Transcript
Research Dissertation for
An Automotive Carbon Dioxide Air-Conditioning System with Heat Pump by
Dipl.-Ing.(FH) C. Böttcher for the qualification
M Tech Eng Magister Technologiae: Engineering (Mechanical)
Promoters: Prof. Dr.-Ing. H. Holdack-Janssen Mr. J. Maczek
i
Copyright Statement This copy of the thesis has been supplied on condition that anyone who consults it is understood to recognise that its copyright rests with its author and that no quotation from the thesis and no information derived from it may be published without the author’s prior consent.
ii
Contents CHAPTER 1 INTRODUCTION ........................................................................................1 1.1 STATEMENT OF THE PROBLEM .................................................................................2 1.2 HYPOTHESIS ............................................................................................................3 1.3 SUB PROBLEMS........................................................................................................4 1.4 DELIMITATIONS ......................................................................................................5 1.5 ASSOCIATION WITH COMPANIES ..............................................................................6 CHAPTER 2 SPECIAL FEATURES AND REQUIREMENTS WHEN USING CARBON DIOXIDE AS A REFRIGERANT............................................7 2.1 PHYSICAL AND THERMODYNAMIC PROPERTIES OF CARBON DIOXIDE .....................8 2.2 THE CONTROL OF A CO2 REFRIGERANT CYCLE ......................................................10 2.2.1
Effect of increased ambient temperature...............................................10
2.2.2
Effect of increased high pressure ..........................................................12
2.2.3
Ideal high pressure................................................................................13 2.2.3.1 The definition of ideal high pressure .......................................13 2.2.3.2 The determination of ideal high pressure ................................14
2.3 THE ARRANGEMENT OF THE COMPONENTS ............................................................16 2.3.1
Possible cycles.......................................................................................16
iii
2.3.1.1 System without high pressure control......................................16 2.3.1.2 System with low pressure receiver (Lorentzen principle) .......18 2.3.1.3 System with a mid-pressure receiver.......................................19 2.3.1.4 System with refrigerant charge control....................................20 2.3.2
Assessment.............................................................................................21
2.3.3
The Lorentzen process...........................................................................22
2.4 THE COMPONENTS .................................................................................................23 2.4.1
The gas cooler .......................................................................................24 2.4.1.1 Parallel Flow principle.............................................................25 2.4.1.2 Serpentine principle .................................................................25
2.4.2
The evaporator ......................................................................................26
2.4.3
The internal heat exchanger..................................................................27 2.4.3.1 Comparison of advantages and disadvantages of using an internal heat exchanger ............................................................27 2.4.3.2 Internal heat exchanger designs...............................................28
2.4.4
The expansion valve ..............................................................................30 2.4.4.1 Solenoid valves........................................................................31 2.4.4.2 Stepping motor valves .............................................................31
2.4.5
The accumulator....................................................................................31
2.4.6
The compressor .....................................................................................32
2.4.7
Tubes and hoses.....................................................................................35
2.5 SUMMARY .............................................................................................................37 CHAPTER 3 THE TEST BENCH ...................................................................................39 3.1 THE REFRIGERANT CIRCUIT ...................................................................................39
iv
3.1.1
The heat exchanger circuit ....................................................................43
3.1.2
The oil return circuit .............................................................................50
3.1.3
The compressor control circuit .............................................................53
3.2 THE MEASURING SYSTEM ......................................................................................55 3.2.1
The computer.........................................................................................56
3.2.2
The signal conditioning unit..................................................................56
3.2.3
The sensors ............................................................................................56 3.2.3.1 The thermocouples...................................................................56 3.2.3.2 The pressure transducers..........................................................57 3.2.3.3 The mass flow meter................................................................58 3.2.3.4 Torque sensor...........................................................................59 3.2.3.5 The humidity sensor ................................................................61
3.2.4
The software ..........................................................................................61 3.2.4.1 The DAC module.....................................................................62 3.2.4.2 The VIEW module...................................................................68 3.2.4.3 The DATA module ..................................................................68 3.2.4.4 GRAPH module.......................................................................70
3.3 SUMMARY .............................................................................................................70 CHAPTER 4 THE FLASH FOGGING PROBLEM......................................................72 4.1 INVESTIGATION OF FLASH FOGGING .....................................................................72 4.1.1
Incidents during the reversal of the cycle .............................................73
4.1.2
Dew point temperature..........................................................................74 4.1.2.1 Graphical method.....................................................................74 4.1.2.2 Calculation of the dew point temperature................................76
v
4.1.3
Flash Fogging in conventional AC cycles?...........................................78
4.2 MEASURES AGAINST FLASH FOGGING...................................................................79 4.2.1
Electrically heated windscreen .............................................................80
4.2.2
Redirection of air flow during change of operation mode ....................81
4.2.3
An additional HX in the refrigerant cycle .............................................81
4.2.4
Use of a secondary (coolant) cycle .......................................................82
4.2.5
Discussion .............................................................................................82
CHAPTER 5 SECONDARY CYCLE SYSTEM ON THE TEST BENCH .................85 5.1 ADDITIONAL COMPONENTS AND THEIR ARRANGEMENT ........................................85 5.2 THE SELECTION OF THE HEAT EXCHANGER FOR THE SECONDARY CYCLE ..............88 5.2.1
Considering a plate type heat exchanger ..............................................89
5.2.2
Considering a coaxial type heat exchanger constructed from stainless steel (material number: 1.4571)............................................................95
5.2.3
Consider a coaxial type heat exchanger constructed from copper (material number 2.0090)....................................................................102
CHAPTER 6 THE TESTS CONDUCTED AND RESULTS OBTAINED ................112 6.1 THE CONTROL CURVES ........................................................................................113 6.1.1
The compressor control valve .............................................................113
6.1.2
The expansion valve ............................................................................117
6.1.3
Conclusion to the controls...................................................................120
6.2 THE AIR-CONDITIONING MODE ............................................................................121 6.3 THE HEAT PUMP MODE ........................................................................................127 6.4 REVERSING FROM AC TO HP MODE ....................................................................131 6.4.1
Reversing without the secondary cycle ...............................................132 vi
6.4.2
Reversing with the secondary cycle.....................................................135 6.4.2.1 Heat exchanger arrangement 1 ..............................................136 6.4.2.2 Heat exchanger arrangement 2 ..............................................139 6.4.2.3 Conclusion to the various arrangements and Flash Fogging .142
CHAPTER 7 FUTURE OPPORTUNITIES..................................................................145 LIST OF REFERENCES ................................................................................................150 APPENDICES ..................................................................................................................154 A1 IMPROVED CYCLE: HEATER ONLY OPERATION .....................................................155 A2 IMPROVED CYCLE: MAXIMUM COOLING OPERATION ............................................156 A3 IMPROVED CYCLE: REHEAT OPERATION...............................................................157 A4 IMPROVED CYCLE: HEAT PUMP (ENGINE COOLANT AS THE HEAT SOURCE) ..........158 A5 IMPROVED CYCLE: HEAT PUMP (AMBIENT AIR AS THE HEAT SOURCE)..................159 A6 IMPROVED CYCLE: HEATER & HEAT PUMP ...........................................................160
vii
Acknowledgements I would like to thank the staff of the Faculty of Engineering, in particular, Joe Maczek for his great support.
Furthermore, I would like to thank Professor Holdack-Janssen who offered me the possibility to accomplish this research work.
Lastly, I would like to thank my colleagues Martin Konz, Thomas Mylo, Frank Kiehne, Arne Hiestermann, Carsten Wachsmuth, Gideon Haasbroek and Charel Marais, from the Institute of Vehicle Construction in Wolfsburg for their support, encouragement and the critical discussions.
viii
Author’s declaration At no time during the registration for the Magister Technologiae Degree has the author been registered for any other Technikon or University degree. Signed.................................... Date....................................
ix
Glossary
Block-orientated
Each component in the measuring software is represented by a rectangle (a block) on the screen. It can be placed there with drag&drop technology. These components can have input and output pins and special parameters which can be entered in pop-up menus.
Counterflow heat
In a counterflow heat exchanger the hot fluid flows in the opposite
exchanger
direction to the cold fluid.
Critical point
In a property diagram for phase change processes the critical point is defined as the point at which the saturated liquid and saturated vapour states are identical.
Crossflow heat
In a cross-flow heat exchanger the direction of fluids are
exchanger
perpendicular to each other.
DIAdem®
A commercial computer program widely used for measurement tasks in the German automotive industry.
HVAC box
This is the part of an automotive air-conditioning system which is located in the cab of the vehicle (normally under the dashboard). The HVAC box contains the blower, the evaporator, the heater core and the flaps for temperature control and air distribution.
LabView®
A commercial computer program for measurement and control tasks. It is widely used in the field of research and development.
x
Material number
Numerical system of classification for ferrous and non-ferrous metals which is defined by the standard DIN 17007.
P,h-diagram
Property diagrams like the p,h-diagram are used to study phasechange processes. In refrigeration technology the most common property diagram is the p,h-digram which shows the enthalpy on the x-axis and the pressure on the y-axis. Every property diagram contains a “saturated liquid line” and a “saturated vapour line”.
Relative humidity
Relative humidity is defined as the actual moisture content of the air relative to the total amount of moisture the air can hold at the same temperature.
Subcooling
The process of removing heat from refrigerant after condensation.
Superheated
If refrigerant vapour is at a temperature that is higher than its boiling
vapour
point at a given pressure, this state is called “superheated”.
