelectric generator; Variable conductance heat pipes; Variable thermal load; Waste heat recovery. vii RESUMO A indústria automóvel é pressionada com objetivos de eficiência e emissões cada vez mais rigorosos. Os veículos apresentam graus crescentes de hibridização, sendo que a produção de eletricidade a bordo parece altamente atrativa. Os geradores termoelétricos (TE) podem converter diretamente o calor de escape em eletricidade sem peças móveis se forem cuidadosamente concebidos. Porém, têm limitações de temperatura, e a sua produção elétrica é muito sensível ao nível térmico. Desse modo, é um desafio conceber um sistema com funcionamento otimizado sob a vasta gama de condições de condução. Um novo conceito de gerador termoelétrico com controlo de temperatura (TCTG), incorporando o controlo térmico através de um dispositivo de distribuição de calor por mudança de fase foi anteriormente proposto e modelado pelo grupo. Este, incorpora tubos corrugados para o fluxo de escape numa matriz de alumínio fundido juntamente com os heat pipes de condutância variável (VCHPs) funcionando como espalhadores de calor em excesso. Este conceito maximiza a absorção de calor de escape sob aplicações de carga térmica variável, evitando ao mesmo tempo a diluição térmica com cargas baixas e o sobreaquecimento com cargas altas. Sempre que o limite de temperatura é excedido nas regiões mais quentes a montante do permutador de calor (HX), o calor é absorvido por vaporização e espalha-se para as regiões mais frias do permutador de calor onde condensa. Portanto, em eventos de carga térmica elevada, o excesso de calor é espalhado ao longo do HX em vez de ser desperdiçado através de sistemas de by-pass, como acontece nos sistemas existentes. O presente trabalho efetuou a primeira validação experimental deste conceito e do modelo numérico correspondente. Primeiramente, um protótipo previamente construído foi instrumentado e testado à pressão ambiente, com uma temperatura de espalhamento de calor de cerca de 100ºC. De seguida, foi construído um novo protótipo TCTG capaz de suportar as pressões de funcionamento necessárias para uma temperatura ótima de espalhamento de calor (∼40 bar / 250ºC), ligado ao escape de um Motor de Combustão Interna (ICE) e testado. Foram assim obtidas, pela primeira vez, provas para o controlo da temperatura anteriormente reivindicado para o conceito. Os resultados indicam que o sistema é capaz de maximizar a absorção de calor sob carga térmica variável e alcançar eficiências de conversão próximas do máximo teórico, com um nível de potência elétrica inovador para geradores TE de veículos rodoviários, mostrando também resultados promissores em termos de consumo de combustível e de redução de emissões de CO2. Palavras-Chave: Controlo térmico; Gerador Termoelétrico; Heat Pipes de Condutância Variável; Recuperação de calor de escape; TEG. viii TABLE OF CONTENTS Acknowledgments ............................................................................................................................... iii Abstract.............................................................................................................................................. vi Resumo............................................................................................................................................. vii Table of contents .............................................................................................................................. viii List of Figures .................................................................................................................................... xii List of Tables .................................................................................................................................... xvii List of Acronyms and Abbreviations ..................................................................................................... 1 List of symbols ................................................................................................................................... 3 1. INTRODUCTION ............................................................................................................................. 6 1.1 Motivation ................................................................................................................................ 6 1.2 Previous Research Group Survey ............................................................................................. 10 1.3 Thermal control through excess heat spreading ....................................................................... 14 1.4 Objectives of the Dissertation .................................................................................................. 17 1.5 Structure ................................................................................................................................ 18 2. FUNDAMENTAL CONCEPTS AND LITERATURE SURVEY ................................................................ 19 2.1 Thermoelectricity .................................................................................................................... 19 2.1.1 The Seebeck, Peltier and Thomson Effect ......................................................................... 19 2.1.2 Thermoelectric Modules ................................................................................................... 22 2.1.3 TEG Electrical Output ....................................................................................................... 25 2.1.4 Maximum Theoretical Power ............................................................................................ 25 2.1.5 Conversion Efficiency and Figure of Merit ......................................................................... 26 2.1.6 Module Maximum Allowed Temperature and Improvements ............................................. 28 2.2 Thermodynamics .................................................................................................................... 30 2.2.1 Basic Concepts ................................................................................................................ 30 2.2.2 Zero law of Thermodynamics ........................................................................................... 30 2.2.3 First law of Thermodynamics ........................................................................................... 30 2.2.4 Second law of Thermodynamics ....................................................................................... 31 2.2.5 Heat Transfer Concepts ................................................................................................... 32 2.2.6 Heat Exchanger Basic Concepts ....................................................................................... 34 xv Figure 74 - (a), (c) side-view and (b) top-view of the moulding box after alluminium casting .............. 104 Figure 75 - Problems found after the casting. .................................................................................. 105 Figure 76 - (a) Moulding Box extraction and cooling (b) inside and (c) outside. ................................. 105 Figure 77 - (a) HX being cleaned and (b) sand being removed. ........................................................ 106 Figure 78 - Verification of the integrity of the end of the corrugated tubes and screws ...................... 106 Figure 79 - Weighted TCTG after casting ......................................................................................... 107 Figure 80 - (a) 6 pad milling cutter tool, (b) milling machine manufacture spindle rotation speeds and (c) machining process of the HX .......................................................................................................... 107 Figure 81 - Preliminary Models of the final prototype. ...................................................................... 108 Figure 82 - Final design of the Prototype. ........................................................................................ 108 Figure 83 - Prototype delivered by ENERGEST. ................................................................................ 109 Figure 84 - (a) Excess condensers positioning and (b) close up view. ............................................... 110 Figure 85 - Expansion Valve. ........................................................................................................... 110 Figure 86 - (a) TEG preparation and (b) cooling plate cleaning. ........................................................ 111 Figure 87 - (a) TEG placement on the HX and (b) alumina sheets covering ....................................... 111 Figure 88 - Prototype after the assembly of the TEGs and the Cooling Plates .................................... 112 Figure 89 - Prototype after load resistance connected to the TEGs ................................................... 112 Figure 90 - (a) Exhaust Gas Collector and (b) possible leakage sealing. ............................................ 113 Figure 91 - Water supply system connections. ................................................................................. 113 Figure 92 - (a) Exhaust gas collector connected to exhaust pipe and (b) coated with glass wool. ....... 114 Figure 93 - (a)Exhaust Gas Collectors position in TCT;(b) Exhaust Collectors Sheet Metal Drawing, (b) MIG Welding on exhaust collectors and (c) junctions to connect the exhaust pipe. ................................... 115 Figure 94 - Dynamometric brake Scheme [92] . .............................................................................. 116 Figure 95 - Brake Dynamometer system. ........................................................................................ 117 Figure 96 - Dynamometer Brake Controller. .................................................................................... 117 Figure 97 - Dynamometric Brake Calibration Procedure. .................................................................. 118 Figure 98 - Calibration Sheet for the Brake Dynamometer. .............................................................. 119 Figure 99 - (a) Engine RPMs and Available Exhaust Thermal Power throughout the final test cycle and (b) Engine/dynamometer brake parameters. ........................................................................................ 120 Figure 100 - Exhaust inlet and outlet gas temperature and HE Effectiveness (Final Test). ................. 121 Figure 101 - a) Hot face temperature at each row of TEGs (top) and (b) Heat Pipe inner pressure Evolution. ...................................................................................................................................................... 122 xvi Figure 102 - (a) TEG Power Level by row and (b) Total Power vs Theoretical Max Power. .................. 123 Figure 103 - TEG efficiency during the final Test Cycle. .................................................................... 124 Figure 104 - TEG (a) voltage and (b) power as a function of temperature differential. ....................... 125 xvii LIST OF TABLES Table 1 -Working fluids and temperature ranges of heat pipes [61] .................................................... 39 Table 2 - Hi-Z 14HV TEG module datasheet [87] ............................................................................... 64 Table 3 - Engine Properties ............................................................................................................... 64 Table 4 - Pressure Sensors datasheet from manufacture ................................................................... 67 Table 5 - Mean Values of Test in Full Operation ................................................................................. 81 Table 6 - Simulation Results for Numerical Validation of the Experimental Test ................................... 89 Table 7 - Simulation results for Full Sized prototype ........................................................................... 93 Table 8 - Alternator and TEG output chain of conversion .................................................................... 95 Table 9 - Vehicle and Alternator Fuel Power savings using TEGs ........................................................ 95 Table 10 - Fuel Power Savings implementing TCTG ........................................................................... 95 Table 11 - Savings from using TCTG (Using TEGs GM250-12) [36] .................................................... 96 Table 12 - Mean Values of Test in Full Operation (New Prototype) .................................................... 125 Table 13 - Vehicle and Alternator Fuel Power savings using new TCTG ............................................. 127 1 LIST OF ACRONYMS AND ABBREVIATIONS 1D One-Dimensional 2D Two-Dimensional AFR Air-Fuel Ratio BEV Battery Electric Vehicle CC Constant Conductance CFD Computer Fluid Dynamics CLT Coolant Temperature Sensor CNG Compressed Natural Gas CO2 Carbon Dioxide EGR Exhaust Gas Recirculation emf Electromotive Force EMU Engine Management Unit GHG Greenhouse-Gas GUI Graphical User Interface HP Heat Pipes HW High Way HX Heat Exchanger IAT Intake Air Temperature ICE Internal Combustion Engines NCG Non-Condensable Gas ORC Organic Rankine Cycle RPM Revolutions Per Minute TCTG Temperature Controlled Thermoelectric Generator TE Thermoelectric 2 TEG Thermoelectric Generator TPS Throttle Position Sensor VC Variable Conductance VCHP Variable Conductance Heat Pipes VCHP Variable Conductance Heat Pipes VCTS Variable Conductance Thermosiphons WHR Waste Heat Recovery WLTC Worldwide Harmonized Light Vehicles Test Cycle ZT Thermoelectric Figure of Merit 3 LIST OF SYMBOLS Latin variables 𝐴 Surface Area m2 𝐶 Electric Conductance S 𝐶𝑝 Specific heat capacity at constant pressure J/kg·K D Diameter m 𝐸 Energy J ℎ Convection heat transfer coefficient W/m2·K 𝐼 Electric Current A 𝑙 Length m 𝑚󰇗 Mass flow rate kg/s 𝑁 Number of Pairs - N Engine Speed RPM 𝑃 Power W 𝑄󰇗 Thermal Power W 𝑄󰇗ℎ Thermal Power Absorbed by the TEG module W 𝑄󰇗𝑐 Thermal Power Released by the TEG module W 𝑅 Resistance Ω 𝑇 Temperature ºC or K T Torque N.m 𝑡 Time s U Internal Energy J 𝑉 Voltage V 𝑉󰇗 Volumetric Flow rate m3/s 4 𝑉𝑐 Cut Speed m/s 𝑊 Work J Greek variables ∆𝑥 Thickness m α Seebeck Coefficient V/K η Efficiency % λ Thermal Conductivity W/(m·K) π Peltier Coefficient J/C-1 ρ Electric Resistivity; Density Ω/m; kg/m3 Subscripts amb Ambient 𝑐 Cold-side ext Exterior ℎ Hot-side I Inlet 𝑖 Internal 𝐿 Load m Material max Maximum 𝑜 Outlet t Total vap Vaporization x Thickness 5 6 1. INTRODUCTION 1.1 Motivation The discovery in 2015 of an illegal exploit on 590 000 Volkswagen diesel engine vehicles led to the manufacturer acknowledging that more than 8 million passenger cars were equipped with a device that would cause the strong under-estimation of emissions to enable the vehicle to comply with the challenging emissions standards [1]. Following the event, several independent European organizations conducted emissions testing under the Euro 5 and Euro 6 in standard passenger cars. This event awakened public concern on whether road vehicles, which in recent years had shown a very promising reduction in greenhouse gas and pollutant emissions and fuel consumption [2], were in fact complying with these targets. Previous research has found that there was a gap between official type-approval and real-world CO2 emission. Some authors [3] stated that the gap has grown from less than 9% in 2001 to around 40% in 2015 (see Figure 1). In terms of fuel consumption, [4] when analysing real-world fuel consumption data from German car manufactures company cars (that represent almost half of new cars registrations in Europe [2]) found that the average divergence between on-road and official values increased from approximately 10% to 50% from 2004 to 2014. Figure 1-Divergence between official and real-world CO2 emissions for new passenger cars in the EU, the United States, China, and Japan [2][3]. 7 The above-mentioned data suggests that car manufacturers are struggling to keep up with targets for fuel consumption and emissions as they are forced to comply with very challenging sustainability goals. The research and development for upcoming years will be looking towards implementing technologies allowing to accomplish targets in energy consumption as well as lower pollutant and greenhouse-gas (GHG) emissions [5]. The EU has recently reviewed their GHG emissions emission cuts goal for 2030 increasing it to 55% in relatively to 1990 levels [6]. Regarding heavy duty vehicles, in 2025 and 2030 these need to achieve a 15% and a 30% reduction, respectively, when compared to 2019/20 levels [7]. These targets are very challenging. Full vehicle electrification is perceived as a very promising approach as GHG emissions during vehicle driving are eliminated. There has been a push by several policy makers to favour exclusively this type of transportation while sometimes pushing for the outright ban of the ICEbased vehicle in the long-term. A notable motivation behind these radical approaches seems to be the illperceived notions of BEVs (Battery Electric Vehicles) as being “emissions-free” vehicles and of ICE-based vehicles as being “emissions-bound” vehicles. Neither notion is accurate, denoting a lack of familiarity with life cycle analysis of emission footprint: The assessment of emissions footprint of BEVs is often reduced to the driving phase, with no acknowledgement to the fact that there are emerging technologies and fuels that can render the ICE an emissions-neutral machine [8], [9]. This reductionist approach risks eliminating valid solutions for the challenge of GHG emission reductions in the mediumand the longterm future. BEVs still have their own sustainability challenges, although these are not always present in policy discussion. The indirect emissions associated with electricity production and the Life Cycle impact at the beginning/end of life of components such as energy storage are issues that still need to be properly addressed [10]. Also, several technical challenges, mostly related to energy storage, still prevent their quicker and wider adoption. Still, light duty BEVs are being increasingly adopted, especially in urban transportation, where there are most advantageous [5], [11]. Their selling share increased from 3.5% in 2019 to 11% in 2020 in the EU [12]. Regarding heavy-duty long-distance transportation, it is expected that the rate of adoption of full electrified powertrains will be much slower, with several decades still needed to pass before mass adoption of full electrical powertrains takes place [5]. A lot of progress has also been made regarding fuels that are low on fossil emissions, such as compressed natural gas (CNG), or achieving a neutral carbon footprint, as the case of synthetic fuels made from renewable sources and waste-derived fuels, in which the CO2 emissions in the tailpipe are neutralized by the CO2 sequestrated during their sustainable production [13], [14]. While the use of ICE-based vehicles is expected to decline steadily over the next decades, they still account for most of the transportation sector nowadays [5]. To neglect the research effort towards solutions that 14 1.3 Thermal control through excess heat spreading In the previous design, thermal control was achieved by delivering all the exhaust heat to the TEG modules through phase change process which, for a given pressure is nearly isothermal. One alternative of performing thermal control with phase change is the one proposed by the authors in [36], which is outlined in Figure 3. It consists of performing the bulk of the heat transfer in a conventional way (absorption of the exhaust heat by convection at the HX, followed by transmission of this heat to the modules by conduction across the HX body). Only a small part of the heat transfer, if any, will be processed through the phase change process. It will only be present in regions where there is excess heat, that is, in regions where the temperature would otherwise exceed the limit if no phase change fluid would be present. Figure 3 - Schematic representation of the excess heat spreading (a) enabled using VCHPs and (b) heat transfer process across the VCHP with the excess heat being absorbed in the vaporization region and released in the condensation region. The idea is to embed chambers along the HX that contain the phase change fluid. Boiling will occur within these chambers wherever the boiling temperature is achieved. Similarly, to previous concepts, this temperature is regulated to be close to the maximum allowable operating temperature of the TEG modules. This is done, again, by regulating the pressure inside the system, using a non-condensable gas (NCG). This boiling will provide two advantages: locally, it will prevent further temperature increase above that limit temperature. But at the same time, the excess energy absorbed by boiling in hotter, upstream regions of the HX, will spread along the HX and will be used to heat up colder regions located downstream (a) (b) 15 in the HX by condensation. That is, the vapour generated at these excess heat regions will spread along the HX chambers and fill them until it condenses at under-heated regions of the HX, heating them up to the optimal temperature. Therefore, a similar effect to the previously described concept is obtained, but in a different way: Similarly to previous concepts, the heat absorbed by the HX will be distributed along an area of TEG modules proportional to the thermal load, with these modules displaying an optimal temperature level. However, in this new concept only the excess heat needs to be transported by the phase change fluid, not the whole heat, as in the case of the previous concept. Also, the system can now absorb a higher fraction of the exhaust heat and, unlike the previous system, exhaust heat will still be absorbed even when its temperature drops below the phase change temperature. Thus, this seems to be a solution able to passively avoid TEG overheating at high thermal load while efficiently using the available heat, because local excess heat is not wasted but spread to under-heated regions. At low thermal loads efficient operation will also be possible because it is possible to minimize the thermal resistance of the HX (maximize HX effectiveness) without the fear of overheating under high loads. This could be achieved using Heat Pipes (HPs) with adjustable boiling temperature, such as the Variable Conductance Heat Pipes (VCHP). A concept of a compact system based on VCHPs which does not requires gravity to operate has been designed and submitted for patenting [37]. One of the advantages of this system is that it can display a very low thermal resistance without the fear of overheating at high thermal loads. This solution allows maximizing the HX effectiveness for variable thermal load conditions. One design incorporating this concept was assessed in a publication of the group [36]. It incorporated corrugated pipes identical to those used in exhaust Gas Recirculation coolers from BorgWarner® that were embedded, along with the VCHPs in a cast aluminium body. The hot face of the modules was attached to the aluminium block, while the cold face of the modules was attached to liquid cooling plates. The concept was modelled and optimized geometrically and it can be seen in Figure 4. 16 (a) (b) Figure 4 - (a) global and (b) cut overview of temperature-controlled thermoelectric generator concept based on corrugated tubes and VCHPs [36] . The results were very promising even when discounting for the backpressure losses induced by the system as it was predicted to be capable of delivering as much as 572 W and 1538 W of average and maximum electric powers during a driving cycle, respectively, and showing a quite promising reduction of 5.4% in fuel consumption and CO2 emissions [36]. Some CFD simulations were included in the publication validating partly the model. However, a full experimental validation was still missing to confirm the excess heat spreading feature and in general the thermal control capability of the concept. The present work aims to address this issue. 17 1.4 Objectives of the Dissertation The main objective of this work is to perform the experimental and mathematical validation of the novel compact Temperature controlled Thermoelectric Generator (TCTG). It embodies corrugated pipes embedded in aluminium matrix working along with variable conductance heat pipes (VCHPs) working as excess heat spreaders for temperature control to be used in highly variable thermal load applications such as driving. This concept, that has been introduced in 1.3 an will be described in detail in 3.1, was patented recently [40]. An embodiment of his concept incorporating embedded corrugated pipes (exhaust flow) and VCHPs in a cast aluminium matrix, was assessed theoretically in recent publications by the LaMoTA team but had not been validated experimentally until now. Therefore, the present dissertation had as a main goal to perform the experimental validation of this concept and of the code that has been proposed to model it. For that, two similar downsized proof-ofconcept prototypes of the generator proposed in [36] have been configured and tested. A first one, that was built within a previous project, was instrumented and tested at gauge pressures close to atmospheric, yielding excess heat spreading temperatures around 100 ºC. A second one was designed and built within the dissertation and relied on stainless steel VCHPs able to withstand the pressures needed to achieve excess heat spreading temperatures around 250ºC. Both prototypes were suitably instrumented and tested after being attached common light duty 1.6L 4-cylinder spark ignition engine in an engine dynamometer. The experimental results were then used to confirm the principle of operation of the concepts conjectured in previous analyses and validate the mathematical models proposed in those analyses. 18 1.5 Structure This dissertation is divided into seven chapters. The first chapter “Introduction,” provides a general outline of the objectives and motivation of the work, as well as some previous research aligned with the already studied concept of excess heat spreading through phase change. Secondly, there is a chapter on the “Fundamental Concepts and Literature Survey” where the central concepts essential to understand the design of prototype tested are described. It includes the basics of thermoelectricity, thermodynamics, heat transfer fundamentals, heat pipes and their principles of operation, a state of the art of exhaust heat recovery systems and compact heat exchanger solutions. Chapter three “Previous Prototype and Modifications” assesses the prototype in which this work started, stating it´s concept of operation and how it was built, as well as the modifications that were made in order to perform the experimental tests. In chapter four, “Experimental Procedure on Existing Prototype,” the experimental system setup and its components are explained. In addition, the experimental test procedure is described. The results of the experimental test and the numerical validation are presented and analysed in chapter five “Results and Validation of the Existing Prototype”. Chapter six “New Prototype Construction and Set Up” presents the new design for a final TCTG prototype, explaining in detail the novel components as well as the construction of this new TCTG. Finally, chapter seven “Results and Validation of the New Prototype,” summarizes the results of the experimental tests performed to the new TCTG along with the main achievements and contributions of the dissertation to the field of study. 19 2. FUNDAMENTAL CONCEPTS AND LITERATURE SURVEY In this chapter the fundamentals concepts and literature survey necessary to understand the covered thematic are summarized. Starting with thermoelectricity, where de concept of Seebeck effect and thermoelectric modules is presented as well as one of its most prominent characteristics: temperature limitations. Followed by a brief summary on thermodynamics and heat transfer concepts. The present chapter also contains a detailed analysis on Heat Pipes technology and its functioning, ending with a state of the art of waste heat recovering thermoelectrical solutions. 2.1 Thermoelectricity Thermoelectricity is a phenomenon dealing with the direct conversion of thermal to electric energy, or the reverse [24]. In parallel with environmental concerns, the use of Thermoelectric technology is of interest for the automotive industry. Thermoelectric generators can transform the exhaust heat directly into electrical energy and therefore decrease the fuel consumption in vehicle. In terms of vehicle application, vibrations, accessibility and maintenance are obstacles to be taken into account. Therefore, the opportunity to design a robust, maintenance-free Thermoelectrical Generator with no moving parts aiming to recover exhaust heat seems a reliable solution. 2.1.1 The Seebeck, Peltier and Thomson Effect In 1821 Thomas Johann Seebeck discovered what is known today as The Seebeck Effect (SE). It is described when an electrical potential (voltage) is generated within any isolated conducting material that is subjected to a temperature difference across its junctions [24]. The ordinary example where this phenomenon is applied is to form a thermocouple composed of two dissimilar current-carrier junctions. The application of a temperature difference across the junctions of the pair of materials will produce a voltage that is approximately proportional to this difference. This Seebeck voltage, also known as thermoelectric electromotive force (emf) is defined as follows: 𝑉𝑜=𝛼∙∆𝑇 (1) The 𝑉𝑜 is the Seebeck voltage generated for the condition of an open circuit, ∆𝑇 is the temperature differential between the junctions. The magnitude of the emf generated is proportional to the temperature 20 difference, and the Seebeck Coefficient (α) of the specific material or pair of materials. The Seebeck Coefficient (α) corresponds to the gradient of the Seebeck Voltage relatively to temperature, as it follows: 𝛼 =[𝑑(𝑉𝑜) 𝑑𝑇 ]𝑇 (2) Later in 1983 Peltier observed that when an electrical current passes through two different conductors connected in a loop, one of the junctions between the conductors absorbed heat while the other one released heat [41]. The heat liberation or absorption depended on the direction of the current. Figure 5 displays a current I flowing through a junction formed between two different materials A and B held at a constant temperature T . Figure 5 - Current flowing through a junction between 2 materials [24]. The Electrical current will generate a heat fux and its magnitude and direction depends on the Peltier Coefficient (𝜋) of each material A and B . If the Peltier Coefficient are different, the thermal energy will leave the junction at a different rate that it is entering [41]. Thus, heat is absorbed or liberated at the junction. The Peltier effect (PE) can be defined as: 𝑞𝜋=𝜋∙𝐼 =𝛼∙𝑇∙𝐼 (3) 𝑞𝜋 is defined as the rate of heat liberation or absorption, 𝜋 is the Peltier coefficient and 𝐼 is the flowing current. It can also be seen from above that in reality, the Peltier coefficient is related to the Seebeck Coefficient, being the latter multiplied by the absolute temperature, T . In 1851, William Thomson postulated the existence of an additional reversible heat generation when a temperature differential is applied on a current carrying material [42]. Therefore, the Thomson heat it is 21 reversible and heat is either generated or liberated when the direction of the current or the temperature different is reversed. It is proportional to the current and the temperature gradient. It is defined as: 𝑞𝜇=𝜇∙𝐼∙∆𝑇 (4) 𝑞𝜇 is the rate of heat absorption per unit volume, 𝜇 is the Thomson coefficient to the applied current and temperature gradient, I is the current and ∆𝑇 is the temperature differential. The Thomson coefficient can also be obtained from the Seebeck Coefficient, as it is a function of its gradient with temperature. For a Seebeck coefficient that would be constant with temperature, a zero Thomson coefficient would result. In reality, the Thomson effect in a conductor results from the variation of the heat transport due to the variation of Peltier effect along the electricity path, due to the variation of temperature. If the capacity to transport heat by the charge carriers reduces along the current path, then heat must be released to the material along the way (Thomson heat source). On the contrary, if the capacity to transport heat by the charge carriers increases along the current path, then heat is absorbed by the charge carriers along the current path (Thomson heat sink). The above-mentioned effects are of practical importance since TE power generation results from the existence of a temperature gradient across a material attached to hot source and a heat sink. In power generation the Seebeck effect is the most important, but the Peltier and Thomson effects are also present due to the current induced by the Seebeck electromotive force. The Thomson effect can be neglected in most applications, but the Peltier effect will have an effect on the heat flow and on the effective temperature difference across the hot and cold junctions, deprecating it. The maximum electrical power generated by a Seebeck effect working device is obtained when the load resistance coincides with the internal resistance, 𝑅𝑖, of the pairs: 𝑃 =(𝑛𝛼∆𝑇)2/4𝑅𝑖 (5) The formal deduction of this equation is performed in 2.1.4. It is a function of the electrical resistance of the module 𝑅𝑖, Seebeck coefficient of the pair of materials and the number of element pairs 𝑛. 22 2.1.2 Thermoelectric Modules When aiming to generate electrical power using temperature differential, Thermoelectric Generators (TEGs) generators can be a possible solution. The TEG functioning principle translates into multiply the number of thermocouples to also multiply the voltage generated. The main components of a TEG are represented in Figure 6. The temperature differential between hot and cold face causes charge carriers of each element to flow from the hotter to the colder regions of the element. The particularity of these two different materials is that in the n-type materials the charge carriers are electrons, which flow from the hot to the cold junction. therefore, the conventional direction of the current (from positive to negative) will be from the cold to the hot junction. On the other hand, in the p-type materials the charge carriers are electron holes (in practice, positive charge carriers), which also flow from the hot to the cold junction. Therefore, the conventional direction of the current (from positive to negative) will be from the hot to the cold junction. Since the p-n pairs are arranged in zig-zag series through the electric connectors (see Figure 6), a current is created along the TEG. Numerous pairs are electrically connected in series and thermally in parallel. Each pair has two different elements, p-type with a positive α and a n-type with negative α [42] [43]. In Figure 7 it is possible to understand the working principle of the TEG. Figure 6 - Detailed thermoelectric module [93]. 