Usable enthalpy
The difference in enthalpy of the refrigerant at the inlet and outlet of
difference
the evaporator is called “usable enthalpy difference”.
xi
List of Figures Figure 1: Effect of increased ambient temperature in p,h-diagram .....................................11 Figure 2: Effect of increased high pressure on the p,h-diagram ..........................................12 Figure 3: Curves of ideal high pressure at different gas cooler temperatures......................13 Figure 4: System without high pressure control ..................................................................17 Figure 5: System with low pressure receiver .......................................................................18 Figure 6: System with mid-pressure receiver.......................................................................19 Figure 7: System with refrigerant charge control ................................................................20 Figure 8: Illustration of the Lorentzen process ....................................................................22 Figure 9: Parallel flow design and principle of flow............................................................25 Figure 10: Serpentine design and principle of flow .............................................................26 Figure 11: A typical co-axial serpentine design for an internal heat exchanger..................29 Figure 12: Internal heat exchanger with a serpentine design...............................................30 Figure 13: Cross-section view of Item® profile ...................................................................40 Figure 14: Double clamping ring fitting ..............................................................................40 Figure 15: Illustration of the whole circuit in AC mode......................................................42 Figure 16: The CO2 compressor...........................................................................................43 Figure 17: Illustration of the whole circuit in HP mode ......................................................45 Figure 18: The gas cooler.....................................................................................................46
xii
Figure 19: The internal heat exchanger................................................................................47 Figure 20: Side view of the parallel-flow evaporator ..........................................................49 Figure 21: The evaporator....................................................................................................49 Figure 22: The principle of the oil separator........................................................................51 Figure 23: The oil separator .................................................................................................51 Figure 24: The wobble plate design principle......................................................................53 Figure 25: The wobble plate integrated into the compressor...............................................54 Figure 26: The forces at the wobble plate............................................................................54 Figure 27: The torque sensor................................................................................................60 Figure 28: The function blocks for data input......................................................................63 Figure 29: The function blocks required for temperature measurement..............................64 Figure 30: The function block for the temperature matrix...................................................65 Figure 31: The test results used for the calibration of the torque sensor .............................66 Figure 32: The function blocks for torque and mass flow measurement.............................66 Figure 33: The function block for measuring pressure ........................................................67 Figure 34: The function block for saving data .....................................................................68 Figure 35: Features of a psychrometric chart.......................................................................74 Figure 36: An example of the graphical determination of dew point temperature ..............75 Figure 37: HVAC box with coolant-to-water heat exchangers............................................86 Figure 38: The different operation modes of the HVAC box ..............................................87 Figure 39: The layout of the secondary circuit ....................................................................88 Figure 40: The principle of the plate-type heat exchanger ..................................................90 Figure 41: The typical dimensioning of the plate-type heat exchanger ...............................93 Figure 42: A sectional view of the connector piece...........................................................103
xiii
Figure 43: Connector block with “T”-piece.......................................................................104 Figure 44: The cross-sectional view of coaxial type heat exchanger.................................105 Figure 45: The refrigerant-to-coolant heat exchanger .......................................................111 Figure 46: Control curves: compressor control valve ........................................................114 Figure 47: Control curves: expansion valve.......................................................................118 Figure 48: Air-conditioning operation: pressure and volume flow....................................122 Figure 49: Air-conditioning operation: system temperatures and volume flow ................123 Figure 50: Air-conditioning operation: temperatures at internal HX.................................125 Figure 51: Heat Pump operation: pressures and volume flow ...........................................128 Figure 52: Heat Pump operation: system temperatures and volume flow .........................129 Figure 53: Reversing the cycle without the secondary cycle.............................................133 Figure 54: Dew point temperature without the secondary cycle........................................134 Figure 55: Cross-sectional view of the HVAC box ...........................................................135 Figure 56: Reversing of the cycle with a secondary cycle to arrangement 1.....................137 Figure 57: Dew point temperature with secondary cycle arrangement 1...........................138 Figure 58: Reversing of the cycle with secondary cycle to arrangement 2 .......................140 Figure 59: Dew point temperature with secondary cycle arrangement 2...........................141
xiv
List of Tables Table 1: Comparison between CO2 and R134a …………………………………………
8
Table 2: Comparison of several methods to determinate the ideal high pressure ……… 16 Table 3: The sensors and their position ………………………………………………… 69 Table 4: The heat exchanger characteristics for the Valeo® Corion 15T14.3 as air cooler ………………………………………………………………………. 92 Table 5: Technical data of the selected plates ………………………………………….. 93 Table 6: The results of the calculation of the plate type evaporator ……………………. 94 Table 7: An abridged overview of the control characteristics for the system ………….. 121
xv
Abbreviations AC
Air-conditioning
CO2
Carbon Dioxide, when used as a refrigerant, is classified as R744
COP
Coefficient of performance
DAC
Data Acquisition
HVAC
Heating Ventilation and Air-conditioning
HP
Heat Pump
HX
Heat Exchanger
LPT
Line Printer Terminal
LSG
Laminated Safety Glass
PWM
Pulse-Width-Modulated
R12
Refrigerant: Difluorodichlorinemethane
R134a
Refrigerant: Tetrafluoroethane
TSG
Toughened Safety Glass
xvi
Chapter 1 Introduction
Chapter 1 Introduction The refrigerant circuits of car air-conditioning systems are fitted with so-called open type compressors, because there is only a lip seal preventing the refrigerant from leaking from the compressor housing to the atmosphere. In addition, the cycle uses damping elements between the compressor and the other components on the suction and pressure lines to reduce vibration and noise transfer from the engine to the car body. Both the lip seal and damping elements result in loss of refrigerant as they are made from elastomers and leak with age, and, under high temperature conditions inside the engine room, these elements also allow a relatively high permeation of the refrigerant gas to the atmosphere.
With very high refrigerant losses in the older R12 -cooling cycles and the damage caused by this gas to the ozone layer in the stratosphere, the Montreal protocol phased out this refrigerant and the car industry was forced to revert completely to R134a until 1994/95. R134a has no ozone depletion potential, but it has a direct global warming potential, and, therefore, leakages also have to be minimised. R134a has, because of its molecular size, a high permeation potential and, hence, all the refrigerant hoses are lined internally. Unfortunately, these hoses also leak with age and significant refrigerant loss will occur [1]
1
Chapter 1 Introduction
R134a can therefore only be viewed as a solution until an alternative refrigerant with no direct global warming potential has been developed. Candidates for new refrigerants are natural substances such as hydrocarbons or carbon dioxide [2]. Unfortunately, both substances have disadvantages and their use is restricted to special cases, for e.g. hydrocarbons are flammable and are not used in car air-conditioners, but in Germany it is used as a refrigerant in household refrigerators with hermetic cycles. What makes the implementation of carbon dioxide (CO2) difficult are the high system pressures and the low critical point [3].
1.1 Statement of the problem CO2 refrigerant cycles are under development in Europe for use in car air-conditioning (AC) systems. Other countries like Australia use R134a as well as propane [4], even though propane is flammable and poses a safety hazard. The European automotive industry considers propane to be too much of a product liability.
As a result of the high pressure levels and the installation of additional safety systems, e.g. CO2 detection in the cabin, CO2 cycles will obviously have higher production costs. On the other hand, this system can have additional functions such as a heat pump mode for heating the cabin. Also, the high suction pressure in the CO2 system provides a very low suction temperature and, therefore, the use of a variety of low temperature heat sources are possible. This allows the heat pump to heat by using a heat source, such as outside air at low ambient temperatures of –20°C or lower, eliminating the necessity for an additional water heating cycle [5].
2
Chapter 1 Introduction
Certain diesel engine vehicles which are optimised for fuel consumption lack “heat up” performance at low ambient temperatures and, hence, are fitted with additional electric heaters. Investigations by AUDI® show that a heat pump cycle has better heating performances in the “heat up” phase than a conventional water heating system with an additional electric heater. The reason for this is explained by the way in which heating is achieved. In the case of a conventional water heating system, it has to heat the entire mass of the engine block and coolant first, and then the air flow to the cabin. The heat pump warms the air flow to the cabin immediately from start-up, and, therefore, the elapsed time for effective heating is significantly reduced.
After the first prototype systems were developed it was realised that CO2 systems would be more expensive than R134a systems. To make CO2 systems more appealing to the market than the existing R134a systems, the CO2 system must be more functional. For this reason the heat pump function is a significant step in automotive CO2 technology. At present the “Flash Fogging” is the major problem in CO2 air-conditioning systems which have a heat pump. This problem can be described as the sudden “steaming-up” of the windscreen by the evaporated water from the heat exchanger’s surface during the process of reversing the cycle.
1.2 Hypothesis The aim of this research is the investigation of the “Flash-Fogging” phenomenon and the development of a suitable prevention or control strategy.
3
Chapter 1 Introduction
By using prototype components a test-bench with a vehicle CO2 air-conditioning system should be developed. The measured data obtained from the tests conducted would be the basis for the analysis of the “Flash-Fogging” Problem. On a theoretical basis different solutions against “Flash-Fogging” will be explained and assessed. The most promising solution will be realized at the test-bench to prove its potential in practice.
1.3 Sub problems The following sub problems apply to this research: •
Establishing a measurement system with DIAdem®
This contains on-screen visualization, recording and analysis of the values for pressure, temperature and mass flow of refrigerant and oil.
•
Air-conditioning mode
The control of a prototype CO2 system is a sensible balance of several adjustments. The setting of the expansion valve effects the high pressure, which is the most important factor concerning the refrigeration power gained and the efficiency of the system. The compressor control valve regulates the pressure level inside the crankshaft case. This pressure affects the position of the swash-plate and, hence, the capacity of the compressor and refrigerant mass flow rate. The oil return mass flow reacts likewise to the refrigerant mass flow and it is necessary to adapt the oil return mass flow to the operating condition. If the oil mass flow is too low (reduced lubrication) damage to the compressor will occur, and a high mass flow reduces the refrigeration performance. During the initiation of the test-bench in the air-conditioning mode the optimised parameters for all control devices must be established.
4
Chapter 1 Introduction
•
Heat pump operation
Following the realisation of the control for the air-conditioning mode, the test-bench must be extended for heat pump operation. Additional pipe work and a 4-way-valve allows for the reversal of the cycle. The parameters for the expansion valve and compressor control have to be determined experimentally.
•
Investigation of “Flash Fogging”
The conditions while switching over from the refrigeration mode to the heat pump operation needs to be investigated, especially the so-called “Flash Fogging”. The conditions for the occurrence of “Flash Fogging” must be established experimentally and, hence, a prevention strategy must be developed.
1.4 Delimitations •
The measurement system used must be based on DIAdem® because the Industrial partner for this project (Valeo Klimasysteme® GmbH) uses this software.
•
The work would be performed on a test-bench using standard available automotive components. The implementation of the system into a vehicle is not part of this work.
•
In heat pump mode, the evaporator will work as the heat sink and the gas cooler as the heat source. A comparison of different CO2 heat pump systems is not planned.
5
Chapter 1 Introduction
•
The Fachhochschule Braunschweig/Wolfenbüttel has no facility to supply an air current with defined temperature and humidity to the heating, ventilation and airconditioning (HVAC) box and the gas cooler. For this reason all tests will be performed at ambient temperature and humidity.