23 Figure 7 - Seebeck operating principle [43] Under open circuit conditions, the voltage generated by the converter is proportional to the number of pn pair junctions multiplied by the difference of Seebeck coefficients between the p-type and n-type thermoelectric elements and the temperature difference between the faces as follows: (Note that 𝛼𝑛 is negative, so in practice, both Seebeck coefficients will be summed in module). 𝑉𝑜=𝑁∙(𝛼𝑝−𝛼𝑛)∙∆𝑇 (6) The module also has an internal resistance 𝑅𝑖. Therefore, it can be represented in a simplified way as consisting of a voltage source with voltage 𝑉𝑜 calculated according to eq. (6), and an internal resistance 𝑅𝑖. To produce power, the module must be connected to a load, that is, the device that will consume the electric power produced by the module. This load can also be represented as a load resistance, 𝑅𝐿. Figure 8 represents a module attached to a load resistance. A switch that can be turned on or off is also represented. Figure 8 - TEG module electrical circuit TEG Module 30 2.2 Thermodynamics Thermodynamics is given as the science that studies the relationship between of heat and other forms of energy, encompassing electrical, potential and kinetic energy and their respective transformations [53]. Engineering thermodynamics traditionally covers the study of applications as diverse as the processes of refrigeration and air conditioning, expansion and compression of fluids and vehicle engines, expanding nowadays to the production of renewable energy and the use of temperature differentials for selfconsumption of energy. 2.2.1 Basic Concepts It is possible to understand thermodynamic phenomena using two different methods. The first one, called classical thermodynamics, is carried out without referring to the nature of the individual particles which constitute the substance, nor to their interactions. It is a macroscopic approach to thermodynamic phenomena. The second method, called statistical thermodynamics, is based on the behaviour of groups of individual particles. The latter has proved to be of great importance in new methods of energy conversion, such as Thermoelectrics [54]. There are 5 laws that describe thermodynamic phenomena, the first and the second one being the most relevant for studying engineering solutions for energy conversion and use and the zero law being related to thermodynamic properties. These three laws will be addressed in the following subchapters. 2.2.2 Zero law of Thermodynamics Law zero of thermodynamics dictates that if a first body and a second body are in thermal equilibrium with a third body separately, then the first and second body are in thermal equilibrium with each other. This basic principle allowed temperature scales such as Celsius and Fahrenheit to be defined. 2.2.3 First law of Thermodynamics The first law of thermodynamics introduces the concept of conservation of energy, which leads to the concept of internal energy of a substance. According to the principle of conservation of energy, energy cannot be created or eliminated, only transformed. This principle makes the creation of a perpetual motion machine impossible. Thus a change in the energy of a body or system is equal to the difference between the final and initial instants [53]. It can be expressed: 31 𝐸𝐹𝑖𝑛𝑎𝑙 −𝐸𝐼𝑛𝑖𝑡𝑖𝑎𝑙 =∆𝐸 (23) In addition, this law establishes a relationship between work (𝑊) and heat (𝑄) exchanged between the system and its external environment. Given an isolated system in which a certain quantity of heat (𝑄) crosses the boundary and considering that the boundary moves, implying energy exchanged in the form of work, the variation of the internal energy ∆𝑈 is expressed as follows [54]: ∆𝑈=𝑄−𝑊 (24) It should be noted that if 𝑄 is positive the system is receiving heat and when 𝑄 is negative the system is giving off heat to the outside. The opposite happens with work: when 𝑊 is positive, it is performing work, so the energy is being supplied from the system to its surroundings. If 𝑊 is negative the reverse applies. 2.2.4 Second law of Thermodynamics In the analysis of real case studies, the first rule of Thermodynamics is not enough to determine the direction of processes as for example the heat transfer in which the heat always flows from the hottest body to the coldest. That is why it is necessary that real processes satisfy the second law of thermodynamics. Work can be more valuable than heat transfer when it comes to energy transfer. Work can be transformed 100% into heat whereas the opposite is not possible. The second law of thermodynamics also shows that the higher the temperature in the medium where the heat transfer takes place, the more heat transferred can be converted into useful work. It should be noted that systems with lower temperatures have "degraded" energy. This law also expresses that the entropy of an isolated system tends to increase with time until it reaches a maximum value [55]. That is, when a closed system interacts with another system, the energy is divided equally, until the system reaches thermal equilibrium [54]. Since no useful work can be extracted from a system in thermal equilibrium, a thermal machine can only function if there is a temperature difference. 32 2.2.5 Heat Transfer Concepts Heat transfer mainly studies energy in heat form since it is the energy that can be transferred from one system to another through a temperature differential. Conduction, convection and radiation (the last not being considered in this dissertation), are the three basic mechanisms of heat transfer. Both require a temperature difference and occur from the highest temperature level to the lowest. Heat transfer ceases when both systems reach the same temperature (thermal equilibrium). Conduction is the transfer of energy, which occurs by diffusion of the kinetic energy of particles within a medium. The particles atoms interact with each other but tend to remain roughly in the same physical position, except for gases, where molecular diffusion also plays a part in conductive heat transfer. The expression of conduction expresses as follows: 𝑄󰇗𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑜𝑛 =𝜆∙𝐴∙𝑇1−𝑇2 ∆𝑥 (25) 𝑄󰇗𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑜𝑛 − ℎ𝑒𝑎𝑡 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑟𝑎𝑡𝑒 [𝑊] 𝜆− 𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑣𝑖𝑡𝑦 [𝑊/𝑚⋅𝑘] 𝐴 − 𝑎𝑟𝑒𝑎 [𝑚2] ∆𝑥− 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 [𝑚] The heat conduction rate, 𝑄󰇗𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑜𝑛, across a plane wall with area, 𝐴 and thickness ∆𝑥, depends on the thermal conductivity of the material, 𝜆 [W/m⋅K], the transfer area, 𝐴 [m2], the temperature differential between two outer surfaces and the thickness, 𝑇1−𝑇2 ∆𝑥 (see Figure 12). Figure 12 - Heat conduction along a body with thickness ∆x and area A. 33 Convection is the heat transfer between a wall and a fluid flow. It occurs in liquids and gases and combines conduction and fluid motion. In this particular case, heat transfer occurs by conduction, that is, by microscopic interactions between contiguous molecules (molecular diffusion) but also by the macroscopic movement and mixing of the fluid that induces the transport of heat. The faster the movement of the fluid relatively to the wall, the greater will be the heat transfer by convection. The convective heat transfer rate is calculated through the convective heat transfer coefficient of the flow, h [W/m2 ⋅K] (which is normally obtained from empirical correlations suitable for the specific case), the heat transfer area, and in this case on the temperature difference between the bulk temperature of the flow and the wall, 𝑇 𝑓 and 𝑇 𝑤 . 𝑄 𝑐𝑜𝑛𝑣𝑒𝑐𝑡𝑖𝑜𝑛 = h∙𝐴 𝑐 ∙ (𝑇 𝑓 − 𝑇 𝑤 ) (26) Radiation heat transfer is related to the energy that is emitted by matter in the form of photons or electromagnetic waves. This heat transfer mechanism will not be further explored in this work. Thermal Inertia Thermal inertia is a property of materials, related to the ability of materials to maintain stored thermal energy over time. It can be defined as the property of a material that expresses the degree of slowness with which its temperature reaches that of the environment [56]. Thus, a material with high thermal inertia will take longer to reach an equilibrium state but is also able to store the accumulated thermal energy for a longer time. It can be defined as follows [57]: 𝐼 =√𝜆𝜌𝐶 (27) Where: 𝐼− 𝑇ℎ𝑒𝑟𝑚𝑎𝑙 𝐼𝑛𝑒𝑟𝑡𝑖𝑎 [𝐽 ∙ 𝑚−2 ∙𝐾−1 ∙𝑠−1/2] 𝜆 − 𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑖𝑣𝑖𝑡𝑦 [𝑊/𝑚∙ 𝐾] 𝜌 − 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 [𝑘𝑔∙𝑚−3] 𝑐− 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐 ℎ𝑒𝑎𝑡 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 [𝐽 ∙ 𝑘𝑔−1 ∙𝐾−1] 34 2.2.6 Heat Exchanger Basic Concepts Absorbed Thermal Power The general equation for describing thermal energy transfer in the absence of phase change is: 𝑄=𝑚∙𝑐𝑝∙Δ𝑇 (28) Where: 𝑄− thermal energy [J] 𝑚− mass of a substance [kg] 𝑐𝑝 − specific heat [J/kg∙K] Δ𝑇− temperature diferential (⁰C) It expresses the heat necessary to raise the temperature of a body of mass m by a certain Δ𝑇 over a period of time. It is a relevant parameter in transient heat transfer. In the present study, this is relevant in order to quantify the portion of the heat absorbed by the TCTG and not transmitted to the TEG modules. This expression can be adapted for energy flows. The heat transfer rate, 𝑄󰇗, from/to the exhaust/cooling stream to/from the TCTG will be proportional to the temperature decrease/increase between the inlet and the outlet. It is expressed in terms of Power [W]: 𝑄󰇗=𝑉󰇗∙ρ∙𝑐𝑝∙Δ𝑇 (29) Where: Q󰇗− heat transfer rate [W] V󰇗− volumetric flow rate [m3/s] ρ − density [kg/m3] cp − specific heat [J/kg∙K] ΔT − temperature diferential [0C] Heat Exchanger Effectiveness: The effectiveness of a heat exchanger is a concept that is slightly different from thermal efficiency. It represents the ratio between the heat effectively absorbed by the HX and the available heat, that is, the maximum heat that could be absorbed in an ideal scenario. From eq. (29) this yields: 35 ∈=𝑄󰇗𝑒𝑥ℎ 𝑎𝑏𝑠𝑜𝑟𝑏𝑒𝑑 𝑄󰇗exh available =𝑇𝑒𝑥ℎ 𝑖𝑛 −𝑇𝑒𝑥ℎ 𝑜𝑢𝑡 𝑇𝑒𝑥ℎ 𝑖𝑛 −𝑇𝑤𝑎𝑡𝑒𝑟 𝑖𝑛 (30) 36 2.3 Heat Pipes (HPs) and Variable Conductance Heat Pipes (VCHPs) As mentioned before, to produce as much energy output as possible TEGs need to maximize the temperature differential, but the modules should have a hot face with a temperature close to the maximum operating temperature, 250ºC. A solution for passively avoid TEG overheating at high thermal load while simultaneously providing efficient operation even at low loads could stand on a system capable of maintaining the desired operating temperature regardless of the exhaust gas regime, by automatically adjusting the heat flow reaching the TEGs. This could be achieved using Heat Pipes (HPs) with adjustable saturation temperature, such as the Variable Conductance Heat Pipes (VCHP). This solution allows constant phase change temperature (saturation temperature) which can be manipulated through pressure. This solution allows maximizing the thermal efficiency for variable thermal load conditions because the temperature will be prevented from going beyond the maximum allowable value. 2.3.1 Historical Development The principle of the heat pipe was first put forward in 1942 by Richard S. Gaugler of the General Motors Corporation in the U.S. Patent No. 2350348 [58]. The advantage of using heat pipe over other conventional methods is that large quantities of heat can be transported through a small cross-sectional area over a considerable distance with no additional power input to the system [59]. Gaugler designed a device that consisted of a closed tube in which a liquid would absorb heat and evaporate, travelling to other end where it would condensate releasing its latent heat and then travel back due to capillary pressure, restarting the cycle. The capillary phenomena seen in the HP working principle allows them to work on microgravitational environments with any external form. This led to its development being firstly aligned with space applications. Eventually energy cost problems and energy saving policies in Europe and Asian countries led to overwhelming development of HP technology in the last two decades. As today, there are HP applications all over the world, in the most diverse areas, from micro and miniature heat pipes, loop heat pipes over to pulsating heat pipes [59]. Likewise, a significant amount of based and applied research and development has been made by the scientific community with a large number of publications, reports and books being published since the 1990´s [60] [61]. 37 2.3.2 Principle of Operation Whereas HPs can be of any shape and size, the cylindrical geometry of the heat pipe is the one best suited to understand how they work. A schematic representation of a conventional cylindrical heat pipe is presented in Figure 13. The three main components of a heat pipe, as described in [59], are a sealed tube (container), a wick structure and a working fluid that is in equilibrium with its own steam. Figure 13 - Conventional heat pipe schematic view [61] The working fluid can vary, depending on the saturation temperature range of use to which they are designed sized, ranging from acetone, ammonia, and sodium, the most commonly used being water. The conventional heat pipe is essentially divided into three operating sections: the evaporator section, adiabatic (transport) section and condenser section. It should be noted that depending on the type and application of the heat pipe, the number of heat sources may vary as well as the existence or not of an adiabatic transport zone. The principle of operation of the HP is as follows: An external heat source is applied to one end of the pipe and heat is transferred by conduction through the pipe wall and the wick structure (metallic porous medium that absorbs liquids due to capillary forces similarly to a sponge) to the working fluid. The fluid vaporizes causing the pressure inside the pipe to drive it through the adiabatic section to the opposite end (condenser section). The different regions of the HP can be seen in Figure 14. 38 Figure 14 – Heat Pipe (a) sections and (b) fluid behaviour inside the pipe [61]. In the condenser section, due to a lower temperature on the outside related to the absence of a hot source, the steam condenses releasing the latent heat of condensation. To understand the phenomenon of fluid return through the Wick capillary it is necessary to understand that the menisci that translate the liquid-vapour boundary have a curved shape in the evaporator area and are flatter in the condenser area (see Figure 14). A capillary pressure resulting from the surface tension of the working fluid and the curved structure of the interface exists and varies in different sections of the heat pipe due to differences in the curved geometry of the menisci. This capillary pressure causes the fluid to circulate in the opposite direction to the pressure losses of the liquid and vapour and against the adverse forces of gravity. This process will continue if there is sufficient capillary pressure to move the working fluid back to the evaporator zone. Although gravity tend have a preponderant role in the mechanism, it is not strictly (a) (b) 39 necessary as, due to capillary action, heat pipes can operate in microgravitational fields without any external force field or pump. Also, horizontal applications can be found with significant efficiency [61]. 2.3.3 Working Fluid and Temperature Ranges The function of the working fluid within the heat pipe is to absorb the heat energy received at the evaporator section, transport it through the pipe and release this energy at the condenser end. Nevertheless, the HP only starts to transfer heat from the hot source to the heat sink once the boiling temperature of the fluid has been achieved (with the exception of the conventional heat pipe (see 2.3.4)) [27]. There is a range of fluids that can be used as condensable fluids and the choice will always rest on their specific application. A good working fluid will require a high thermal conductivity, high latent heat, and surface tension. The design of the HP must consider the temperature range in which it will work and its corresponding pressure range, specifying which fluid should be used, considering that it will be expected to vaporise and condense during the operation of the HP. Table 1 lists some of the vaporization and condensing temperatures of commonly used working fluids. Table 1 -Working fluids and temperature ranges of heat pipes [61] Working Fluid Melting Point @ ºC at 1atm Boiling Point @ ºC at 1atm Useful Range @ ºC at 1atm Helium -272 -268 -271 -269 Hydrogen -259 -268 -259 -242 Nitrogen -210 -246 -246 -236 Argon -189 -185 -203 -170 Oxygen -218 -218 -189 -157 Ammonia -77.6 -182 -200 -154 Ethanol -114 78 -60 100 Water 0 100 30 232 Dowtherm-A 12 253 30 277 Cesium 28 669 150 395 Lithium 180 1341 450 900 Silver 960 2211 1000 1800 46 (a) (b) In the following years, the thermoelectric generator was modified to adapt a Hybrid SUV to generate 180W of electricity and with the capacity to charge 300V batteries (Figure 21a). In 1998, the Nissan Research Centre from Japan [75] developed a TE generator with a rectangular crosssection of 72 modules (see Figure 20). Each one of these modules contained eight pairs of Si-Ge elements with a ZT of 0.6 to be applied in gasoline-powered vehicles. Cold water was used for the cold plates to provide a heat sink for the TEGs to generate electric power. The electrical power supplied by the generator was 35.6 𝑊 using a 3000 cc petrol engine under a 60 𝑘𝑚/ℎ in a hill climb. The power output ratio to the total volume of the TEG was 5.6 𝑊/𝑚3, and the ratio of the total electric power to area of the hot sides of all modules was 1.2 𝑘𝑊/𝑚2 [76][77]. Figure 20 - Nissan´s 1998 TEG Prototype [76]. Figure 19 – (a), (b) 1kw Thermoelectric generator installed in a 550hp diesel engine [71]. 47 Again in 1999, Hi-Z Technology in collaboration with the Clarkson University, General Motors and Delphi Corporation began the study to build and test a thermoelectric generator to generate 300W with a system capable of charging 12 and 24 V batteries to power the lights and other on-board devices on a GM Sierra Pickup with a V8 220 hp diesel engine [78][79]. The HE and TEGs are shown in Figure 21b and the results presented a power output of 255W (of the expect 300 W) with an average coolant temperature of 25ºC. (a) (b) In 2005 BMW and other organisations published at the DEER Conference (Directions in Engine-Efficiency and Emissions Research) the results of a study to implement a thermoelectric exhaust gas generator on a BMW 5 Series (petrol, Inline 6 cylinder, 3.0 L and 254 hp) capable of producing 500 W with a temperature difference between 250ºC and 390ªC [60][70][80] (see Figure 22). One of the goals of this project was to achieve a 10% improvement in fuel economy but no reports were found stating the final conclusions of the project. (a) (b) Figure 21 - a) 180 W thermoelectric generator for Hybrid SUV and b) 330 W for Jeep Sierra Pickup from General Motors [94]. Figure 22 – (a), (b) TEG Installed in a BMW 530i [82] [83]. 48 In 2008 Volkswagen presented the results of its first implementation of a heat exchanger with thermoelectric generators at the Thermoelektrik-Eine Chance Fur Die Atomobillindustrie with an output of about 600W, announcing a filling of the electric needs of the car by 30% and a saving of 5% in fuel through the load relief of the alternator [81]. In 2014 GMZ presented the results of the study of its 1 kW heat exchanger connected to a 15 L V8 diesel engine [82], [83]. This exchanger consists of a junction of five TEGs of 200 W each, incorporating GMZ's TG8-1.0 TE modules, capable of operating at temperatures of up to 600°C. Figure 23 - Exhaust TEG system presented by WV in 2008 [81]. Figure 24 - GMZ's 1KW Exhaust Heat TEG [74]. 49 Despite the constant innovations in the area of heat exchangers and materials of thermoelectric modules, the figure of merit of the materials, ZT, is still a barrier to overcome in order to enhance applications of this type of technology in the automotive industry. Some theoretical predictions were made and later experimentally validated [84] of the increased performance of thermoelectric modules using nanostructured materials. In summary, new prototypes of thermoelectric modules have been studied and have considerable advances in this area. However, the non-implementation of these technologies in vehicle exhaust gas recovery today can be blamed on the thermal stability that these materials require, design of heat exchangers with low effectiveness, production costs and development [24]. Currently one of the biggest constraints for these systems is the difficulty to achieve average efficiencies that may be closer to their theoretical maximum efficiency. This is due to the difficulty of achieving an optimal temperature level in the modules that operate under the highly variable thermal loads that are present in realistic driving cycles. Having a thermal control strategy enabling the modules to operate always near their optimal temperature (which is their maximum temperature) without the risk of overheating and with a high HX effectiveness could finally enable the viability of automotive TEGs. The concept assessed in the present study aims at these objectives. 50 3. EXISTING PROTOTYPE AND MODIFICATIONS For this work, a concept prototype modelled and numerically simulated by LaMoTA group was used as a base to build a downsized proof-of-concept prototype within Project Exhaust2Energy to be tested with an engine. The design and numerical assessment performed on this concept are described in [36]. This chapter describes the initial construction of the prototype presented in Figure 25, carried out by the LaMoTA research group, as well as the problems arising from this construction process that influence the types of tests performed and the finalization phase of this construction that was part of the work performed in the scope of this thesis. Figure 25 - TCTG Prototype concept. 3.1 TCTG Prototype Concept Based on the operating principles explained in chapter 2 regarding TEGs, excess heat spreading and VCHPs, a novel conceptual prototype design was proposed, as seen in Figure 26. Figure 26a displays the built and tested TEG generator concept design. The prototype tested is a partial version of the proposed final system, being one quarter of the full system (recall Figure 4). 51 (a) (b) Figure 26 - a) Full TEG generator concept; (b) Section View of the corrugated tubes and VCHPs embedded in an aluminum matrix, along with the TEGs (red) and cooling plates. Figure 26b shows a cross section view of the prototype in which it is possible to distinguish the corrugated tubes at the center, within which the exhaust gases flow. Two horizontally disposed batteries of VCHPs are positioned parallel to the exhaust flow, between the corrugated pipes and the upper and lower sets of TEG modules. These VCHPs are made of copper, incorporating an inner wick of porous copper aimed at promoting capillary pumping of the condensed fluid returning to the vaporization region. Both the corrugated pipes and the VCHPs are embedded in within a die-cast aluminum matrix that ensures a low thermal resistance heat path from the exhaust gases to the modules and allows structural stability during operation. The hot faces of the TEGs are in direct contact with the top and bottom surfaces of the cast aluminum matrix, while the cold faces of the TEGs are attached to the cooling plates, through which water flows as coolant in a controlled water circulating circuit. The CAD drawing represented in Figure 26a represents the size of the experimental prototype built and tested in the present study, that is presented later in this work. It comprises a total of 16 TEGs (8 below + 8 above). The working principle of this novel TCTG (recall Figure 3) is as follows: the HPs inner pressure is set to value that the condensable fluid only starts to vaporize on a controlled temperature set to optimize the system efficiency. In this case, the optimal temperature will be the TEG’s modules maximum working temperature (250ºC), in which case they only start to vaporize and transfer heat to next upstream thermoexchange zone whenever reaches the temperature limit of the modules. Then, when engine is running the exhaust gases dissipate heat while running through the corrugated pipes to transfer heat to the TEG’s hot face as desired. The heat dissipated by exhaust gases at an early stage, arrives by conduction to the TEG's, however when the point of saturation of the fluid inside the VCHP is reached (dependent to the pressure set initially), condensable fluid starts to evaporate and subsequent condensation allowing the Corrugated Tubes Heat Pipes TEGs Cooling Plates 52 phenomenon of heat dissipation along the exchanger. Based on this description, it is possible to identify 3 operating regimes: Low thermal power operation: the temperature is lower than the fluid saturation temperature inside the VCHPs, so they are inactive throughout their length. In this scenario heat travels by conduction through the aluminum matrix to the TEG's. Under these conditions, the VCHPs are filled only with the NCG and are approximately adiabatic. Average thermal power operation: there is a region in the vicinity of VCHPs where there is enough thermal power to raise the temperature above saturation conditions. This causes local fluid vaporization inside the VCHP, which absorbs heat from the aluminium matrix. A portion of the VCHP will thus be filled with vapour. This vapour will condense at the colder regions reached by vapour (recall Figure 3). However, there is not enough thermal power to fill the VCHP with vapour, so it works in a mixed state and VCHP are only partially active. In zones where it is active, there is still a component of the heat that is transmitted by conduction, but VCHPs absorb energy by vaporization (hotter regions) and release energy by condensation (colder regions). In the regions where it is inactive (with NCG only) energy is transmitted exclusively by conduction through of the aluminium matrix. High thermal power operation: the thermal power is such that the VCHP is filled by vapour along its entire length - the VCHP is fully active. In this case, the non-condensable gas is pushed out of the module region and into the expansion tanks. It is important to remember that maximum energy production is achieved when all TEG modules are active at maximum operating temperature, 250ºC. For this reason, the increase in thermal power source only causes to have more active modules and not increase the TEG's hot face temperature. Meaningfully for electric energy production in TEG's is the differential of temperature between the cold face and the hot face so it is vitally important to cool down effectively the opposite face of the TEG's. For this purpose, at the opposite side there are the cooling plates, through which a coolant (in this case water) flows driven by a pump. In a real-life application this fluid would have to be cooled through a radiator, and use a fluid such as antifreeze engine cooling fluid, to prevent system incapacitation during cold scenarios. Regarding the location of the TCTG, it should be placed directly after the catalyst aiming to minimize energy losses due to heat dissipation. Another interesting way to reduce heat dissipation losses would be the exhaust pipe insulation. Regarding the use of stainless-steel corrugated exhaust pipes (see Figure 27), these seem quite convenient as they are used extensively in exhaust applications such as Exhaust Gas Recirculation (EGR) coolers. They can be easily embedded within cast aluminium with good thermal continuity. Moreover, 53 their geometry is able to provide a good degree of mixing and boundary layer renewal while still providing a comparatively low pressure drop [85]. Figure 27 - Corrugated Pipes CAD Drawing. Besides the specific corrugated pipe geometry choice, the main variable to consider is the spacing between them. It should be as small as possible so that the largest volume of exhaust gases can flow with a sufficiently low pressure drop so that it would not negatively affect the engine performance. Another advantage of having a high number of corrugated tubes is the increase in the heat transfer area which maximizes exhaust heat absorption. A greater number of corrugated pipes also reduces the molten aluminum mass needed and therefore the overall weight of the generator. Of course, the configuration presented in Figure 26 has still not been optimized for this purpose. A lighter configuration would be possible to design. In the case of this specific prototype, due to the aluminum melting process the minimum allowable spacing was 2 mm. When working at their maximum temperature, the thermoelectric generator modules absorb their peak thermal power, so it is advantageous in these situations of high thermal power to have a by-pass system that offers a path of escape from excess exhaust gases, which will allow to reduce the back pressure phenomenon at the exhaust. This system is triggered whenever the TEG limit temperature is reached, and the exhaust gases are diverted so as not to damage the TEG modules. Unlike by-pass systems of existing concepts, this by-pass system does not waste otherwise useful heat because it will only activate once all the modules are operating at maximum power. 54 The condensable fluid used in VCHP's could be water, which boils at 250ºC (maximum allowable temperature of the modules) at a pressure of around 40 bar. Alternatively, another fluid that has a saturation point around the same temperature of a much milder pressure would be also attractive. Using such fluid would substantially reduce the pressure inside the VCHPs as compared to water, thus decreasing the required level of robustness, complexity and precision of construction. As an example, a condensable fluid such as DowTherm-A boils at 250ºC at a pressure of around 1 bar, making it an interesting alternative for this equipment (recall Table 1). However, as shown in [86], some experiments were made for maximum axial heat transfer. It can be stated that using water as working condensable fluid has a much greater potential than DowTherm-A as the heat transfer was double of that using the latter. As analysed throughout the work, TEGs are an essential component of the functioning of a TCTG, so the choice of TEG is crucial for the success of the equipment and will always depend on the type of vehicle operating conditions in which it will be implemented. For example, a TEG with a high thermal power transfer such as the Adaptive GM127 should perform better on driving cycles with abundant available exhaust power, such as the custom highway (HW) cycle. TEGs with a lower thermal power transfer such as the Adaptive GM49 TEG will be best suited in driving cycles where the available exhaust power is lower, such as the WLTC3 cycle (Worldwide harmonized Light vehicles Test Cycle), therefore more suitable for light cars used in urban environment. On this prototype, HZ-14HV TEG modules were used. Although their matched load electrical power output values are placed in between the values of the TEGs presented above (17 W against 15 W in GM49 and 28 W in GM127) and the value of the maximum temperature of the hot face of the modules is the same (250ºC) they have been chosen given their high robustness compared to the other modules, making it highly suitable for the type of use given the purpose of system implementation in a light duty vehicle. 55 3.2 TCTG Prototype Construction After obtaining satisfactory results from the numerical simulation of the prototype [36], the research group started building what would be a downsized model of the final prototype for the heat exchanger designed to be tested and thus validate the concept. In this phase, the casting components, which, due to their small dimensions, require precision, were planned and designed. Some casting attempts were made in order to perfect the technique and to understand how each material would behave with the aluminium casting at high temperatures. In Figure 28, one can see some of the attempts made by the LaMoTA group. It is possible to verify in Figure 28b that the aluminium has some solidification problems in colder areas of the moulding box, causing nonuniform distributions. The complete and uniform coating of aluminium throughout the prototype is important to ensure an even and planned heat transfer from the corrugated tubes to the TEGs that are placed in contact with the aluminium matrix. Figure 28 - a) HPs and corrugated pipes placement planning; (b) Corrugated pipes casting tests and (c) 3D printed Positioners prototypes. The prototype construction process started as follows: with the help of previously designed and tested positioners (displayed in Figure 29a and b), the corrugated pipes were aligned with the heat pipes and placed at a distance that allows them to be placed evenly inside the molding box. Figure 29 - (a) Corrugated pipes and (b) HPs arrangement before casting; (c) Corrugated Pipes and HPs inside de moulding box ready for casting. 62 4. EXPERIMENTAL PROCEDURE FOR THE TESTS OF THE EXISTING PROTOTYPE A test rig for the TCTG and an experimental procedure were configured to verify experimentally the excess heat spreading phenomena, that hitherto had only been hypothesized theoretically [39], and to evaluate the system performance when attached to a light vehicle engine. 4.1 Experimental System Setup To evaluate the energy conversion and electric power production, the TCTG was integrated into the experimental system. This experimental system consists of the TCTG, which is attached to the exhaust line of a light-duty spark ignition 1.6L engine connected to a brake dynamometer and to an open watercooling circuit. Electrically, it is connected to a data acquisition system (see Figure 37). Figure 37 - Schematic representation of the facility. 