1.5 Association with companies Valeo Klimasysteme® GmbH and the Fachhochschule Braunschweig/Wolfenbüttel are cooperative partners for the CO2 project.
Valeo Klimasysteme® GmbH was the development partner for a prototype CO2 system for the current Mercedes-Benz® A-Class®. All prototype parts used for this project have been sourced from the before mentioned development work.
The compressor used during the research work has is origins from Zexel Valeo Compressor Europe® GmbH. This company is a joint venture between Valeo® and Zexel® for the development of a serial CO2 compressor.
The components sourced from Valeo® have prototype status and, therefore, only information essential for their implementation into the system was made available. Detailed information (material specification, design details etc.) beyond this was not disclosed on grounds of confidentiality.
6
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant Normally the introduction of an alternative refrigerant affects the design of the systems’ components and the materials used. With CO2 as a refrigerant, there are additional changes needed for the whole process. The reason for this is the relatively low critical temperature of 31.1°C. In most operational conditions of an automotive AC system, this temperature will be exceeded at the heat emission side of the system. Thus the automotive CO2 system is a transcritical process, i.e. the evaporation pressure is below the critical pressure but the heat emission takes place in the over-critical area.
In comparison to standard refrigeration systems, a CO2 circuit shows radical differences concerning material properties, the components used, the arrangement of the components and the method of control, all of which are described in this chapter.
7
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
2.1 Physical and thermodynamic properties of Carbon Dioxide The best way to look at the thermodynamic properties of CO2 is by a comparison to a standard industrial refrigerant. The comparison is made to R134a, the standard refrigerant used in current automotive AC technology. The following table shows the comparison of the thermodynamic properties between CO2 and R134a.
Property
R134a
CO2
0
0
GWP100, Global Warming Potential for 100 years
1200
1
GWP20, Global Warming Potential for 20 years
3200
1
-
0.1889
102.03
44.01
no
no
2.93
34.85
Density of saturated vapour @ 0°C [kg·m ]
14.44
95.20
Evaporation Temperature @ 1bar [°C]
-26.8
-78.5
Critical Temperature [°C]
101.2
31.1
Critical Pressure [bar]
40.56
73.77
515.3
467.6
Volumetric Evaporation Enthalpy @ 0°C [kJ·m ]
2863
22547
Specific Evaporation Enthalpy @ 0°C [kJ·kg-1]
196.5
230.9
ODP, Ozone Depletion Potential
-1
-1
Specific Gas Constant [kJ·kg ·K ] -1
Molar Mass [g·mol ] Combustibility Pressure @ 0°C -3
-3
Critical Density [kg·m ] -3
Table 1: Comparison between CO2 and R134a Carbon dioxide is a colourless and non-combustible gas. There are several ways to produce CO2: through the combustion of carbon containing materials; as a product of the metabolism from the respiration of an organism; or through the process of fermentation. Carbon dioxide has a higher density than air, and, therefore, it accumulates as an
8
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
underlying layer to air. In dealing with CO2 caution is required, because it restricts respiration and is odourless and, hence, cannot be detected through smell.
Carbon dioxide has a molecular lattice with a face centred cubic atomic arrangement. Both oxygen atoms lie symmetrical to the carbon atom in a straight line. Carbon dioxide has poor electrical conductivity and water solubility characteristics. It possesses a chemically stable connection and, only at very high temperatures, it may start to break up and then only in small portions. It is an inert gas so it should not corrode any of the parts within the circuit.
An elastomer material must not be used for seals in high pressure CO2 applications as, at high pressure, the CO2 molecules diffuse into the material structure of the elastomer. When the pressure is released too quickly, the escaping CO2 molecules cause serious damage to the elastomer’s material structure. This process is called “explosive decompression”.
CO2 as refrigerant has many advantages, especially its ecologically friendly characteristics. As a natural substance it has no Ozone Depletion Potential. The Global Warming Potential of CO2 in comparison to R134a is infinitesimally small. It is possible to produce carbon dioxide as a waste product of industrial processes. If exhaust gas is used for the production of CO2, the effective GWP would approach zero, because the refrigeration system would serve as a buffer storage for CO2.
By using carbon dioxide as an automotive refrigerant, there would be less expenditure on servicing and recycling because it would be possible to release the refrigerant into a well ventilated area without any harm to humans and the environment, this being one reason 9
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
why it so popular in the food industry and other sectors. Using Carbon Dioxide as a refrigerant will cost approximately 50 times less than using conventional R134a [6], as the volumetric refrigeration power of CO2 is eight times higher. Thus, the relatively small volumetric refrigerant flow results in compressors with small cubic capacity and refrigerant lines of small bore.
2.2 The control of a CO2 refrigerant cycle In the previous section describing the compressor it was mentioned that, in a CO2 cycle, the expansion valve cannot be used as in a standard refrigerant cycle. In conventional cycles, the expansion valve adjusts the refrigeration power by controlling the filling of the evaporator with refrigerant. In a CO2 cycle, the expansion device is needed for adjusting the high pressure of the system. The original task of the expansion valve, the adjustment of the refrigeration power, is taken over by the variable displacement compressor.
In order to explain the intention of high pressure control, this section describes the thermodynamic effects of an increased ambient temperature and an increase in high pressure on a CO2 cycle. The section is concluded by considering the different methods for the determination of the ideal high pressure.
2.2.1 Effect of increased ambient temperature To understand the necessity for high pressure control of a CO2 refrigerant system, it is important, first of all, to investigate the effect of increased ambient temperature at a constant interior temperature. This examination is done on a system with a low pressure receiver. The pressure/enthalpy diagram provides the best illustration of the thermodynamic changes in the process. 10
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
If the ambient temperature increases, the refrigerant temperature of the gas cooler outlet rises and the evaporation pressure stays nearly the same. In addition a slight increase in high pressure occurs. This process is illustrated in Figure 1.
Figure 1: Effect of increased ambient temperature in p,h-diagram The cycle changes from process 1-2-3-4 to process 1-2’-3’-4’. The increase in the refrigerant temperature at the gas cooler outlet causes a shift of point 3 towards the right hand side. The shift of the expansion process to the right hand side effects a decrease in usable enthalpy difference and refrigerating effect inside the evaporator. This causes a reduction of refrigeration power and Coefficient of Performance (COP).
11
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
2.2.2 Effect of increased high pressure To offer sufficient refrigeration power, even at high ambient temperatures, two solutions are possible. Firstly, an internal heat exchanger could augment the usable enthalpy difference. Secondly, it is possible to improve the refrigeration power by an increase of high pressure because of the s-shaped isotherms. Figure 2 illustrates the effect of an increase in high pressure at constant refrigerant temperature at gas cooler outlet.
Figure 2: Effect of increased high pressure on the p,h-diagram The cycle changes from process 1-2-3-4 to process 1-2’-3’-4’. As the high pressure is shifted to a higher level, the process of gas cooling (2’-3’) ends on the same isotherm as before but at a much lower enthalpy. After the isenthalpic expansion to the low pressure level, point 4’ is reached. This is the condition of the refrigerant at the evaporator intake. The gain in specific refrigeration power is the difference between the enthalpies at point 4
12
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
and point 4’. In this particular case the gain in refrigeration power, by inspection, is approximately 70%. Besides the increase in refrigeration power, there is also an improvement in COP.
2.2.3 Ideal high pressure The previous section explained why the control of the high pressure is important for the operation of a CO2 refrigerant cycle. This section defines the ideal high pressure, boundary conditions and their determination. 2.2.3.1 The definition of ideal high pressure The improvement of refrigeration power is limited by the greater losses of the compressor at higher pressure. By calculating the theoretical COP as a function of the high pressure, it becomes obvious that at most gas cooler outlet temperatures the COP shows a well defined maximum. Figure 3 [7] illustrates the selection of these curves.
Figure 3: Curves of ideal high pressure at different gas cooler temperatures In an ideal process (isentropic compression, no pressure loss) at e.g. a gas cooler outlet temperature of 35°C, the high pressure which produces the best COP is approximately 13
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
87bar. If the gas cooler outlet temperature increases to 50°C, the optimum high pressure increases to 121bar.
In order to ensure the best economic operation of the cycle, an appropriate control system should follow the curve of the ideal high pressure. Different evaporation pressures produce slightly different curves. 2.2.3.2 The determination of ideal high pressure There are several methods to determine the ideal high pressure for a given process: by measurement, by calculation using data on chemical media, graphic method and by using an empirical equation.
Measurement It is possible to determine the curves in the above figure by measurements on a test bench CO2 cycle. This method delivers the most accurate values because all the influencing factors on the cycle are included. Unfortunately the procedure is very time consuming and the measured values are valid for the given test bench only. This method could be used to determine the ideal characteristic curve for an electronic expansion device.
Calculation (using data for chemical media) The curves can be calculated by using the values from the p,h diagram or with the help of a simulation program. The accuracy depends on the influencing factors which are taken into account.
14
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Graphical method This method was introduced by Inokuty in 1928 [8]. Considering the p,h-diagram, the ideal high pressure is determined by constructing a tangent to the gas cooler outlet isotherm, which should cross the low pressure isobar at 90% vapour content. The point where the tangent contacts the isotherm determines the ideal high pressure. This procedure leads to a result very quickly but it is valid for the theoretical cycle only.
Calculation (using an empirical equation) The optimal high pressure can be calculated by using the equation below. If the isentropic efficiency is considered as constant, the curve representing the ideal high pressure as a function of gas cooler outlet temperature, is nearly linear. This is described by the following equation [9]:
optimal high pressure
pH , opt = (2,778 − 0,0157 ⋅ to ) ⋅ tG 2 + 0,381 ⋅ to − 9,34
Eq. 1
pH, opt
optimal high pressure
to
evaporation temperature (in this case 5°C)
tG2
gas cooler outlet temperature
This formula corresponds to the graphical method. Therefore, it is also valid for the theoretical cycle. The error between the graphical procedure and the formula is less than 1%.