63 4.1.1 Heat Exchanger and TEG Modules The Heat Exchanger was integrated into the experimental system to evaluate the operating performance of energy conversion and electric power production. There are 16 TEG modules displaced and attached to the lower and upper surfaces of the HE, 8 on each surface, as shown in Figure 38. In this figure is possible to see the upper half surface of the HE with its corresponding 8 TEGs, numbered from 1 to 8 (one should recall that the lower half of the HE contains the remaining TEGs numbered from 9 to 16). Figure 38 - TCTG TEG modules upper-half. Since the TEG modules are connected in series in pairs of two (recall chapter 3.3.5), the nomenclature used throughout the work is presented as follows: -When referring to the upper half of the prototype, Row 1 consists of TEG 1 and TEG 2 (the first TEG row presented in the exhaust flow direction, see Figure 38), Row 2 consists of TEG 3 and TEG 4, Row 3 consists of TEG 5 and TEG 6 and Row 4 consists of TEG 7 and TEG 8. -When referring to the lower half of the prototype the same principle applies, being this time Row 1 the first two TEGs on the exhaust gas flow direction (TEG 9 and TEG 10), and Row 4 the last TEGs on the prototype (TEG 17 and TEG 18). The TEG modules used are the Hi-Z 14HV optimized for waste heat recovery and their manufacture datasheet properties are described in Table 2. Row 1 Row 2 Row 3 Row 4 64 Table 2 - Hi-Z 14HV TEG module datasheet [87] Thermal and Electrical Characteristic Parameter Conditions min typ max units Power Th=250ºC, Tc=50ºC 14 15.5 17.0 Watts Open Circuit Voltage Th=250ºC, Tc=50ºC 7.6 8.0 8.4 Volts Matched Load Voltage Th=250ºC, Tc=50ºC 3.8 4 4.2 Volts Internal Resistance Th=250ºC, Tc=50ºC 0.9 1.0 1.1 Ω T=50ºC 0.55 0.6 0.65 Ω Heat Flux Th=250ºC, Tc=50ºC (matched load) 410 430 450 Watts Th=250ºC, Tc=50ºC (open circuit) 200 210 220 Watts 4.1.2 Engine The engine used in the experimental validation of the concept is a light duty spark ignition engine (presented in Figure 39) and his specifications are displayed in Table 3. The engines that were available in the Lab was this one and also a Diesel one (1.6 HDI from PSA group). This one was chosen because petrol engines display a higher exhaust temperature on average and thus display a higher recovery potential than Diesels. Table 3 - Engine Properties Engine Properties Engine Spec. TU5JP4 from PSA Engine Capacity 1587 𝑐𝑚3 atmospheric Cylinders 4 Volumetric compression ratio 11:1 Maximum Power 88 kW @6000 rpm Maximum Torque 160 N.m @4250 rpm Injection System Multipoint Bosch Spark ECU ECUMaster® EMU Fuel RON95 Petrol Figure 39 - TU5JP4 Engine. 65 4.1.3 Cooling water system The cooling system for the cold sides of the TEGs is composed of water collectors designed to distribute the flow along the several cooling plates, the piping system that connects the assembly and a rotameter water flow meter (see Figure 40). The detailed explanation of most of the water supply system components is outlined in chapter 3.2 and 3.3 with the exception of water flow meter. It was connected to the water supply system of the prototype and the water circuit of the lab. The flowmeter measurements [l/m] were used in thermal absorbed power calculations to evaluate the performance of the cooling plates. Figure 40 - Water Flow meter. 4.1.4 Data Acquisition system To assess the TCTG performance, a range of data acquisition techniques aimed at describing the heat exchanger and electric power production capabilities was used (see Figure 41a). The CompactDAQ C Series Multifunction modular system from NI-National Instruments that includes analog input, analog output, and 5 V TTL digital I/O channels was used (see Figure 41b). For the temperature readings two NI-9214 were used, for voltages a NI-9205 D-SUB and for pressures a NI-9201. 66 Figure 41 - (a) Data acquisition system table and (b) NI CompactDAQ modular system in blue. 4.1.5 Pressure In this experimental system two types of pressure measurements were deployed. The first one was the pressure measurement inside the HP system, which allowed the indirect monitoring and control of the vaporization temperature of the condensable fluid. The second one measured the differential pressure drop of the cooling water flow between the entrance and exit of the cooling plates. The pressure drop of water flow on the cold-side cooling system was measured using a JUMO® MIDAS differential pressure sensor. This pressure transmitter acquires the relative pressure between two points of liquid or gaseous media and transforms it into an electrical signal. The two differential pressure acquisition points are highlighted in Figure 42a in the red arrows. The HP pressure sensor consists of a GEMS® 3500 series high pressure sensor which aims to measure the evolution of the relative pressure inside the system. This measurement device is presented in Figure 42b. The measurement parameters of each sensor are displayed in Table 4. (a) (b) Figure 42 - (a) Jumo Midas® differential pressure sensor; (b) GEMS® 3500 pressure sensor. 67 Table 4 - Pressure Sensors datasheet from manufacture Pressure Sensors GEMS® General technical data Input Max. Overload Bursting Pressure Current Output Temperature range Overall Error 0.35 +40 bar 80 bar 120 bar 4 to 20 mA 0ºC/+ 120ºC 0,0025*P and long-term drift of 0,2% JUMO® MIDAS General technical data -1 +5 bar relative pressure 7.5 bar 60 bar 4 to 20 mA two-wire; DC supply 0 to 10V three-wire -20ºC / + 125 ºC 0,0025*P 4.1.6 Temperatures To measure the different temperatures of the system, 2 types of thermocouples were installed at the aforementioned points. The first ones are K-type thermocouples 1.5mm x 150 mm and the second ones are K-type 0.2 mm x 1 m thermocouple. According to the manufacturer datasheet [88], these thermocouples are capable of recording temperatures from a range of -40ºC to 1100ºC and they are in compliance with IEC584 Standards approval, which states the measurement error for K-type thermocouples is 2.5ºC or 0.75% of the measurement temperature. The temperatures measured at the HX were the hot and cold face temperatures of the TEGs and the inlet and outlet exhaust temperatures. The water temperature of the cooling system was also measured at the inlet and the outlet of the cooling plates. This measurement was made at the junction of all the water tubes coming to and from the cooling plates. 4.1.7 Voltage The main voltages measured during the test were the ones generated by the modules due to the Seebeck effect (recall chapter 2.1.1). This voltage is then dissipated by the load resistances attached to them. The specific load resistances used in the prototype are described in chapter 3.3.4. This voltage is measured by the NI card, recall subchapter 4.1.4 . Other voltages were measured, such as those from the pressure transducers and thermocouples. 68 4.1.8 LabVIEW® The software LabVIEW® was used to control the data acquisition cards, provide a graphical user interface (GUI), and allow the programming of the channels to be acquired and displayed in real time in the computer screen. A custom LabVIEW® program was previously developed to record all main parameters. At the end of a test the software would generate a text file containing the values of the measured parameters such as temperatures, pressures, and TEG voltages. These parameters were then processed and displayed graphically for the later analysis of system performance. Figure 43 - LabView® software used to monitor and record TCTG data acquisition parameters during the test cycle. 4.1.9 ECUMaster® To control and record some engine operating parameters, a programmable engine control unit (ECU) was used (see Figure 44a). The engine control hardware and software are from ECU Master®, model EMU® (Engine Management Unit). This allowed changes in parameters such as ignition advance and air-fuel ratio. It also allowed real-time monitoring and storage of engine operating data during the performed tests (see Figure 44b). After the test, the software would generate a .csv file containing the records of predefined parameters. These parameters were the engine RPM, TPS (Throttle Position Sensor), IAT (Intake Air Temperature), CLT (Coolant Temperature), Fuel Usage and Lambda. These parameters were used in the performance study of the prototype. 69 4.1.10 MATLAB® A MATLAB® code was developed to analyse the bulk data acquired during the tests, and to perform calculations related to the performance of the system. This code would read the two files produced by the NI data acquisition system and the ECU software mentioned. Then it would perform some calculations to obtain the performance parameters in numerical and graphical form. These included the instantaneous or time-averaged values for the exhaust mass flow rate, thermal power absorbed by the TCTG and released to the cooling system, real and estimated matched load power produced by the TEGs. Some of the calculations performed by the code to obtain the TEG power output and efficiency, TEG´s maximum theoretical power, absorbed thermal power and Heat Exchanger Effectiveness are presented in subchapters 2.1 and 2.2. a) b) Figure 44 - a) Engine Management Unit and b) ECUMaster software used to change and record engine parameters during the test cycle. 70 4.2 Test Procedure All sensors such as thermocouples, voltage and pressure sensors were connected to the data acquisition system which was bridged to a computer running LabView®. The ECU was connected to a second computer running the ECUMaster® data acquisition and engine control software (recall Figure 37). The TCTG heat pipes were previously filled with water (60ml in each HP system). As a result of the casting process both HP batteries had pressure leaks, making it impossible to pressurize the system to achieve the intended boiling temperature of 250ºC. Nevertheless, it was deemed that performing the tests at ambient pressure, and thus with a corresponding saturation temperature around 100ºC, would still be insightful, as it would still allow to observe the phenomena of temperature control and excess heat spreading and thus validate the concept. The engine operating conditions were defined based on preliminary tests. It was decided that the engine would work near 3000 rpm and with no load (no braking torque in the dynamometer). This was indeed sufficient to produce enough heat to observe excess heat spreading phenomena and achieve full TCTG operation. The volumetric water flow rate used was 10.3 L/min. The ambient temperature in the lab was roughly 23ºC. Figure 45 - Ready to test prototype. 71 In this experimental procedure, the following steps were followed during each test 1. Turn on the data acquisition systems; 2. Start the recording of values from LabView and ECU acquisition system simultaneously (facilitating synchronization for later analysis); 3. Start the combustion engine and begin engine warm up conditions (2400 rpm for 100 seconds); 4. Turn on the water flow of the cooling plates, and perform leakage verification; 5. Measure water flow rate; 6. After 100 seconds, change Engine speed to 3000 rpm; 7. Gas and water leakage verification under test-conditions; 8. Thermal image acquisition after system thermal stabilization; 9. After 700 seconds from test beginning, reduce engine speed to idling conditions; 10. Turn off combustion engine; 11. Turn off water flow of the cooling plates; 12. Stop and save the LabView and ECU data, performing this simultaneously; 13. Export recorded data acquisition for further treatment in MATLAB and later analysis. 78 ensures that the temperature differential between the faces of the TEGs increase throughout the cycle which will be directly related to the electrical output of the HE. 5.1.4 Electrical output and efficiency As the primary goal of this prototype consists on the conversion from thermal to electrical power, the voltage and power produced by the TEGs are among the main indicators to analyse the system performance. Based on eq. (1) the voltage is expected to be proportional to the temperature differential obtained for each TEG. It is possible to verify that this premise is true, and that the curve of the voltage graph seen in Figure 50a, obtained by direct measurement, evolves similarly to the temperature curve (seen in Figure 49). The limited number of voltage input channels available in the data acquisition system led to the choice of connecting in series the load resistances of the left and right modules of each row. Therefore, the voltages measured and the corresponding obtained powers correspond to the pairs of modules belonging to each row (recall outline of Figure 38). The electrical power generated is a function of voltage and of load resistance (recall eq. (14)), thus it is possible obtain the electrical power curve through the voltage curve, as seen in Figure 50. (a) (b) Figure 50 - a) Row 1 TEG Voltage Output and b) Row 1 TEG Power Output. To obtain maximum power the load resistance should be equal to the internal resistance of the modules (matched load conditions). The load resistance used was very close to the internal resistance reported 79 for the modules at ambient temperature. Nevertheless, a correction for estimating the theoretical matched load power was also used (eq. (18)). The differences obtained were negligible. The electrical power output of each module increases with the square of the voltage (eq. (14)), so it is expected to see a sharper increase than the voltage. In terms of electrical output, it is noteworthy the functioning of the first TEG rows in comparison to the last ones (see Figure 51). The closer the Row is to the exhaust gas inlet, the greater the power generated. This results from the heat absorption from the exhaust gases throughout the system. Thus, the Row 4 is the one that receives the least heat as much of it has already been absorbed by previous rows. By consequence it is the TEG row that produces less electrical power. In Figure 51 it is also possible to see the system total electrical power production which represents the sum of the electrical power output of the TEG modules. The maximum value achieved during this test was around 90W. Being the steady state point where the heat pipes are filled with steam and the TEG’s hot face is the same. For a setpoint of 100ºC, if an excess condenser or a by-pass valve were used, the system should stabilize around 350 to 400 seconds, meaning a maximum power output of around 60W. If it was possible to reach 250ºC the maximum output should be considerably greater as power increase with increases with the square of the temperature (recall eq.(14)). Figure 51 - TEGs Power by Row and TEGs Total Power Output. 80 The electrical efficiency expresses a ratio between the electric power generated and the thermal power absorbed by the system and by the TEGs (recall eq. (19)). It is as important as the electrical power output, especially in limited thermal power events, such as urban driving with cold engine. The efficiency of the system tested was studied (see Figure 52). Figure 52 - TEGs efficiency during Test Cycle. As the cycle progresses, the electrical efficiency increases as a result of the increase in the temperature differential between the TEG’s faces throughout the cycle, which is directly related to their efficiency. The maximum efficiency recorded was 3.5%., this efficiency is at the upper spectrum of what could be expected from the Hi-Z 14HV TEG. This is surprising because during the cycle under-study, the maximum operating temperature of the TEGs (∼120ºC) was far from the maximum operating temperature of 250ºC, where they are most efficient (∼5%). This indicates that factors that normally reduce efficiency, such as thermal contact resistances, here have been minimized (thermal grease was used). The analyses performed were able to calculate some mean values for a given steady state of choice. For this test, an interval from 300 seconds to 400 seconds was chosen given to be moment when all the HP’s system was full and the Heat Exchanger was in full operation. Table 5 shows some of the calculated mean values. 81 Table 5 - Mean Values of Test in Full Operation Variable Mean Value Units EGTin 420 [ºC] EGTout 75 [ºC] Cooling water Pressure drop 40 [mbar] Exhaust mass flow rate 41.1 [g/s] Fuel consumption 7.44 [l/h] Power absorbed by exhaust heat exchanger 15.6 [kW] 5.1.5 Voltage and Power as a function of Temperature TEG's energy production is related to the temperature differential between their hot and cold faces. It is therefore interesting to establish a relationship between the electrical power produced and the temperature differential of the TEGs. Figure 53 displays the Voltage and Power of the TEG´s output for each temperature differential registered throughout the test cycle. (a) (b) Figure 53 - TEG a) voltage and b) power as a function of temperature differential. Under steady state operation, for the same temperature difference the TEGs should all produce the same power output, however, this was not the completely the case. Small differences are visible depending on TEG position at the HX. These results were obtained under transient regime, so that thermal equilibrium has not yet been reached, and there are differences due to thermal inertia. This may mean that the temperature measured by the thermocouples might not be fully representative of the average temperature 82 of the hot faces. The greater the temperature difference, the greater the power generated (as already been mentioned on the functioning of the TEGs). The graphics of Figure 53 present some undesirable noise, as well as anomalous behaviours that can be related with the change of the flow regime of the exhaust. For example, when the engine is switched off but the temperature in the prototype remains. A comparison was made between the performance of the TEG in the experimental test with the performance data provided by the manufacturer. This information data had to be corrected, as the internal resistance of the module was actually substantially lower than that stated in the datasheet. While this does not affect the voltage (matched load voltage is half of the open circuit voltage, which in turn is solely a function of the temperature differential and the Seebeck effect according to eq. (1)), it affects the power (recall eq.(14)). In Figure 53a and Figure 53b it is possible to see the green line that represents the performance of the TEGs based on the corrected manufacturer data. In conclusion, it can be stated that the performance of some modules (more specifically of the ones from Row 1 and 2) is higher than that announced by the manufacturer, especially when the temperature difference between the TEG’s hot and cold face is greater than 60ºC. The difference seems to be mainly in the internal resistance of the modules, which seems to have been improved over the datasheet information. This yields a similar open circuit voltage but a much higher matched load power output than reported in the datasheet. It is worth noting that the lower internal resistance of the modules purchased was actually provided by the manufacturer when supplying the modules. 83 5.2 Numerical Model Validation The experimental results obtained allow the validation of the numerical approach that has been previously used by the group to model the heat spreader concept that was proposed earlier [39]. This subchapter outlines the mathematical model used, presents the simulation results when using the same exhaust input as in the experimental results and compares the simulation results against the experimental results. 5.2.1 Outline of The Model The LaMoTA research group developed a mathematical model to predict the system behaviour. The main goal of this approach was to predict the electric power output of this prototype during predefined driving cycles. The model was described in [34], while the electrical output of the TEGs, as well as its impact on fuel economy and GHG emissions were assessed under the Worldwide Harmonized Light Vehicle Test Cycle class 3 (WLTC) and a custom highway (HW) cycle. In the present work, a custom cycle which mimics the experimental test conditions (inlet exhaust thermal power as seen in Figure 46 and inlet exhaust temperature according to Figure 47) was simulated as it is described afterwards. The model’s engine and driving cycle energy analysis was developed in earlier work [34] [89]. It predicts the instantaneous engine map position such as speed and torque to complete the driving cycle and the corresponding instantaneous exhaust mas flow rate and inlet temperature. The heat transfer and thermoelectric model was originally proposed in [38]. It is a model with discretization in the longitudinal direction into 72 sections in which 1D thermal calculations are made, and where the gas/coolant outlet temperature of each section is the inlet temperature of the next section, thus guiding the thermal simulation calculations of the system as a whole. In each of these sections, heat is transferred from the exhaust gases (corrugated pipes) to the coolant (cooling plates): - Heat transfer calculations are performed for each section, namely, the thermal resistances between the exhaust flow, the HX body (including HPs if active), the modules and finally the cooling flow. For the convective heat transfer calculations in each section empirical correlations for the convection at the exhaust flow and coolant flow are used. For the heat flow across the HX, 1D conduction equivalent thermal resistances were derived from 2D numerical heat transfer calculations done with commercial packages, as seen in Figure 54b. - Two different heat transfer calculations are made in each section depending on whether there is excess heat spreading condition (active HPs) or not. 84 - In each section heat fluxes and temperatures are calculated. vapour production, accumulation and depletion are also calculated in case the HPs are active. - Once the calculations in one section are completed, energy balances are made to calculate the temperature of the exhaust gases and water at the entry of the next section. Also, depending on the vapour produced, transported or depleted on the preceding section, the vapour accumulated for the next section is obtained. - Simplified thermoelectric calculations are made based on the resulting temperatures at the hot and cold faces of the modules and the manufacturer datasheet [38] The average Peltier effect is incorporated into an effective thermal conductivity that already accounts for this effect. All the physical properties of the materials, fluids and components were extracted from the manufacturer datasheets aiming to confirm that the electrical output of the system corresponds to the prototype built in real world. For simplification reasons, conduction transfer along the length of the HX was not considered. In the built physical prototype, which was described previously, there is a discontinuity/separation of the aluminium matrix for each 2 TEG module rows, so that this longitudinal heat transfer in the real prototype is limited. Under these conditions, this simplification seems acceptable. The thermal power flowing from the corrugated tubes to the cooling water on each of the 72 sections depends on the heat pipe thermal resistance which varies depending on whether the HP are active or not. For the calculation of the 1D equivalent thermal resistance, either an adiabatic or isothermal wall boundary 2D condition solver was used to represent the conditions where the HPs were inactive or active, respectively, as seen in Figure 54b and Figure 55. Figure 54 - (a) CFD simulation of a corrugated pipe; (b) 2D heat conduction simulation of the exhaust heat exchanger [39]. The prototype studied in the present work (depicted in Figure 26) represents one fourth of the total system length of this model. This means that in the input and output of the modelling and simulation process, some adaptations to the model had to be considered. (a) (b) 85 Figure 55 shows an example of the geometry used in the 2D calculations to extract the 1D thermal resistances of each one of the 72 sections of the HX. One can see the main block which contains the corrugated pipes embedded with aluminium and the HP region divided into upstream and downstream region, R1 and R2, respectively. By performing 2D simulations of the heaty transfer occurring in these complex shapes for a given temperature differential, it was possible to extract an equivalent 1D thermal resistance that would be valid for other operating conditions. In this way, it is possible to adapt this complex shape to a 1D analysis. These thermal resistances are different whether HPs areas active or not. An adiabatic HP surface corresponds to the situation where there is no phase change occurring (temperature has not exceeded the HP saturation temperature). The situation with active HPs corresponds to isothermal HP surface (temperature has achieved saturation temperature). The detailed outline of this model, as well as the condition solver simplifications were presented in [38]. Figure 55 - 1D Thermal resistance scheme extracted from the 2D heat transfer simulation of a section of the HX under active and inactive HP condition. The numerical model requires two main input parameters: the exhaust mass flow rate or power (Figure 46) and the exhaust gas inlet temperature (Figure 47) as a function of time. Any driving cycle, be it more variable or more stable can be programmed as long as the aforementioned values are extracted for each time step (in this case every second) [27][34], [89]. In the present case, these values were extracted from the engine ECU file (mass flow rate data) and from the data acquisition system (exhaust temperature data). Adiabatic (inactive HPs) or Isothermal (active HPs) Thot Tcold Adiabatic (inactive HPs) or Isothermal (active HPs) Tcold 86 5.2.2 Numerical model correction for heat capacity It must be noted that the model is a quasi-steady state model, meaning that it does not incorporate the influence of the heat capacity of the solid and liquid masses. What happened in the tests was that under transient operation (heating over time) a fraction of the heat absorbed by the HX was stored in it due to its heat capacity and was not transmitted to the modules. In order to overcome this model limitation, a correction of the inlet exhaust power was performed so that the heat stored in the heat exchanger and the working fluid through heat capacity was discounted from the heat absorbed by the HX. This correction brings the mathematical model closer to reality. The heat required to heat the heat exchanger from ambient temperature to the vaporization temperature of the working fluid (100 ºC) was calculated. For this calculation the mass of the aluminium matrix and the working fluid (water) were used as follows: 𝑄𝐻𝑒𝑎𝑡 =(𝑚𝑎𝑙 ∙𝑐𝑝𝑎𝑙 +𝑚𝑤𝑎𝑡𝑒𝑟 ∙𝑐𝑝𝑤𝑎𝑡𝑒𝑟)∙(𝑇𝑣𝑎𝑝 −𝑇𝑎𝑚𝑏) (34) After calculating the heat value that is necessary to heat the HE from ambient temperature to 100 ºC (working fluid vaporization temperature), the next step was to calculate the average heating power. In this way it was noticed that, in the results of the experimental tests, the system needed in average about t=200 seconds to reach the vaporization temperature in the hot faces of the modules. So, for the same engine parameters, the heating power can be calculated as follows: 𝑃𝐻𝑒𝑎𝑡 =𝑄𝐻𝑒𝑎𝑡 𝑡 (35) The resulting power due to heat capacity was subtracted at each instant from the exhaust gas power in order to obtain a corrected exhaust power. A corrected exhaust gas temperature for each time step was derived and used in the heat transfer calculations. With this model adaptation, it was possible to incorporate the influence of thermal inertia in a simplified way and thus enable a valid comparison between theory and experiments. 5.2.3 Numerical Validation Results Figure 56a and Figure 56b displays the comparison between the experimental and theoretical values of the electrical power output as a function of time, as in total electrical output and electrical output for each 87 TEG Row. The maximum value of total electrical output predicted is reached at around 260 s (see Figure Figure 53a). By comparison, the full operation of the prototype, identified where all four rows of modules achieve a plateau was attained after 300 s (see Figure 53b). Given the simplifications done, namely regarding the heat capacity, the fact that on the experimental results the prototype is on transient regime, the comparison seems to be acceptable. It has already been said that the reason why the value of the total experimental power then continues to rise has to do with the fact that the existing prototype still does not incorporate an excess vapour condenser and therefore, once the vapour becomes excessive the pressure continues to build-up and so the saturation temperature also continues to grow, inducing a higher excess heat spreading temperature and thus a higher power. However, it may be seen that the maximum total power predicted, around 64 W, is similar to the value achieved experimentally slightly after, at around 400 s. (a) (b) Figure 56 - Electrical Power output (a) experimental vs (b) predicted for the downsized proof-of-concept TEG operating with an excess heat spreading temperature of 100ºC. Although in the experimental results more time to reach full operation was needed, the heat spreading effect was confirmed for the same TEG Row electrical power values. The results of the electrical power output as well as the efficiency organized as a function of the available exhaust energy or the heat absorbed by TEGs can be seen in Figure 57. 0 10 20 30 40 50 60 70 80 90 100 0 100 200 300 400 500 600 700 800 Electrical Power Output [W] Time [s] TEGs Predicted Power Electrical Output [W] 1st TEG row 2nd TEG row 3rd TEG row 4th TEG row 94 5.4 Fuel and CO2 Savings The increasing need for onboard electricity to supply modern vehicle transportation components, such as the air conditioner, digital displays, headlights and tail lights, cockpit ambient lights, sound system are accountable for decreasing the fuel efficiency. This phenomenon results from the fact that onboard electricity comes as a direct result of fuel being consumed within the engine to feed an energy conversion chain [90]. This chain starts with chemical energy stored in the fuel and ends with electrical energy from the alternator. Along the system functioning power losses happen and are associated with every energy conversion process. Even focusing only on the alternator there are several associated losses, such as mechanical losses, electrical losses and magnetic losses. [90] explained that although alternator efficiency varies with speed, load (amps) and electrical power output, commercially available alternator’s efficiency range between 50 to 60%. This combined with conversion chain and ICE’s efficiency make onboard electricity a significant impact on vehicle fuel consumption. 5.4.1 Compact TCTG Implementation Although the average electrical power produced by this concept seems to have a relatively low value for light duty vehicles it is important to remember that this is a proof of concept system, with a quarter of the size of the final system and where it was only possible to reach less than half of the hot face temperature of the modules, staging them in underproduction. Using the results obtained in the experimental tests , the fuel consumption and CO2 emission savings were calculated. In these calculations, the engine power loss due to the back-pressure imposed by the exhaust heat exchanger was not considered. For this, the bench test ICE’s stock alternator specifications were used. All the calculations presented afterwards were made by creating an Excel spreadsheet for further prototype studies (see Annex 2). First, with the manufacture’s datasheet values for voltage (V) and current (A) the alternator´s power consumption (full-load) was calculated (see Table 8). Then using a mean value for alternator efficiency stated by [90] (55%) and the petrol engine average efficiency of 18% regarding to the average engine efficiency data from the WLTC Driving cycle presented in a recent publication [36], the mechanical power and the fuel power consumption were calculated. It is notable that to produce 1.4 kW of electricity the engine must spent 14.45 kW of fuel power. Also, using the same chain of conversion note that to produce the TEG output (64W) in the alternator it would be necessary to use 646 W of fuel power. 95 Table 8 - Alternator and TEG output chain of conversion Parameter Value Unit Alternator Power 1,44 kW Alternator Mechanical Power (η=55%) 2.62 kW Alternator Fuel Power (η=18%) 14,45 kW TEG Output 64 W TEG Fuel Power (if produced by the Alternator) 646 W Calculations of the fuel consumption were made using the lower heating value of fuel used (Petrol RON95). The vehicle’s mean fuel power during the test performed was calculated using the exhaust mass flow rate of the engine (whose value was previously estimated using MATLAB®). Then it was possible to study the fuel consumption of either the vehicle and the fuel savings of the alternator if the energy was produced by the TCTG (see Table 9). Table 9 - Vehicle and Alternator Fuel Power savings using TEGs Parameter Value Unit Vehicle Fuel Power 90 kW Vehicle Total Fuel Power Consumption 2 g/s Alternator Total Fuel Power Consumption 0.32 g/s Alternator Fuel saving Using TEGs 0.014 g/s Alternator Fuel saving Using TEGs 0.039 l/h Table 10 presents the fundamental savings obtained by applying the TEG system assuming the mentioned alternator with 55% efficiency for the test conditions performed. Table 10 - Fuel Power Savings implementing TCTG Savings Value Unit Alternator Fuel Power Savings using TEGs 4,44% % Vehicle Fuel Power Saving using TEGS 0.72 % 96 In respect to the implementation of this proof-of-concept prototype, the electrical production does not completely replace the alternator specially. This happens because the electrical output of the system is low compared to the alternator full load power, since the maximum achieved TEG module hot face temperature was 100 ºC, less than half of the maximum allowed temperature. Also, the size of this system is one quarter of the final proposed system. Since the results of the mathematical model have been validated with the experimental tests, it is possible to conclude that the simulations performed previously by the research group [36] for a model with final dimensions and hot faces temperature of 250 ºC are close to realistic electrical output. The same results pointed out a fuel saving between 4-5% and CO2 emission savings of around 11-12 g/km depending on the driving cycle conditions (see Table 11). Table 11 - Savings from using TCTG (Using TEGs GM250-12) [36] Driving cycle data: HW WLTC Distance [km] 27.1 23.3 Average engine efficiency [%] 25.3% 17.7% Average required mechanical power [kW] 18.7 8.58 Using TCTG: Electric power produced (ave.) [W] 627 267 Energy produced [kJ/km] 26.6 20.7 Average required Mechanical Power [kW] 17.7 7.00 Fuel savings [L/100km / %] 0.50 / 5.4% 0.47 / 4.2% CO2 emissions savings [g/km] 11.99 11.14 97 6. NEW PROTOTYPE CONSTRUCTION AND SET UP At this point, the tests on the existing prototype were completed. The TCTG was tested with the heat pipes working at atmospheric pressure, therefore the working fluid vaporization temperature achieved was 100 ºC. Nevertheless, the excess heat spreading feature was confirmed. The construction of a new prototype was planned. This time, overcoming the initial construction difficulties that led to the appearance of leaks in the heat pipes. This chapter reports in detail on the construction work of the final prototype, as well as changes to the design, the methods used and the operation of the system. 