Comparison of the different methods The following table compares the results of the different methods for the determination of the ideal high pressure. The calculation based on the data for chemical media was not taken 15
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
into account, because the accuracy of this method shows a strong dependence on the influencing factors. temperature at gas cooler outlet measurement graphical method use of empirical equation
35°C
40°C
45°C
50°C
86 bar 86 bar 87 bar
99 bar 99 bar 100.5 bar
111 bar 113 bar 114 bar
123 bar 127 bar 127 bar
Table 2: Comparison of several methods to determinate the ideal high pressure
Generally the results are well matched. The differences of the measurement method at higher pressures were caused by the use of an internal heat exchanger. This feature reduces the ideal high pressure.
2.3 The arrangement of the components The previous sections described the unique control characteristics of a CO2 cycle and the importance of the high pressure control. This section describes the different ways of arranging the components of a CO2 refrigerant cycle while simultaneously considering the high pressure control. The explanation of four different circuits and cycles is followed by an assessment of their pros and cons and a further description of the most appropriate cycle.
2.3.1 Possible cycles A CO2 refrigerant circuit offers several possibilities to arrange the system’s components. This section describes four different possibilities with their advantages and disadvantages. 2.3.1.1 System without high pressure control
This system works with so called “dry expansion”, i.e. there is no liquid refrigerant left at the evaporator exit. Like in conventional refrigerant systems the filling of the evaporator is 16
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
controlled to a certain superheat value by a thermostatic expansion valve. Control of the high pressure is not possible. The value of the high pressure depends on the refrigerant temperature at the gas cooler and evaporator, and the refrigerant charge. Figure 4 illustrates the arrangement of the system.
Figure 4: System without high pressure control
Advantages: •
Very simple layout.
•
No need for a liquid separator after the evaporator because of dry expansion.
•
Oil returns to the compressor without problems because no liquid separator is installed.
•
Only one control valve is needed.
Disadvantages: •
Without a receiver there is no buffer storage for the refrigerant. Even a small leakage spoils the performance of the system.
•
High non-operational pressure of the system because of small internal volume (no refrigerant receiver).
17
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
•
Control of the high pressure is not possible, i.e. no COP-optimised operation of the system is possible.
•
High superheat at evaporator, i.e. not the whole surface of the evaporator is used for evaporation of the refrigerant.
2.3.1.2 System with low pressure receiver (Lorentzen principle)
In this system the high pressure is controlled by the expansion valve. The evaporator works under flooded conditions. The amount of unevaporated liquid refrigerant is absorbed by the receiver. If the opening of the expansion valve is reduced, the liquid level inside the receiver drops. As a result, the refrigerant is transferred from the receiver to the high pressure side and creates an increase in pressure. This increase of high pressure is necessary to provide the full refrigeration power at high ambient temperatures. An increase of the valve opening results in higher liquid level inside the receiver and a decrease of the high pressure. As the oil is eliminated from the refrigerant inside the receiver a feature for the oil return is necessary. Figure 5 illustrates the arrangement of the system.
Figure 5: System with low pressure receiver
18
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Advantages: •
Only one control valve is necessary.
•
Lower non-operational pressure of the system (the receiver increases internal volume).
•
The receiver can serve as a storage tank to compensate for leakages.
•
Better performance of the evaporator because there is no superheat.
•
Simple design.
Disadvantages: •
An oil return feature is necessary.
•
The receiver is an additional component (cost).
2.3.1.3 System with a mid-pressure receiver
For the control of this system two valves are necessary. The valve at the inlet of the receiver controls the high pressure. The thermostatic or electronic expansion valve at the receiver’s outlet controls the refrigerant flow to the evaporator. The temporarily unnecessary amount of refrigerant is stored in the receiver. Figure 6 illustrates the arrangement of the system.
Figure 6: System with mid-pressure receiver
19
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Advantages: •
No need for an oil return feature.
•
Insensitive to leakage of refrigerant.
Disadvantages: •
Two control valves are necessary.
•
The receiver has to withstand high pressure.
•
Complicated control strategy.
2.3.1.4 System with refrigerant charge control
In this system the receiver is arranged parallel to the circuit. Valve 1 and Valve 2 control the high pressure inside the gas cooler. If the pressure is too high, control valve 1 opens to reduce the amount of refrigerant in the circuit. If the gas cooler pressure is too low, control valve 2 opens and releases refrigerant to the system. The expansion valve controls the refrigerant flow to the evaporator like in a conventional refrigeration circuit. It is possible to use either a thermostatic or electronic expansion valve. Figure 7 illustrates the arrangement of the system.
Figure 7: System with refrigerant charge control
20
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Advantages: •
No need for a oil return feature.
•
Receiver serves as buffer storage for the refrigerant. Small refrigerant leakage rates can be compensated.
•
Because of the parallel arrangement of the receiver, its position in the system is flexible.
Disadvantages: •
Three control valves are necessary.
•
The receiver has to withstand high pressure.
•
Complicated control strategy.
2.3.2 Assessment To decide which system is the best, the special requirements for automotive duty have to be taken into account.
In contrast to stationary refrigeration plants, mobile refrigeration systems always have a small leakage rate. Therefore, a system which cannot compensate a small loss of refrigerant is not acceptable.
All HVAC systems for vehicles have problems with space limitations, hence, every component has to be very effective. An evaporator in “dry expansion” operation wastes usable heat exchanger surface by using some parts of the evaporator only for superheating and not for evaporation. An evaporator working under flooded conditions is more effective and this is the preferred condition. 21
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Automotive CO2 technology is under pressure regarding cost because it has to compete against the cheaper conventional R134a technology. Therefore, systems with two or more control valves and a complicated control strategy cannot be justified for cost effective serial production. This being a very restrictive limitation, only one other promising system remains, this being the one with a low pressure receiver.
2.3.3 The Lorentzen process Because the system with a low pressure receiver is the most appropriate solution for a CO2 refrigerant cycle inside a vehicle, this section describes it in more detail.
The CO2 process with a suction side receiver for the use in vehicles was introduced in 1990, in a patent specification of Gustav Lorentzen. Figure 8 clearly illustrates the components and the principle of flow of the Lorentzen process.
Figure 8: Illustration of the Lorentzen process
22
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
The Lorentzen process works as follows: •
The compressor sucks superheated vapour at pressure pO from the internal heat exchanger. By absorption of the mechanical power PV from the combustion engine an isentropic compression to the high pressure pC and the temperature tV2 takes place.
•
During the isobaric cooling of the refrigerant inside the gas cooler, there is heat flow to the environment. During this heat transfer, a floating temperature difference will exist.
•
The refrigerant then passes through the internal heat exchanger for additional isobaric cooling where heat is transferred to the low pressure side. The internal heat exchanger increases the specific evaporation power by lowering the enthalpy at the evaporator inlet.
•
At the internal heat exchanger’s outlet the expansion valve throttles the refrigerant to the pressure pO in an isenthalpic process. The refrigerant gas turns into partially saturated vapour and the temperature drops to tO1.
•
By absorption of the heat flow, the two-phase refrigerant is evaporated completely. The heat flow is supplied by the air which is blown through the evaporator.
•
Finally, the refrigerant crosses the internal heat exchanger and “picks up” the heat from the high pressure side through an isobaric process.
2.4 The components In this section the special demands and designs of the components in a vehicle’s air conditioning systems are described when CO2 is used as the refrigerant.
23
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
2.4.1 The gas cooler The refrigerant absorbs heat energy from the environmental air at the evaporator. Additional heat energy is transferred to it by the compressor. Similar to the condensor in a R134a system, the gas cooler has to remove the heat energy from the refrigerant. There is a continuously decreasing temperature difference during the emission of heat, because it takes place above the critical point, where no condensation is possible. This thermodynamic process is the cooling of over-critical gas, thus the name of the heat exchanger used, changes from condensor to gas cooler.
As a result of the given boundary conditions in vehicles, the design of the CO2 components follows existing R134a parts. Like in conventional systems, the gas coolers of vehicles equipped with prototype CO2 systems are still located under the hood in front of the water cooling system. To guarantee the operation of the refrigeration system even at low air velocity, an additional blower has to be mounted at the gas cooler. This blower should be controlled by a temperature sensor in the gas cooler outlet. In several tests the classical tube-and-fin design proved to have good characteristics concerning heat transfer and strength. The disadvantage of this particular design was the large internal volume, which leads to a high refrigerant charge.
The best solution for the gas cooler, as well as for the evaporator, seems to be a design with special extruded flat tube sections containing several parallel capillary tubes; which are called “multi-flow tubes” or “multi-port profiles”. When using these types of tubes, the high pressure drop should be taken into account. The parallel flow and serpentine principles are the two typical designs available and are described in the following sections.
24
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
2.4.1.1 Parallel Flow principle
In this design the refrigerant flows in several parallel multi-port profiles from one collector pipe to another, which are on opposite sides, back and forth though the whole heat exchanger. To achieve this, the collector pipes contain separators. The gaps between the multi-port profiles are filled with aluminium fins to increase the heat exchanger’s surface area. Figure 9 illustrates the parallel flow design and the principle of flow through the heat exchanger, with separators after every eight passages.
Figure 9: Parallel flow design and principle of flow
2.4.1.2 Serpentine principle
In this design, the multi-port profiles follow a serpentine arrangement with wavy fins located between the profiles. By using a multiple pass design, an increased heat flux and reduced pressure drop can be achieved compared to one pass designs. Figure 10 illustrates the design and principle of flow of a multiple pass serpentine heat exchanger.
25
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Figure 10: Serpentine design and principle of flow
2.4.2 The evaporator The evaporator transfers heat energy from the inlet air to the refrigerant. Additionally, dehumidification of the air takes place with the condensation of vapour in the air on the evaporator’s surface. With the transfer of heat, the refrigerant evaporates and is superheated (depending on the setting of the expansion device). Under ideal conditions, i.e. no pressure drop in the evaporator, evaporation occurs as an isobaric process. Designs with tube-and-fin or with serpentine multi-port profiles are available.
An ideal design for a multiflow-tube evaporator is a two-pass construction. The tubes run parallel and side-by-side in order to attain a crossflow-counterflow heat transfer. The tubes with the greater amount of liquid refrigerant (low dryness fraction) run on the air outlet side, while the tubes with less liquid refrigerant (high dryness fraction) are on the air intake side. This construction makes it possible to achieve a constant temperature drop.
26
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
2.4.3 The internal heat exchanger The internal heat exchanger transfers the heat energy from the hot refrigerant, on the high pressure side following the gas cooler outlet, to the cold vapour exiting the evaporator outlet. Thus, the hot gas is cooled further and the useable enthalpy difference in the evaporator is increased thereby increasing the refrigerating effect. At the same time the heat transfer results in a higher superheating on the low pressure side. The necessity of an internal heat exchanger at high environmental temperatures in particular, has been proved in several tests [10]. The following section summarises and discusses the advantages and disadvantages of using an internal heat exchanger. 2.4.3.1 Comparison of advantages and disadvantages of using an internal heat exchanger
Advantages: •
Inside the internal heat exchanger surplus liquid refrigerant can be evaporated. This serves as an additional protection for the compressor because it helps to avoid the destructive consequences of compressing liquids.
•
At high environmental temperatures the internal heat exchanger increases the efficiency of the refrigeration system by increasing the useable enthalpy difference.
•
Lower refrigerant mass flow because of higher efficiency.
Disadvantages: •
The increase in the suction temperature of the compressor leads to higher temperatures after compression and, hence, the stress on the material is higher.
27
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
•
As an additional component the internal heat exchanger increases the internal volume of the system (a greater amount of refrigerant is required for charging) and increases the pressure drop.
•
The higher temperature at the compressor inlet decreases the density of the refrigerant and, therefore, at the same volumetric flow, the mass flow rate is lower.
In conclusion, the refrigeration power of a system with an internal heat exchanger is greater than in a system without one, because the higher refrigerant mass flow (in a system without the heat exchanger) cannot compensate for the increase of usable enthalpy difference and refrigerating effect. By increasing the high pressure in a system without an internal heat exchanger it is theoretically possible to reach the same refrigeration power as in a process with this additional component, but there are obvious limitations on pressure based on material properties. 2.4.3.2 Internal heat exchanger designs
As an additional component in the engine compartment, and, because of the restricted space, a very compact design is needed. One possibility would be a design that is shaped in such a way that it is integrated into the existing components. A co-axial design could be integrated into a refrigerant line, and a serpentine design into the accumulator.
The co-axial design: This type of internal heat exchanger design can easily be fitted along the radiator. It is a simple construction made of extruded profile tubes. The operational principle is counterflow heat transfer between high and low pressure refrigerant flowing through different
28
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
channels of the same extruded pipe. The heat transfer takes place by thermal conduction inside the material of the tube.
A disadvantage of this design is the tube length needed for the exchanger’s surface, although the high flow rates result in a good heat transfer performance. The small internal diameters of the tubes are ideal for withstanding high pressures. Figure 11 illustrates a coaxial serpentine design for an internal heat exchanger.
Figure 11: A typical co-axial serpentine design for an internal heat exchanger
Serpentine design: In this design the high pressure pipe is spiral wound and located in a pressure tank through which low pressure refrigerant flows. This type is particularly suitable for systems with low pressure accumulators. If the pressure tank of the heat exchanger is correctly sized, it can work as an accumulator and, hence, an additional component can be saved.
Under good conditions (high filling level of the pressure tank) this design provides good heat transfer performance but the unstable level of liquid inside the tank can be a disadvantage because the heat transfer of refrigerant in the vapour phase is lower than that
29
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
for liquid. The structure of this internal heat exchanger is more complicated than the coaxial design and will cost more to produce. Figure 12 illustrates an internal heat exchanger with a serpentine design.
Figure 12: Internal heat exchanger with a serpentine design
2.4.4 The expansion valve The purpose of the expansion valve is to decrease the pressure from the gas cooler outlet to a pressure level appropriate for evaporation. The throttling is an isenthalpic process. The difference of the expansion valve used in a transcritical CO2 process to that of one used in normal compression type refrigeration systems, is that it has the task of controlling the high pressure. To achieve economic operation of the process the high pressure has to be adjusted according to the environmental temperature. Suitable valves for this application are pulse-width-modulated (PWM) solenoid valves or stepping motor valves.
30
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
2.4.4.1 Solenoid valves
The refrigerant flow is controlled by the periodic opening and closing of the valve. The time ratio between opened and closed state determines the effective opening area of the valve. An advantage of the PWM valves is their simple construction. The continuous rapid changes between opened and closed state cause pulsation of the fluid, which leads to additional material stress and high noise emission. 2.4.4.2 Stepping motor valves
This type of valve uses a stepping motor to drive the spindle of a needle valve. The rotating motion of the motor is converted into linear motion, which the needle follows. Stepping motor valves offer an accurate and infinitely variable control characteristic.
2.4.5 The accumulator The accumulator is fitted at the low pressure side between the evaporator and the internal heat exchanger. According to the operation conditions, the refrigerant in a compression type refrigeration processes is continuously shifted back and forth from the high pressure to low pressure side. The accumulator is needed as a buffer reservoir to compensate for surplus or lack of refrigerant while the operating conditions are changing. An additional function of the accumulator could be that it accommodates a refrigerant dryer and acts as a damper for pulsation caused by the compressor. This pulsating pressure can cause additional stress in the material and acoustic problems inside the passenger compartment. The size of the accumulator should be such that it is never filled completely with refrigerant, hence the volumetric capacity of the accumulator should be 110% of the refrigerant volume [11]. The design pressure of the accumulator is determined by the nonoperational pressure. The working pressure of the accumulator is approximately 30bar
31
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
[11], but once the compressor is stopped, the different pressures in the system will equalise, and depend only on the system volume and the ambient temperature. Typical pressures of 100bar and higher may be reached at equalisation [11].
In a CO2 refrigeration process the expansion valve controls the high pressure, but not simultaneously the filling of the evaporator. As a result of this, it is possible for liquid refrigerant to reach the compressor and cause serious damage. Therefore, the accumulator also functions as a liquid/gas separator and serves to protect the compressor.
Further, the accumulator increases the internal volume of the system and, thus, the stagnation pressure, especially at high ambient temperatures, can be reduced. Without this additional volume the stagnation pressure of the whole system could exceed the pressure found under working conditions [11].
2.4.6 The compressor The function of the compressor is to move refrigerant from a lower pressure state to a higher pressure state. In most modern automotive refrigerant systems the compressor has the additional task of controlling the refrigerant mass flow. This makes it possible to adjust the refrigeration power according to specific air-conditioning requirements.
Carbon dioxide was used as a refrigerant during the early days of refrigeration technology. The main areas of application were the food industry and on-board refrigeration on ships. The compressors that were used were stationary types and were state of the art at that time.
32
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
This technology cannot be used in today’s mobile refrigeration systems, as new developments for CO2 compressors have to meet the standards set by current R134acompressors. An increase in the mass or volume of the compressor is not acceptable. Despite the fact that the CO2 compressor has to withstand higher pressures than the current R134a-compressor, it is possible to reduce the latter’s dimensions. The advantage of CO2 is its very high volumetric refrigerant power. A fifth of the cubic capacity of a standard R134a compressor is sufficient for a CO2 compressor (approximately 30cm3).
A CO2 compressor should be able to provide variable capacity. Due to the special control characteristics of a CO2 cycle the expansion valve cannot be used to control the refrigeration power of the system, therefore, a variation of the compressor capacity is the only way to adjust the refrigeration power. An axial piston compressor with a variable wobble plate offers good volumetric control characteristics. Other advantages of this compressor type are the compact layout and a delivery of flow with less pulsation and as a result, nearly all prototype CO2 compressors are axial piston types.
The variable axial piston compressors are well known for their present application in R134a technology. This type allows the variation of the piston stroke by changing the angle of the wobble plate, thus, only the bottom dead centre changes, the top dead centre is fixed, therefore, the wasted space of the compressor is not influenced by the variation of the stroke. The angle of the wobble plate depends on the difference between high pressure and crank case pressure (pressure conditions at the base and top of the piston). In an internally controlled compressor an integrated valve controls the crank case pressure in order to achieve a certain difference between high and low pressure. An externally controlled compressor uses the same principle but the control valve can be adjusted by an 33
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
external signal. Due to the control characteristics of CO2, the compressors used should be externally controlled.
To illustrate a few difficulties experienced when using other compressor types for mobile CO2 refrigeration: •
Trunk piston compressors require more installation space and are too heavy for automotive use.
•
Rotary piston compressors are not suitable for high pressures because of sealing problems between the working chambers.
A critical point of a CO2 compressor is between the connecting rod and piston. The pistons of a CO2 compressor are very small, due to its small cubic capacity. The connecting rods have to transfer great forces because of the high pressure differences in a CO2 system. This leads to high surface pressures inside the bearing. The lubrication at this point has to be greatly improved over that of a standard R134a-compressor.
Another problem of a CO2 compressor is the changing solubility of oil. At the high pressure side the oil is dissolved by the refrigerant and transported into the system. After the expansion to a lower pressure level the oil is deposited and it is impossible to get it back into the compressor and the compressor experiences a continuous loss of oil. To prevent this problem, an effective system for oil separation after the compression process is needed. For many prototype compressors an external oil separator is used. Future serial CO2 compressors will have to be designed with an internal oil separation system.
34
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
The current solutions for the shaft seals of R134a compressors are not suitable for CO2 compressors. The seal has to withstand the pressure difference between crank case and ambient (up to 50 bar). Furthermore, the material used has to resist the diffusion of CO2. The conventional rotary shaft lip seals are unsuitable for the CO2 technology. Today’s prototypes use floating ring seals. A major factor in the design of this kind of seal is the correct choice of the material. According to information from Burgmann Automotive Seals GmbH, lubrication difficulties result in heat generation of up to 100W at the floating ring seal. Therefore the sealing material should be capable of withstanding high levels of thermal stress.
2.4.7 Tubes and hoses The use of tubes throughout the system would be ideal for the refrigerant circuit, but hoses are essential for a refrigerant system inside a car. It makes allowances for engine movements (e.g. relative motion between compressor and gas cooler), road excitations and mounting tolerances in the piping system.
Due to the special characteristics of CO2 at high pressure, the flexible polymer hoses currently used in R134a air-conditioning systems cannot meet the requirements for safe operation of CO2 systems. The permeation rate of polymer materials for CO2, and the phenomenon of explosive decompression, require flexible hose materials that build a barrier for CO2 molecules and, therefore, prevent leakage and absorption of CO2. Additionally, these materials have to withstand operating conditions of 180°C and 140 bar [12]. The materials must be chemically resistant against the fluids and lubricants present, inside and outside of the system.