6.1 New Design The new TCTG was based in the existing prototype principle of operation with the addition of some design improvements resulting from the learning made throughout the testing process of the previous prototype. Since the main objective was the construction of a TCTG prototype capable of withstanding high operating pressures (around 40bar), the construction of the part that is associated to this operation was planned and designed in cooperation with a company specialized in industrial boilers and Heat Exchangers, Energest®. The final design (see Figure 63) was achieved, satisfying the safety concerns of the company as a designer of high-pressure devices and the scientific needs of the research group to have this type of operation specified. Figure 63 - New prototype design. 98 6.1.1 New Heat pipes One of the most relevant changes in this prototype when compared to the previous one was the material of the Heat Pipes. The previous prototype was equipped with copper heat pipes, which were the reason why they did not resisted the aluminium matrix casting process and broke in the most critical areas of their structure. The new prototype was equipped with stainless steel HPs manufactured specifically for this prototype with welding made by specialised personnel, capable of withstanding both the temperatures of the aluminium matrix casting process and the pressure range of operation. The new HPs can be seen in Figure 64. A copper fabric was added to the interior of the pipes in order to provide the wick structure needed for the capillary pumping effect. Figure 64 - (a) CAD model and (b) built stainless-steel Heat Pipes. 6.1.2 Corrugated Pipes The exhaust gases circulate in the Corrugated pipes and exchange heat along their length. This technology can be found also in EGR (Exhaust Gas Recirculation) coolers, in order to rapidly reduce the temperature of the hot exhaust gases [91]. The corrugated pipes were manufactured by BorgWarner® and its configuration was designed specifically for this prototype (see Figure 65). 99 Figure 65 - Corrugated Pipes. Some positioners were placed in the corrugated pipes to position the heat pipes inside the moulding box for casting the aluminium, but also for the creation of the separation/discontinuity zone of the aluminium matrix, discussed previously in chapter 3.1. The spacing between the tubes was the same used in the existing prototype, which was optimized for minimizing thermal resistance and pressure drop, as published in a paper of the group. 100 6.2 New TCTG Construction 6.2.1 Casting Process The moulding box for the aluminium casting was prepared. This box created the aluminium body that promotes the heat transfer between the corrugated tubes, the heat pipes and the hot faces of the TEG modules. The moulding box was designed to position the corrugated pipes and the heat pipes at the desired position in order to properly cast the aluminium. Figure 66 displays the CAD drawing of the set. Figure 66 - CAD drawing of the moulding box. The process of building this new prototype began with the construction of its body. For this, the corrugated tubes and the HPs inside the moulding box were positioned for subsequent casting of the aluminium. In Figure 67 it is possible to see the positioning of the corrugated tubes and their centring with the moulding box. Figure 67 - (a) Moulding Box and (b) Heat Pipe positioning inside the moulding box. 101 It was necessary to ensure that during the casting process, the positioners placed on the new HPs prevent the aluminum casting from filling unwanted regions. A thermal silicone paste, capable of withstanding 1400ºC, was applied to the HPs so as not to suffer any damage due to the high temperatures of the casting process. The openings between the positioners and the heat pipes were then sealed as shown in Figure 68b. Figure 68 - (a) high temperature silicone and (b) after application on positioners. Next, the corrugated tubes and the heat pipes were placed inside the moulding box. The design of this TCTG required the creation of three areas that could not be contaminated by aluminium: the area of discontinuity mentioned above, and the two ends of the corrugated tubes. Therefore, these areas were coated with fine silica sand. Figure 69a shows the closed moulding box with the three specified zones where the aluminium must not enter as well as the preparation of the special sand that was used. Next, Figure 69b shows the respective areas already covered with compressed sand. Finally, Figure 69c shows in detail the future discontinuity zone of the aluminium matrix. Figure 69 - Moulding Box (a) before and (b) after the sand application and (c) close up of the discontinuity zone 102 After the moulding box preparation was complete, the aluminium casting and casting process started. First, the moulding box containing the HPs and the corrugated tubes was placed in the muffle furnace. The purpose of this machine is to heat the box so that the aluminium does not cool down abruptly when it is cast. The moulding box was placed inside the muffle furnace at 350 ºC for 1 hour. The muffle furnace used as well as its display are shown in Figure 70. Figure 70 - (a) Muffle Furnace and (b) digital display. At the same time as the moulding box was in the muffle furnace, the aluminium ingot was placed in the furnace to melt. The ingot used was an aluminium AlSi12 alloy, due to its high silicon content this alloy has a high fluidity, which will be important due regarding the small size of the cavities it has to fill when cast. Its thermal conductivity is around 120 W/m.K. The ingot was cast inside the furnace at 720 ºC and after 30 minutes of verification of complete melt the alloy was de-gazified for 20 minutes using argon gas. The alloy was then refined and modified with titanium-bor aluminium and strontium aluminium for 20 minutes. The furnace and the de-gasification process are presented in Figure 71. 103 Figure 71 - (a) Melting Furnace, (b) de-gasification of the alloy using (c) Argon. After the aluminium casting process in the furnace was completed and the pre-set time for preheating the moulding box in the muffle had elapsed, the moulding box cointaing the HPs and corrugated pipes was removed. Figure 72 shows the moulding box exiting the muffle. Figure 72 - Moulding box exiting the Muffle Furnace. The aluminium in the furnace was then cooled down to 700ºC and cast under mechanical vibration to facilitate the casting process considering the geometrical complexity of the various cavities of the piece (see Figure 73). 110 Figure 84 - (a) Excess condensers positioning and (b) close up view. 6.2.4 Relief Valve For pressure control stability and safety reasons, a relief valve was added to the highest part of the prototype. Figure 85a shows the expansion valve placement in the TCTG above the expansion vessel. It was designed to be connected to both upper and bottom HP system and dimensioned to withstand the TCTG inner HP’s pressure during high thermal load events. It is activated at 60 bar, preventing the system from exceeding pressures that may jeopardise its structural integrity. This valve is a SAFE-TCP 960 Series and can be seen in Figure 85b. Figure 85 - Expansion Valve. The relief valve manufacture datasheet can be found in Appendix 2 – Security Valve Datasheet. 111 6.2.5 TEGs and Cooling Plates placement Next, the TEGs were prepared for placement in the HX (see Figure 86a). Since these are the same ones that were used in the first prototype, they were individually tested to ensure their proper functioning Figure 86 - (a) TEG preparation and (b) cooling plate cleaning. The cooling plates were also cleaned and prepared for fitting to the new prototype. For that an electric bench polisher was used (see Figure 86). To minimize thermal contact resistances the HX was coated with the same thermal paste used in the first prototype (recall Figure 34a). The TEG modules were placed on top of this paste. Thin alumina sheets were also placed between the TEGs to provide electrical insulation with a low thermal resistance, the HX and the cooling plates. In Figure 87 it is possible to see the placement of the TEGs and in Figure 87b the alumina sheets. (a) (b) Figure 87 - (a) TEG placement on the HX and (b) alumina sheets covering The TCTG prototype after the assembly of the TEGs and the Cooling Plates can be seen in Figure 88. 112 Figure 88 - Prototype after the assembly of the TEGs and the Cooling Plates 6.3 System Set-up The main components of the system were built, lacking only the assembly of everything so that the system would be ready for testing. All the connections between the TEGs and the load resistors were made and connected to the previously built “custom box” for connection with the data acquisition plug (see Figure 89). Figure 89 - Prototype after load resistance connected to the TEGs 113 The exhaust gas collectors were fitted to the prototype. Figure 90b shows that they were sealed with hightemperature silicone to prevent any exhaust gases leakage (highlighted in red). (a) (b) Figure 90 - (a) Exhaust Gas Collector and (b) possible leakage sealing. The water system pipes from the cooling plates (in red) as well as those of the new excess condensers (in black) were connected to the system (see Figure 91). Figure 91 - Water supply system connections. 114 The exhaust pipe coming from the engine was connected and the exhaust manifolds, both inlet and outlet were covered with glass wool and a film of refractory tape to thermally insulate these parts (see Figure 92). Figure 92 - (a) Exhaust gas collector connected to exhaust pipe and (b) coated with glass wool. All thermocouples, pressure sensors and load resistors were connected to the data acquisition system. This connection is described in detail in chapter 4.1.4. (a) (b) 115 6.3.1 Exhaust Collectors The exhaust manifolds/collectors for the inlet and outlet of the HX were built. Stainless steel was used for this piece. It was firstly designed using SolidWorks ® SheetMetal to obtain the necessary bends, to achieve the desired form. Then, it was cut and bent in order to obtain the final shape. Finally, it was MIGwelded. The CAD drawing and the built collector are displayed in Figure 93. (a) (b) (c) Figure 93 - (a)Exhaust Gas Collectors position in TCT;(b) Exhaust Collectors Sheet Metal Drawing, (b) MIG Welding on exhaust collectors and (c) junctions to connect the exhaust pipe. 116 6.4 Brake Dynamometer In order for the engine to produce torque and generate the maximum possible heat it is necessary to simulate a mechanical load to the engine. The dynamometer brake consists of a rotor connected to the engine, a stator supported on bearings which is fixed by a load cell placed at a known distance to the axis of rotation (𝑑), where the load cell measures a force which is being exerted to keep the stator stationary (𝑓). The torque is obtained by multiplying 𝑓 by 𝑑 as shown in Figure 94. Figure 94 - Dynamometric brake Scheme [92] . The brake dynamometer available in the laboratory was used for this purpose. It is an electric dynamometer working on the Eddy currents principle. A throttle was used to control the engine load as the control variable. The dynamometer was linked to a digital readout system which contains the digital RPM meter. In Figure 95 it is possible to see the break dynamometer complete system. 117 Figure 95 - Brake Dynamometer system. The cooling fan has the role of reducing the temperature of the brake during its operation, preventing it from reaching temperatures that may damage its internal components. This system was not used in the first prototype because the electronic controller was missing. Thus, considering the importance of this feature, a controller was developed for the project. Figure 96 displays the controller. Figure 96 - Dynamometer Brake Controller. Cooling Fan Dynamometric brake Fan Controller 118 This controller enables a set rpm to be defined for the engine so that, regardless of the position of the throttle, the brake locks the engine, maintaining the same engine speed, producing more or less torque. The controller is equipped with a display zone that allows the user to check the set rpm, the real rpm and the torque produced. 6.4.1 Calibration A calibration of the brake was also necessary as the values shown by the display were not in conformity with reality. For that, several weights with known masses were placed on an arm centred on the brake rotation axis (see Figure 97). The values presented on the display were noted. Figure 97 - Dynamometric Brake Calibration Procedure. The moments of inertia of the arm were previously calculated to make a calibration curve that states the real torque value for each corresponding value that appeared on the display by the time of calibration (see Figure 98). 119 Figure 98 - Calibration Sheet for the Brake Dynamometer. This calibration enabled to determine the real torque values produced by the engine during the test cycle of the experimental test performed. . 126 Despite the undesirable noise, the graphics of Figure 104 present a comparison made between the TEG’s performance in the experimental test cycle with the manufacture’s performance datasheet. The same correction presented in Chapter 5.1.5 was made, where the value of the internal resistance of the module was brought close to the real one. In Figure 104a and Figure 104b the green line represents the performance of the TEGs based on the manufacturer datasheet. In conclusion, it can be stated that, on this experimental test, the performance of the modules is in accordance with what is announced by the manufacturer (after correction of the resistance). 127 7.2 Fuel and CO2 Savings Although the new prototype could not maintain pressure stability due to the expansion valve leakage problem, it was possible to test the system with a module hot face temperature close to 250ºC (TEG module optimum/maximum allowed temperature). The same fuel savings and CO2 emissions analyses presented in chapter 5.4 were made regarding the experimental results obtained with the new system. Table 13 presents the fuel and emissions savings for the case where the TCTG would be implemented for reducing the use of the alternator on the test cycle performed. Table 13 - Vehicle and Alternator Fuel Power savings using new TCTG Alternator and TEG chain of conversion Value Unit Alternator Power 1.44 kW Alternator Mechanical Power (η=55%) 2.62 kW Alternator Fuel Power (η=18%) 14.5 kW TEG Output 300 W TEG Fuel Power (if produced by the Alternator) 3.03 kW Parameter Vehicle Fuel Power 130.5 kW Vehicle Total Fuel Power Consumption 29.0 g/s Alternator Total Fuel Power Consumption 0.32 g/s Alternator Fuel saving Using TEGs 0.067 g/s Savings Alternator Fuel Power Savings using TEGs 20.83% % Vehicle Fuel Power Saving using TEGS 2.3 % The new prototype can produce around 300 W (when fully operative) and replace the alternator in 20% of its load. It is noteworthy that this estimate is made considering alternator maximum load events. This is a significant increase when comparing with the previous prototype working at atmospheric pressure (only 4%). Although the value for the vehicle fuel power savings is small, the fact that the system is harvesting otherwise wasted energy and providing fuel power savings of 2.3% is breakthrough to this type of energy recover systems. 128 8. CONCLUSIONS AND FUTURE WORK The present work dealt with the development, manufacture and full experimental validation of a novel Compact Temperature Controlled Thermoelectric Generator (TCTG) concept to recover the exhaust heat of engines. An original concept proposed by the team of the Engine Lab of the University of Minho (LaMoTA) had been previously proposed but had never been tested and the concept validated. This concept consists of a TCTG using high performance stainless steel corrugated pipes for the exhaust flow, which are embedded in a cast aluminium matrix, along with variable conductance heat pipes (VCHPs) that work as excess heat spreaders for temperature control under the variable thermal load found in real driving. When TEG modules temperature limit risks being exceeded at the hotter upstream regions of the HX, the VCHPs start absorbing heat by vaporization and spread the excess heat by condensation at the colder regions of the heat exchanger. Firstly, an existing, limited pressure downsized prototype was instrumented and tested. Subsequently, an enhanced experimental downsized prototype was built and also tested. Due to technical problems in the casting process of the aluminium matrix, the copper HPs of the existing prototype presented leaks, making it impossible to pressurize them above atmospheric pressure, thus the working fluid inside the HPs could only operate as excess heat spreader at 100ºC, a temperature that is sub-optimal for the TEG generators. Nevertheless, this prototype could be used to validate the excess heat spreading and temperature control features claimed by the LAMoTA team in their patent submission and assessed numerically in previous publications. Therefore, the construction of the existing prototype was completed and it was instrumented. Namely, a water-cooling system was built including optimized finned cooling plates and pipelines as well as temperature, pressure and voltage data acquisition sensors that were incorporated into the prototype. An experimental validation was carried out where the prototype was tested with a 1.6L petrol engine and the following results were obtained: -TCTG heat exchanger effectiveness between 80-90% throughout the test cycle, indicating a good thermal design for the Heat Exchanger. -The use of heat pipes proved to have a significant role in thermal control, allowing the system to operate at constant temperature even with excess heat, therefore avoiding thermal degradation of TEG modules under variable loads and providing a controlled thermal level to the TEGs. It was possible to observe the sequential (cascading) module activation during the heat spreading phase and the limitation in temperature. 129 -An average electrical power output of 60W was obtained when the HPs became fully activated, with a peak of 90W with fluid vaporization temperatures slightly over 100ºC. -The TEG´s recorded a maximum efficiency of 4%, which seems promising at this low temperature, as they are close to the manufacturer’s maximum efficiency of 5% for a considerable higher temperature difference. The engine exhaust data recorded in the experimental test was used as input for the existing numerical model. The results allowed to preliminarily validate the model of the concept: - The prediction of the evolution of the total electric power produced was reasonably predicted, with the full HP operation occurring at 260s against the approximately 300 s recorded experimentally. The predictions of the power under full system operation were also relatively close to the value recorded experimentally (64W). Due to the lack of excess vapour condenser, the prototype continued to increase the output beyond the full system level. After confirming the mathematical model validation, a simulation for the same driving cycle, this time with the vaporization temperature of 250ºC was made, predicting that the system could achieve a maximum power output of 350W, which seems substantial, given that it is a downsized prototype (4 times smaller than future full-size prototypes). -For the simulation of a fully sized prototype on a standard WTLC, using the working fluid vaporization temperature/modules maximum temperature of 250ºC the system could produce more than 1.5kW. These figures seem to be highly promising for TEG systems in such small vehicles and are unparalleled in literature. Also, a similar but more robust TCTG concept using stainless steel Heat Pipes instead of copper one’s, aiming to overcome casting difficulties and therefore test the concept under saturation temperatures of 250ºC and inner pressures in excess of 40 bar, was built and tested experimentally with a 1.6L petrol engine connected to a brake dynamometer. The following results were obtained: - Minimum TCTG concept effectiveness between 55 and 65%, indicating a lower effectiveness than the previous prototype but at a much higher thermal load, still a convenient thermal design for the Heat Exchanger as it enabled the full operation of the VCHPs (maximum TEG power) -The use of stainless-steel heat pipes proved to overcome the casting process difficulties and to have a significant role in thermal control while withstanding high pressure, allowing the system to operate at nearly constant temperature, therefore avoiding the thermal degradation of TEG modules under variable loads and spreading local excess heat instead of wasting this heat. 130 -Average electrical power output around 300W under the condition of HPs fully activated (optimal TEG temperature level, i.e., close to 250ºC). -The TEG´s maximum obtained efficiency of 7% suggests that the actual performance of TEGs might be better than the performance advertised by the manufacturer. The prototype concept seems to validate previous thermal design optimization proposals done by the research group and provide a temperature controlled thermoelectric generator for vehicle application with no moving parts and little to no maintenance needs. The experimental and numerical validation of the systems seems to indicate that it has a high potential for waste energy recovery applications on vehicles, because it seems able to maintain the TEG modules close to their optimal hot side temperature while avoiding overheating, spreading excess heat instead of wasting it through by-pass systems, and also avoiding thermal dilution in low heat load events because it is able to concentrate the limited amount of heat into fewer modules because the heat spreading only occurs when heat is excessive. In terms of thermal design, the use of corrugated pipes from Exhaust Gas Recirculation (EGR) cooler manufacturers and Heat Pipes to promote thermal control seems clearly an advantage, being the copper heat pipes problematic in terms of aluminium casting due to their thermal dilatation, and the stainless-steel ones more suitable to the manufacture process and to withstand system function in high pressure. While being true that TEG modules still provide a low conversion efficiency from heat to electrical power at the present, typically below 5-7% for the temperature ranges of these applications, it is noteworthy that the studied concept allows a thermal optimization that will be even more valuable as higher figure of merit thermoelectric materials are steadily appearing in the market. Also, it is harvesting otherwise wasted energy that was not converted into useful work by the engine. Given its free availability and capability of conversion, it is only a question of evaluating its implementation viability in terms of economics, energy and environmental payback. 131 8.1 Future Work In regards to the novel stainless steel HP prototype, future tests are required after valve repair to study in full the pressure evolution of the TCTG. Also, a wider engine speed range tests are required to fully study the excess condenser operation feature. Regarding the Exhaust2Energy project, after the present validation of the thermal design and the mathematical model, the LaMoTA research group can confidently take the step of beginning the construction of a full-size prototype to experimentally validate the prediction of 1.5kW made by the mathematical model. Also, a full study of the HP behaviour under dynamic conditions, including adverse forces during breaking and accelerating, and a study of a by-pass valve implementation upstream of the HX is required. Note that this valve is designed to open once the thermoelectric generator achieves its full power (unlike competitive alternative designs) since under these conditions the heat exchanger cannot absorb more heat. The system has still not been optimized in terms of mass, and form factor. This optimization could be done, along with the use of phase change fluids that could operate at lower pressures (i.e. DowTherm-A) to allow a lighter concept VCHPs could be directly machined or formed into the body of the heat exchanger (as currently explored in project COOLSPOT), and laser textured surfaces used to improve phase-change Suitable compact heat exchangers (i.e. wavy fins) combined with the use of these systems with heavy duty vehicles (as explored in a parallel dissertation) could also be further explored numerically and experimentally to make this concept further promising. 132 133 REFERENCES [1] R. Muncrief and J. German, “Defeat devices under the U . S . and EU passenger vehicle emissions testing regulations,” no. March, 2016. [2] M. P. Tietge U, Sonsoles D, Zifei Y, “From laboratory to road international: A comparison of official and real-world fuel consumption and CO2 values for passenger cars in Europe, the United States, China, and Japan,” ICCT White Pap. , no. November, p. 69, 2017, [Online]. Available: www.theicct.org. [3] U. Tietge, P. Mock, V. Franco, and N. Zacharof, “From laboratory to road: Modeling the divergence between official and real-world fuel consumption and CO2 emission values in the German passenger car market for the years 2001–2014,” Energy Policy , vol. 103, no. January, pp. 212– 222, 2017, doi: 10.1016/j.enpol.2017.01.021. [4] N. E. Ligterink and A. R. A. Eijk, “Update analysis of real-world fuel consumption of business passenger cars based on Travelcard Nederland fuelpass data,” no. June, p. 25, 2014, doi: 10.13140/RG.2.1.3212.5040. [5] D. F. Dominković, I. Bačeković, A. S. Pedersen, and G. Krajačić, “The future of transportation in sustainable energy systems: Opportunities and barriers in a clean energy transition,” Renew. Sustain. Energy Rev. , vol. 82, no. June 2017, pp. 1823–1838, 2018, doi: 10.1016/j.rser.2017.06.117. [6] “European Commission, ‘Delivering the European Green Deal | Climate Action,’” 2021. [7] European Commission, “Reducing CO2 emissions from heavy-duty vehicles | Climate Action,” 2019. . [8] W.-R. Chang, J.-J. Hwang, and W. Wu, “Environmental impact and sustainability study on biofuels for transportation applications,” Renew. Sustain. Energy Rev. , vol. 67, pp. 277–288, Jan. 2017, doi: 10.1016/J.RSER.2016.09.020. [9] A. Alagumalai, “Internal combustion engines: Progress and prospects,” Renew. Sustain. Energy Rev. , vol. 38, pp. 561–571, Oct. 2014, doi: 10.1016/J.RSER.2014.06.014. [10] W. Chen, J. Liang, Z. Yang, and G. Li, “A Review of Lithium-Ion Battery for Electric Vehicle Applications and Beyond,” Energy Procedia , vol. 158, pp. 4363–4368, Feb. 2019, doi: 10.1016/J.EGYPRO.2019.01.783. [11] P. Letmathe and M. Suares, “A consumer-oriented total cost of ownership model for different vehicle types in Germany,” Transp. Res. Part D Transp. Environ. , vol. 57, pp. 314–335, 2017, doi: 10.1016/j.trd.2017.09.007. [12] E. E. Agency, “Sharp decrease in CO2 emissions of new cars in 2020 — European Environment Agency,” 2021. . [13] T. Hoffman, “Alternative Fuels and Advanced Vehicle Technologies for Improved Environmental Performance: Towards Zero Carbon Transportation,” Woodhead Publ. Energy , vol. 57, pp. 565– 581, 2014. [14] J. Martins and F. P. Brito, “Alternative fuels for internal combustion engines,” Energies , vol. 13, p. 4086, Aug. 2020, doi: 10.3390/en13164086. [15] F. P. Brito, J. Martins, F. Lopes, C. Castro, L. Martins, and A. L. N. Moreira, “Development and assessment of an over-expanded engine to be used as an efficiency-oriented range extender for electric vehicles,” Energies , vol. 13, no. 2, 2020, doi: 10.3390/en13020430. [16] F. Zhang, X. Hu, R. Langari, and D. Cao, “Energy management strategies of connected HEVs and PHEVs: Recent progress and outlook,” Prog. Energy Combust. Sci. , vol. 73, pp. 235–256, Jul. 2019, doi: 10.1016/J.PECS.2019.04.002. 134 [17] J. Martins, Motores de Combustão Interna , 5a Edição. Engebook, 2016. [18] S. N. Hossain and S. Bari, “Waste heat recovery from the exhaust of a diesel generator using Rankine Cycle,” Energy Convers. Manag. , vol. 75, pp. 141–151, Nov. 2013, doi: 10.1016/j.enconman.2013.06.009. [19] C. RE and C. D, “Combined cycle for hybrid vehicles,” SAE , 2015. [20] E. Massaguer, A. Massaguer, T. Pujol, M. Comamala, L. Montoro, and J. R. Gonzalez, “Fuel economy analysis under a WLTP cycle on a mid-size vehicle equipped with a thermoelectric energy recovery system,” Energy , vol. 179, pp. 306–314, 2019, doi: 10.1016/j.energy.2019.05.004. [21] T. Wang, Y. Zhang, Z. Peng, and G. Shu, “A review of researchs on thermal exhaust heat recovery with Rankine cycle,” Renew. Sustain. Energy Rev , vol. 15, 2011. [22] Z. Peng, T. Wang, Y. He, and L. Lu, “Analysis of environmental and economic benefits of integrated Exhaust Energy Recovery (EER) for vehicles,” Appl. Energy , pp. 238–243, 2013. [23] T. Wang, Y. Zhang, J. Zhang, G. Shu, and Z. Peng, “Analysis of recoverable exhaust energy from a light-duty gasoline engine,” Appl. Therm. Eng. , vol. 53, no. 2, pp. 414–419, 2013, doi: 10.1016/j.applthermaleng.2012.03.025. [24] D. Rowe and C. Bhandari, CRC Handbook of Thermoelectrics . 1995. [25] F. P. Brito et al. , “Analysis of thermoelectric generator incorporating n-magnesium silicide and ptetrahedrite materials,” Energy Convers. Manag. , vol. 236, p. 114003, May 2021, doi: 10.1016/J.ENCONMAN.2021.114003. [26] L. M. Goncalves, J. Martins, J. Antunes, R. Rocha, and B. F., “Heat-Pipe Assisted Thermoelectric Generators for Exhaust Gas Applications,” vol. 5, 2010. [27] F. P. Brito, J. Martins, L. M. Goncalves, and R. Sousa, “Temperature controlled exhaust heat thermoelectric generation,” SAE Tech. Pap. , no. January 2016, 2012, doi: 10.4271/2012-011214. [28] F. P. Brito, J. Martins, L. M. Gonçalves, and R. Sousa, “Temperature Controlled Exhaust Heat Thermoelectric Generation,” SAE Int. J. Passeng. Cars - Electron. Electr. Syst , vol. 5, pp. 561– 571, 2017. [29] and J. A. J. Martins, F. P. Brito, L. M. Goncalves, “Thermoelectric Exhaust Energy Recovery with Temperature Control through Heat Pipes,” 2011. [30] F. P. Brito, J. Martins, L. M. Goncalves, and R. Sousa, “Temperature controlled exhaust heat thermoelectric generation,” SAE Tech. Pap. , 2012, doi: 10.4271/2012-01-1214. [31] F. P. Brito, L. M. Goncalves, J. Martins, N. Antunes, and D. Sousa, “Influence of heat pipe operating temperature on exhaust heat thermoelectric generation,” SAE Int. J. Passeng. Cars - Mech. Syst. , vol. 6, no. 2, 2013, doi: 10.4271/2013-01-0559. [32] F. P. Brito, J. Martins, E. Hançer, N. Antunes, and L. M. Gonçalves, “Thermoelectric Exhaust Heat Recovery with Heat Pipe-Based Thermal Control,” J. Electron. Mater. , vol. 44, no. 6, pp. 1984– 1997, Jun. 2015, doi: 10.1007/s11664-015-3638-3. [33] F. P. Brito et al. , “Analysis of the Effect of Module Thickness Reduction on Thermoelectric Generator Output,” J. Electron. Mater. , vol. 45, no. 3, pp. 1711–1729, 2016, doi: 10.1007/s11664-015-4182-x. [34] F. P. Brito et al. , “Analysis of a Temperature-Controlled Exhaust Thermoelectric Generator During a Driving Cycle,” J. Electron. Mater. , vol. 45, no. 3, pp. 1846–1870, 2016, doi: 10.1007/s11664015-4258-7. [35] F. P. Brito et al. , “Efficiency improvement of vehicles using temperature controlled exhaust thermoelectric generators,” Energy Convers. Manag. , vol. 203, no. November 2019, p. 112255, 2020, doi: 10.1016/j.enconman.2019.112255. [36] N. Pacheco, F. Brito, R. Vieira, J. Martins, H. Barbosa, and L. M. Goncalves, “Compact automotive thermoelectric generator with embedded heat pipes for thermal control,” Energy , vol. 197, p. 135 117154, Feb. 2020, doi: 10.1016/j.energy.2020.117154. [37] and R. V. F. P. Brito, J. Martins, L. M. Goncalves, J. Teixeira, N. Pacheco, “System for efficient heat recovery and method thereof,” 2018. [38] N. Pacheco, F. P. Brito, R. Vieira, J. Martins, H. Barbosa, and L. M. Goncalves, “Compact automotive thermoelectric generator with embedded heat pipes for thermal control,” Energy , vol. 197, 2020, doi: 10.1016/j.energy.2020.117154. [39] N. Pacheco, F. P. Brito, R. Vieira, J. Martins, H. Barbosa, and L. M. Goncalves, “Compact automotive thermoelectric generator with embedded heat pipes for thermal control,” Energy , vol. 197, p. 117154, 2020, doi: 10.1016/j.energy.2020.117154. [40] F. P. Brito, J. Martins, L. M. Goncalves, J. Teixeira, N. Pacheco, and R. Vieira, “System for efficient heat recovery and method thereof,” US Patent application, Attorney Docket number 10224/008071/US0, submitted August 8th 2019, originally filed in Portugal in August 8th. 2018 (application number 111106), 2019. [41] a Thesis, “Modeling of an Automotive Exhaust,” no. June, 2005. [42] H. Julian Goldsmid, Introduction to Thermoelectricicty . 2010. [43] “Figura Seebeck,” 2008. https://searchnetworking.techtarget.com/definition/Seebeck-effect. [44] R. M. Sousa, “Gerador Termoelétrico para Escape do Automóvel com Controlo de Temperatura,” 2011. [45] D. Ji et al. , “Thermoelectric generation for waste heat recovery: Application of a system level design optimization approach via Taguchi method,” Energy Convers. Manag. , vol. 172, no. April, pp. 507–516, 2018, doi: 10.1016/j.enconman.2018.06.016. [46] S. Siouane, S. Jovanovic, and P. Poure, “A Novel Identification Method of Thermal Resistances of Thermoelectric Modules Combining Electrical Characterization Under Constant Temperature and Heat Flow Conditions,” Trans. Environ. Electr. Eng. , vol. 1, no. 4, p. 44, 2016, doi: 10.22149/teee.v1i4.72. [47] T. M. Tritt, X. Tang, Q. Zhang, and W. Xie, “Solar thermoelectrics: Direct solar thermal energy conversion,” Fundam. Mater. Energy Environ. Sustain. , vol. 33, no. January, pp. 289–294, 2011, doi: 10.1017/CBO9780511718786.026. [48] S. S. Siouane, S. Jovanovic, and P. Poure, Influence of contact thermal resistances on the Open Circuit Voltage MPPT method for Thermoelectric Generators . 2016. [49] R. Bjørk, “The Universal Influence of Contact Resistance on the Efficiency of a Thermoelectric Generator,” J. Electron. Mater. , vol. 44, no. 8, pp. 2869–2876, 2015, doi: 10.1007/s11664015-3731-7. [50] T. P. Hogan et al. , “Nanostructured thermoelectric materials and high-efficiency power-generation modules,” J. Electron. Mater. , vol. 36, no. 7, pp. 704–710, 2007, doi: 10.1007/s11664-0070174-9. [51] A. Polozine, S. Sirotinskaya, and L. Schaeffer, “History of development of thermoelectric materials for electric power generation and criteria of their quality,” Mater. Res. , vol. 17, no. 5, pp. 1260– 1267, 2014, doi: 10.1590/1516-1439.272214. [52] J. Yang et al. , “On the tuning of electrical and thermal transport in thermoelectrics: An integrated theory-experiment perspective,” npj Comput. Mater. , vol. 2, no. October 2015, 2016, doi: 10.1038/npjcompumats.2015.15. [53] Y. A. Çengel, J. M. Cimbala, and R. H. Turner, Fundamentals of fluid-thermal sciences , vol. 2, no. 27. 2017. [54] W. K. J. and R. D. E., Termodinámica , 6th ed. Madrid: McGraw-Hill, 2001. [55] C. (Claus) Borgnakke and R. E. Sonntag, “Fundamentos da termodinámica,” 2018. [56] J. Lizarraga and A. Perez, Exergy Analysis and Thermoeconomics of Buildings . 2020. [57] M. S. Veto and P. R. Christensen, “Mathematical Theory of Thermal Inertia Revisited 1: Improving