35
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
The basis of a CO2 hose is a corrugated pipe made of stainless steel, which acts as a barrier layer against the CO2 molecules. To achieve a high pressure resistance, the corrugated pipe is equipped with an outer braiding that prevents lengthening and widening of the hose. Due to pressure pulsation caused by the refrigerant compressor, mechanical wear takes place between the hose and the braiding. The braiding would grind through the corrugations and this would lead to leaks. To combat this, an additional layer of silicon coating between the hose and the stainless steel braiding is used. A secondary effect of this additional layer is the damping of the resonance frequencies of the hose.
An important difference between metal hoses and polymer hoses is the ageing or fatigue of the material. Pressure pulsation and vibrations are detrimental to metal hoses. These problems are uncommon to polymer hoses currently in use. Another difference is that of the bending radius: the minimum bending radius of metal hoses is larger than that of polymer hoses. The integration of metal hoses in the piping system cannot be carried out with press-fittings, and, hence, these hoses are joined by means of welding or soldering. Besides the fact that the materials are more costly for metal hoses, this extravagant production method also greatly increases costs above that of the R134a hoses currently used.
However, the small volumetric flow in a CO2 refrigerant system enables small diameter hoses and tubes to be used. On the high pressure side, internal diameters range between 4 and 6mm and at the low pressure side between 6 and 8mm. These dimensions are one third of the tube and hose diameters that are commonly fitted in R134a systems. This makes the installation of the tubing inside the vehicle much easier. Also, due to the small internal diameters, a wall thickness of 1mm is sufficient for tubes of a CO2 refrigerant system. 36
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
Steel or aluminium are materials that can be considered for use as tubing and, due to cost considerations, aluminium comes out as the best choice [12].
2.5 Summary This chapter gave an overview about the special features of CO2 as refrigerant for automotive AC systems and helps to understand the design of the test bench which is described in the following chapter.
The Lorenzen process came out as the most reasonable cycle for automotive use. As this cycle uses the principle of wet evaporation, an accumulator has to be located after the evaporator in order to protect the compressor from drawing in liquid refrigerant. The internal volume of the accumulator has to be calculated very carefully because of its additional task of decreasing the stagnation pressure of the system.
Concerning the heat exchangers (gas cooler and evaporator) in the system the similar principles of design and materials like in conventional R134a cycles are possible. The major difference is the much smaller cross-sectional area and the increased wall thickness of the tubes used for the manufacturing of the CO2 heat exchangers.
In addition to the conventional heat exchangers, an additional internal heat exchanger is needed for the efficient operation of a CO2 air-conditioning system at high ambient temperatures. The design of the internal heat exchanger complies with the available space in this particular application.
37
Chapter 2 Special features and requirements when using Carbon Dioxide as a refrigerant
A CO2 compressor is similar to a conventional compressor; both are externally controlled axial piston types with variable capacity. Due to the volumetric refrigeration power of CO2 a fifth of the cubic capacity of a standard compressor is sufficient for a CO2 compressor.
Investigating the thermodynamic characteristics of CO2, the low critical point catches the eye and explains why the CO2 refrigeration cycle at high ambient temperatures is a transcritical process with the heat emission in the overcritical area. The s-bend characteristic of the isotherms in the overcritical area of the p,h-diagram is the reason for the high pressure control strategy in CO2 cycle.
Conventional thermostatic expansion valves cannot be used for CO2 systems because for high pressure control strategy, externally controlled valves are needed. As the expansion valve is used for the high pressure control, its original task of controlling the refrigeration power is replaced by the compressor control valve. By adjusting the cubic capacity of the compressor the refrigeration power is changed.
38
Chapter 3 The test bench
Chapter 3 The test bench This chapter describes the layout of the test bench, the components, materials and equipment used to construct the combined refrigeration and heat pump circuit.
Firstly, the refrigerant circuit with its three sub-circuits: heat exchanger circuit, oil return circuit and compressor control circuit are described. Secondly, the measuring system including hardware, software and sensors are described.
3.1 The refrigerant circuit All components and tubing were mounted to a space frame, made of aluminium (material number: 3.3206.72) profiles. This lightweight and stable part was manufactured by Item®. They also offer a wide range of connectors and fixtures suitable for this profile. A square profile in the dimensions 40x40mm was chosen. The following picture shows a crosssectional view of the profile.
39
Chapter 3 The test bench
Figure 13: Cross-section view of Item® profile
The refrigerant tubing was made from seamless stainless steel (material number: 1.4571) tubes which have dimensions of 8x1mm and 6x1mm respectively. The 8mm tube was rated to a working pressure of 320bar and the 6mm tube was rated to a pressure of 430bar. The tubes were connected by double clamping ring fittings manufactured by Swagelok®. Figure 14 illustrates the design of this component.
Figure 14: Double clamping ring fitting
In this design, the back ring keeps the front clamping ring in position as the fitting is tightened. All the parts were constructed from stainless steel (material number: 1.4571). The 6mm fittings were rated to a pressure of 450bar and the 8mm fittings to 360bar.
40
Chapter 3 The test bench
Similary, all the valves and gaskets used in the circuit were made from stainless steel (material number 1.4571). The valves were connected to the tubes with double clamping ring connections. As mentioned before, elastomer seals are not suitable for CO2 because of the problem with explosive decompression.
Figure 15 illustrates the entire circuit in air-conditioning (AC) mode, where the arrows indicate the flow direction. The heat exchanger circuit is indicated in red, the oil return circuit in blue and the compressor control circuit in green.
41
Chapter 3 The test bench
Figure 15: Illustration of the whole circuit in AC mode
42
Chapter 3 The test bench
3.1.1 The heat exchanger circuit The compressor draws in low pressure vapour and raises its pressure and temperature state. This component was a prototype and was made available by Zexel/Valeo Compressor Europe® GmbH and was an axial piston compressor with a variable wobble plate. The five cylinders provide an overall displacement of 22 cm3. All of the compressor parts were made of steel. Figure 16 shows the compressor and fittings.
Figure 16: The CO2 compressor
A bursting disc was located immediately after the discharge port of the compressor and was a requirement of the compressor manufacturer to protect the compressor in the event of overload or malfunction. The bursting disc was designed to rupture at 180 bar and 160°C.
The next component in the high pressure line was the oil supply device, which consisted of a stainless steel (material number: 1.4571) needle valve and a transparent container with a tube extending from the top opening to the bottom. If the circuit was evacuated, oil could
43
Chapter 3 The test bench
be drawn into the system by opening the valve. The compressor manufacturer calculated the amount of oil needed in the system to be 200g.
Following the oil supply vessel, the refrigerant passes through the oil separator (a detailed description is provided in section 3.1.2) and reaches the four-way valve. This device was used to switch between AC operation and heat pump (HP) operation. In AC operation the valve connects, on the one hand, the oil separator outlet with the gas cooler inlet, and, on the other hand, the evaporator outlet with the low pressure inlet of the internal heat exchanger. In heat pump operation the valve connects, on the one hand, the oil separator outlet with the evaporator, and, on the other hand, the gas cooler with the low pressure inlet of the internal heat exchanger.
The four-way valve was very practical, as the four refrigerant lines can be connected or disconnected with one adjustment. On the other hand the valve allows heat exchange between the hot and the cold refrigerant flow which leads to a loss of refrigeration power. Figure 17 shows the circuit and the four-way valve in heat pump mode.
44
Chapter 3 The test bench
Figure 17: Illustration of the whole circuit in HP mode
45
Chapter 3 The test bench
If the four-way valve is in the AC position, the refrigerant flows from the oil separator to the gas cooler. This heat exchanger was also a prototype and was made available by Valeo Klimasysteme® GmbH.
The gas cooler was a parallel flow type and was of aluminium construction. The passing tubes were flat-type multi-port profiles (dimension: 17mm x 2mm) and the collectors were made from thick-walled round aluminium. The heat exchanger contained three parallel flow passes, with each pass consisting of 10 multi-port profiles. Now, with reference to Figure 18, the inlet was located at the top left hand side and the outlet at the bottom right hand side. The outer dimensions of the gas cooler were 590mm x 370mm x 24mm, so as to correspond to the dimensions of the original condensor of the Mercedes A-class®, which made it possible to mount the gas cooler to the original front unit. This mounting position allowed the use of the radiator fan which provided sufficient air flow through the gas cooler.
Figure 18: The gas cooler
46
Chapter 3 The test bench
The expansion valve for heat pump mode was located at the outlet of the gas cooler and was a stainless steel (material number 1.4571) needle valve, and during the AC operation the valve was bypassed by a non-return valve.
The non-return valve connected the outlet of the gas cooler and the high pressure inlet of the internal heat exchanger and was a serpentine design with a spiral wound high pressure pipe which was located in a low pressure tank. As seen in Figure 19, the tank was a horizontally arranged cylinder, made of thick-walled aluminium. The internal volume of the cylinder was sized in such a way, that the internal heat exchanger could work as a low pressure accumulator. This prototype was also provided by Valeo Klimasysteme® GmbH.
Figure 19: The internal heat exchanger
47
Chapter 3 The test bench
An analogue manometer was located between the high pressure outlet of the internal heat exchanger and the expansion valve for the AC operation, and was used during the charging of the system to monitor the increase in pressure. The expansion valve for AC operation was similar to the one for heat pump operation, with a non-return valve in a bypass line which was used during heat pump operation to bypass the needle valve.
After throttling through the expansion valve, the low pressure refrigerant flowed to the evaporator to be evaporated. This prototype heat exchanger was located inside the
original HVAC box of the Mercedes A-Class® and was provided by Valeo Klimasysteme® GmbH. To be able to fit the evaporator into the box, the dimensions had to follow the size of the original part. The outside dimensioning of the all-aluminium CO2 evaporator was 230mm x 200mm x 65mm. As in the gas cooler, it was also a parallel flow design, made of multi-port profiles of dimensions: 60mm x 3mm.
The collectors, made from large aluminium blocks, were located on top and bottom of the heat exchanger. The top collector had only one separator. In contrast to the design of the gas cooler, this separator was not arranged crosswise, but longitudinally to the collector. The bottom collector had no separator at all. As the refrigerant enters the evaporator, it fills the intake section of the top collector, while the outlet section is isolated by the separator. The refrigerant flows through the first half of the 21 multi-port profiles. At the bottom of the heat exchanger the collector reverses the flow direction of the refrigerant and it streams through the other half of the multi-port profiles to the outlet section of the top collector. This is illustrated in Figure 20.
48
Chapter 3 The test bench
Figure 20: Side view of the parallel-flow evaporator
The method of directing the refrigerant entering the air outlet side of the evaporator and the returning refrigerant to the air intake side, allowed the best possible use of the available temperature difference. The evaporator was manufactured and provided by Valeo Klimasysteme® GmbH and is shown in Figure 21. The wire screen carries the 16 thermocouples for the measurement grid located behind the air outlet.
Figure 21: The evaporator
49
Chapter 3 The test bench
Following the evaporator, the tubing leads to the four-way valve. This connection line included a stub line, where a needle valve was fitted. This valve was used to release the refrigerant and to connect the charging cylinder or vacuum pump.
In AC operation the four-way valve connects the evaporator outlet and the low pressure inlet of the internal heat exchanger. On leaving the low pressure outlet of the internal heat exchanger, the refrigerant flows to the mass flow meter. A ball valve was located in a bypass line to the flow meter. Normally the valve is closed and the refrigerant flow passes through the mass flow meter. If a test run without the pressure drop inside the flow meter was performed with no need for measuring the mass flow, the valve could be opened and the refrigerant flow would bypass the measuring instrument. From the mass flow meter the refrigerant line meets with the bypass line and runs to the suction port of the compressor.
3.1.2 The oil return circuit The oil return circuit also starts at the compressor. From the discharge port of the compressor, the gas, which contains refrigerant and dissolved oil, passes the bursting disc and the oil supply device to reach the oil separator. This prototype component was provided by Valeo Klimasysteme® GmbH. The oil separator was made of steel and was designed as a cylinder with its top end closed by a sealed and bolted cover. The refrigerant enters the separator through the intake port, which was located on the side wall of the cylinder. When the gas fills the cylinder, the heavier elements (oil) sink to the bottom and the lighter (refrigerant) rises to the top. While the refrigerant gas leaves the separator through the outlet at the top of the cover, the oil is collected in the bottom of the cylinder and leaves the cylinder through a port in the bottom. Figure 22 shows the flow through the oil separator.
50
Chapter 3 The test bench
Figure 22: The principle of the oil separator
This type of oil separator could only be a solution for a prototype compressor as, a serial type had to have an integrated oil separator. Figure 23 shows the oil separator as it was installed on the test bench.
Figure 23: The oil separator
51
Chapter 3 The test bench
On leaving the oil separator, the oil flows to an empty filter housing where filter elements made of sintered metal in the range of 0,5 - 440µm pore size could be fitted. This filter was never used during the tests as, it was planned as a backup solution in the event that the 40µm-filter was blocked too soon.
The outlet of the filter housing was connected to a spring loaded relief valve. The spring pushes on the spindle, which was connected to the cone. If the force on the bottom of the cone resulting from the pressure in the system is lower than the force exerted by the spring, the valve remains closed. If however, the system pressure reaches the adjusted opening pressure, the resulting force is greater than the spring load and the cone is lifted. Thus, the valve is opened and the oil is released. The opening pressure could be adjusted by using different removable springs. Eight different springs cover an opening pressure range from 3,4 to 413bar. The fine adjustment of the opening pressure could be done by tightening or loosening the cap of the spring housing, which varies the pre-load on the spring. In contrast to the bursting disc, the relief valve is a reversible safety device.
Another filter housing was located after the relief valve and was fitted with a sintered metal filter with a pore size of 40µm. The tests revealed very little contamination of the filter and, hence, the upstream filter located between the oil separator and the relief valve for pre-filtering was not required.
From the filter, the oil flows to the fine metering valve which was a special type of needle valve. To enable very fine adjustment of the flow through the valve, there was a large ratio between the angular movement of the spindle and the needle movement. The fine metering valve was needed to adjust the amount of oil flowing back to the compressor. A lack of oil 52
Chapter 3 The test bench
would damage the compressor, but, conversely, a surplus of oil would adversely effect the refrigeration power. The compressor manufacturer suggested a mass flow between 1500 and 3000g·h-1. Again this method of adjusting the oil return flow was a prototype solution, as in a serial compressor the oil return line would be integrated and a defined opening would adjust the flow.
The outlet of the fine metering valve was connected to a coriolis-type mass flow meter which provides the measurement for the adjustment of the fine metering valve.
From the mass flow meter, the oil flows back to the compressor through the oil return port from where it is distributed to the required parts of the compressor.
3.1.3 The compressor control circuit In CO2 applications, axial piston compressors with variable displacement are used. The variation of the displacement results from a change in the stroke of the pistons, which are driven by a wobble plate. The angle of this wobble plate can be adjusted. The displacement depends on the angle “α” of the wobble plate from the radial direction as illustrated in Figure 24.
Figure 24: The wobble plate design principle
53
Chapter 3 The test bench
Basically, the angle of the wobble plate results from a balance of forces caused by the difference in pressures inside the compressor. Figure 25 illustrates how the wobble plate was integrated into the compressor and indicates the different pressures.
Figure 25: The wobble plate integrated into the compressor
As the resultant force is directing at the pivot point, the high pressure has no effect on the balance of forces as it only depends on the ratio of crank case pressure, low pressure and spring force. Figure 26 illustrates all the forces which influence the angle of the wobble plate.
Figure 26: The forces at the wobble plate
54
Chapter 3 The test bench
The crank case pressure, which effects the bottom end of the piston, is used as the controlling factor. There is a defined “leakage” from the high pressure side of the compressor to the crank case. The crank case pressure can be adjusted by a defined opening to the low pressure side, which is done by the compressor control valve which connects the crank case port of the compressor to the low pressure side. The control valve is bypassed by an overflow valve. If the control valve is completely closed the crank case pressure rises because of the continuous high pressure blow-by. At a certain level the overflow valve opens and releases the pressure. [13]
As the compressor is driven by a combustion engine, it has to compensate very quickly to changes in input speed. In order to achieve a quick control characteristic of the compressor, it is also necessary to change the pressure inside the crank case rapidly. Following a decrease of the crank case pressure via the compressor control valve, it takes some time until the high pressure blow-by increases the pressure again. Thus, a serial type compressor also needs a high pressure connection to the crank case. Furthermore, the control device should be integrated into a serial type compressor.
3.2 The measuring system The measuring system contains four separate parts; namely the software, computer, signal conditioning unit and sensors. The sensors convert physical values to electric signals. They are transmitted to a signal conditioning unit, which changes the analogue signals to digital signals. By using the LPT(Line Printer Terminal)-interface also known as a Parallel Port, the digital data is transferred from the signal conditioning unit to the personal computer.
55
Chapter 3 The test bench
The measuring software, installed on the computer enabled the user to visualise and analyse the delivered data in real time.
3.2.1 The computer For the measuring system a standard personal computer with Mircosoft® Windows 2000 system software was used. The computer was equipped with an Intel® Pentium III central processing unit.
3.2.2 The signal conditioning unit A signal conditioning unit, or a so called “LPT measuring system” manufactured by the company “Institut für explorative Datenanalyse GmbH” was used. The LPT measuring system contained two 19 inch housings connected by a 20 pin ribbon contact with 96 analogue input channels available, the first 32 of which were voltage inputs (K0 – K31) and the others (K32 – K95) were thermocouple inputs. The system was connected to the measuring computer by using the standard printer port (LPT port). As far as the power supply was concerned, the LPT measuring system accepts direct current voltage between 10V and 30V.
3.2.3 The sensors This section describes the measuring principle and the properties of the sensors used. The system contained more sensors than were needed for the Flash Fogging investigation to facilitate further testing outside the scope of this research. 3.2.3.1 The thermocouples
Two wires from different materials are connected at one end, e.g. by welding, soldering or crimping. The connected end is exposed to the medium investigated temperature, “T2”. 56
Chapter 3 The test bench
The open ends of the wires are connected to the data acquisition system. This “reference junction” is exposed to the ambient temperature “T0”. The combined wires A and B are called a “thermocouple”. The output voltage “U” of a thermocouple depends on the transfer coefficient “kAB” of the different materials A and B and the temperature difference T2 - T0. output voltage
U = k AB ⋅ (T2 − T0 ) U
output voltage (V)
kAB
transfer coefficient (V·K-1)
(T2 - T0)
temperature difference (K)
Eq. 2
It is obvious from the equation that thermocouples do not measure a certain temperature but the temperature difference between the hot connector at temperature “T2“, and the cold connector at temperature “T0”.
For this application type K thermocouples were used, i.e. one wire is made from nickel, the other one from a nickel/chrome alloy. Type K thermocouples have a resolution of 40µV/°C and in the range between –200°C and –0,1°C they have an accuracy of ±0,8°C and between 0°C and 999,9°C, an accuracy of ±0,4°C. [14]
3.2.3.2 The pressure transducers
The pressure transducers used were manufactured by Druck®. The pressure transducers type PMP 1400 and the type PMP 317-2489 were used. Both models have a stainless steel body and silicon pressure diaphragm. The design of this diaphragm is based on semiconductor technology. Tiny resistor lines are etched on the diaphragm’s surface having the same function as a strain gauge. They convert the mechanical deformation of the
57
Chapter 3 The test bench
diaphragm into a change in resistance. The sensor also contains integral electronics which convert the signal from the silicon diaphragm to a three wire voltage output which is proportional to the applied pressure. In addition, the electronics provide power supply regulation. The accuracy of this pressure transducers are ±0,15% of the measured pressure value. 3.2.3.3 The mass flow meter
The measuring equipment contains two mass flow meters, one for the refrigerant mass flow and the other for the oil return mass flow. Both are manufactured by Danfoss®. A mass flow meter system consists of two separate parts, the measuring tube and the control unit. The tube is part of the circuit through which the fluid flows that is to be measured. For the CO2 mass flow, the tube “Mass 2100 DN6” was used. The oil return mass flow was measured by the tube “Mass 2100 DN1,5”.
The control unit regulates the oscillation of the tube and displays the measured value and converts it to a 4 – 20mA output signal that can be fed to the LPT-measuring system. For the CO2 mass flow, control unit “Mass 3000” was used. The tube inside the oil return circuit was connected to control unit “Mass 6000”. A custom made hardware chip, connected to the control unit contains the information about the connected tube and provides the physical data of the measured fluid.
Both sensors use the Coriolis principle. The basic element of a Coriolis type mass flow meter is a straight tube with rigid supports at both ends. It oscillates at its fundamental eigen frequency, controlled by a digital signal processor and is stimulated via two drivers.
58
Chapter 3 The test bench
In the ideal case, at zero mass flow, all parts of the measuring pipe will vibrate synchronously.
When the mass flow enters the pipe, Coriolis forces are caused by two orthogonal velocities, one representing the fluid’s velocity and the other the velocity of the pipe. These Coriolis forces acting on the pipe’s wall are in opposite directions in the upstream and downstream side. In the resultant oscillation, the upstream side will lag and the downstream side will lead with respect to the pipe’s centre, so when the flow enters the pipe, the oscillation will no longer be synchronous.
The Coriolis mass flow meter can thus be viewed as an instrument that measures a kind of “non symmetry”. The delay in time between the velocities of two given points along the length of the measuring pipe is due to this asymmetry and it is proportional to the mass flow rate. [15] 3.2.3.4 Torque sensor
A torque meter was positioned between the shaft of the electric motor and the pulley of the compressor. The torque meter provided measured data of the torque, which was used by the compressor.
There are essentially two different methods of measuring torque, i.e. by angle and stress measuring methods. In contrast to stress measuring methods, angle measuring methods require a particular length of torsion shaft over which the torsion angle can be read off. Due to the limited space inside the compressor drive box, the compact stress measuring method was selected. This method is based on the strain gauge principle. A strain gauge is
59
Chapter 3 The test bench
a displacement transducer which converts a mechanical deformation into the form of a change in electrical resistance [14].
The sensor used was a so called torque flange, model T10F, manufactured by Hottinger Baldwin Meßtechnik® GmbH. It consists of two separate parts, a rotor and a stator where the rotor contains the adapter flange and the measuring body. The strain gauge is fixed to the measuring body. The electronics for the transmission of supply voltage and measuring signal are located inside the flange. The measuring body carries the transmission coils for the contactless transmission of supply voltage and measuring signal on the outer circumference. The signals are received and transmitted through a ring antenna which is mounted on the stator case. It contains the electronics for the modulation of the supply voltage and the processing of the measured data. Figure 27 illustrates the individual parts the torque sensor.
Figure 27: The torque sensor
60
Chapter 3 The test bench
3.2.3.5 The humidity sensor
The air humidity sensor used was manufactured by Mawi Therm® GmbH and has a measuring range of 0 – 100% relative humidity (RH) and in a temperature range of 10°C to 40°C it has an accuracy of ± 2% RH. Relative humidity is defined as the actual moisture content of the air relative to the total amount of moisture the air can hold at the same temperature.
In addition to the humidity sensor, the instrument also contains a resistance thermometer which is a PT-100 measuring device with a range of –30°C to +70°C and accuracy of ± 0,2K. Both, the humidity measuring device and the resistance thermometer, have an output signal of 0V to 10V. The sensor is positioned at the air outlet of the HVAC box.
The principle of humidity measurement is based on the capacitive principle. This is the most common principle for electronic measurement of humidity. This principle is based on the different dielectric constants of water (ε ≈ 81) and dry air (ε ≈ 1). By placing a hygroscopic dielectric material between two electrodes, the relative humidity of the air becomes a function of the changing capacitance. The dielectric should be able to pick up humidity from the air and release it again very quickly. Common materials used are a polyamide-film or a porous layer made of aluminium oxide (Al2O3). The sensor in this application uses a polyamide film as the dielectric and together with the electrodes is located on a ceramic carrier material.
3.2.4 The software The DIAdem® measuring software is popular in the German automotive industry. One major advantage of DIAdem® in comparison to other systems like LabView® is its user 61
Chapter 3 The test bench
friendliness. DIAdem® is not as flexible as other competitors concerning additional control tasks, but for pure measuring tasks it is a very good choice because it allows the user to get the needed measured data with minimum effort.
DIAdem® was used for visualisation, recording and analysis of the data, which was sent through the LPT-port of the personnel computer by the LPT-measuring system. DIAdem® is able to handle large data sets of more than a billion data points in up to 65000 columns and is divided into several modules. This section describes how the different modules of DIAdem® like DAC, VIEW, DATA and GRAPH were used to build a measuring system. 3.2.4.1 The DAC module
With help of the DAC (Data Acquisition and Control) Module the user can describe the measurement and control tasks of the particular application and is a visual type of programming. The user chooses function blocks and connects them via data lines and both are parametrizable. The connections between the blocks are data bus lines where several signals can be transferred on one line. There are two different types of lines, namely, yellow lines which convey a measurement frequency and the green lines contain the measured values.
Input
The frequency of the measurement is defined in the first function block. The LPT measuring system is able to process 1000 data samples per second (1kHz). In order to find the highest possible measuring frequency of a particular application the total frequency of 1000Hz has to be divided by the number of channels used. In this case a maximum frequency of 1000Hz/55 = 18Hz was possible.
62
Chapter 3 The test bench
As the measured signals are in the microvolt range, they have to be amplified but the interferences are also amplified. Interference effects were reduced by using the “Mittelung” software module. This function box produces an average value over an adjustable interval.
The signal of the measurement frequency and the information about the interval for the average values are transferred to the incoming signal block which is called “Messeingang”. This function block contains the driver for the connected measuring system which, in this case, is the LPT measuring system from IED®. Figure 28 shows the three function blocks which define the input of data.
Figure 28: The function blocks for data input Temperature Measurement
Thermocouples generate a voltage corresponding to the temperature difference between the connection and the measuring point. The LPT measuring system measures the temperature
63
Chapter 3 The test bench
of every thermocouple terminal block and adds the corresponding voltage to the measured voltage. In addition, the LPT measuring system amplifies the thermoelectric voltage in order to increase the resolution of the analogue to digital (A/D) converter. The standard output of the type K thermocouples used, were 40µV/°C and this value was amplified to 10mV/°C. In the range up to 150°C the type K thermocouples were linear, but above this temperature, software linearization was necessary. This action takes place in the middle function block, called “Thermo_Lin”.
The software linearization expects an unamplified thermoelectric voltage and, thus, it was necessary to retransform the incoming amplified signal to the unamplified standard output of the type K thermocouples. This was done in the first function block, called “Teiler1” by multiplying the signal by the matching factor.
The third block, called “Schreiber_Temp” generates a visual image of the measured temperature values in the DIAdem module “VIEW”. The readings are displayed in a similar way as on a line recorder. Figure 29 illustrates the three function blocks, needed for the temperature measurement.
Figure 29: The function blocks required for temperature measurement
64
Chapter 3 The test bench
The function block “Matrix1” generates a different view of the temperature readings where the temperature values of the air side grid measurements after the evaporator are shown in a colour matrix. The purpose of this type of graph was to make the identification of hot or cold spots on the evaporator easy. This would be the result of an unhomogeneous distribution of the refrigerant. Figure 30 shows the function block for generating the temperature matrix.
Figure 30: The function block for the temperature matrix
Torque and mass flow measurement
The function blocks for the measuring of torque and mass flow include a scaling block and a visualisation block. The visualisation block creates a numeric display of the measured values with the module “VIEW”. The scaling block allows the conversion of the measured voltage to the corresponding physical value. The function blocks used only support a linear relationship and two points of the straight line define the conversation.
In order to acquire the linear relationship of the torque sensor of the compressor drive, a test was performed. A lever arm with a defined length was fixed to the compressor’s pulley. The lever was loaded with different weights and the voltage output of the torque sensor was recorded. The result of the test is shown in Figure 31.
65
Chapter 3 The test bench
Calibration of Torque Sensor 3,500 3,000 2,500 Voltage [V]
y = 0,0995x - 0,0409 2,000 1,500 1,000 0,500 0,000 0
5
10
15
20
25
30
35
Torque [Nm]
Figure 31: The test results used for the calibration of the torque sensor
The values for the conversion of the voltage output of the mass flow meters for oil and refrigerant could be calculated from the information provided by the manufacturer. Standard outputs of these types of mass flow meters are currents between 4 to 20mA. By using a shunt circuit with a matching resistor the current is converted to a voltage signal in the range of 0 to 10 volts. Figure 32 illustrates the function blocks for the measuring of torque and mass flow.
Figure 32: The function blocks for torque and mass flow measurement
66
Chapter 3 The test bench
Pressure and humidity measurement
The function blocks for the pressure and humidity measurement are similar to the configuration for the torque and mass flow measurement. The scaling block converts the measured voltage to a pressure value. There are two blocks for the visualisation of the pressure values. One generates a virtual line recording and the other shows the values as numbers. Figure 33 illustrates the function blocks for the pressure measurement.
Figure 33: The function block for measuring pressure
Saving data
There is a function block, which allows the saving of the measured values on the hard drive of the computer. To extract the data saved, all data bus lines of the converted values have to be connected with the block, called “Speichern1”. In addition to this, the block needs an input value for the desired frequency for saving data (yellow line). The following picture shows the function block for the saving of measured data.
67
Chapter 3 The test bench
Figure 34: The function block for saving data
3.2.4.2 The VIEW module
The VIEW module allows the on-line visualisation of the measured values during the measuring process. The DAC module generates function blocks for visualisation. These function blocks create graphs or numeric views of the measured values on the VIEW surface. The VIEW module contains several tools to modify these displays. During the tests the refrigerant circuit was monitored using the VIEW module. 3.2.4.3 The DATA module
The DATA module can be used to manage and process the recorded data. The data was arranged in channels, each displaying a data series. It was possible to undertake all mathematical operations using the values and the results could be stored in existing or new channels. There are several tools for editing the data and for searching for special values (minimum, maximum, etc) within a data set.
Table 3 summarises all the sensors and their position in the circuit. The names, “gas cooler” and “evaporator” indicate the positions of these heat exchangers in air-conditioning mode, irrespective of the mode of operation (air-conditioning or heat pump). In heat pump
68
Chapter 3 The test bench
mode, one must bear in mind that the sensor labelled as, “gas cooler”, is actually the evaporator, and that the inlet is the outlet. Name of sensor Evap_1 - _16 GC_1 - _16 Mass_CO2 Mass_Oil p_v2 p_gc1 p_gc2 p_ihx_HP2 p_E1 p_o1 p_o2 p_ihx_LP2 p_V1 p_CC t_v2 t_gc1 t_gc2 t_ihx_HP2 t_E1 t_o1 t_o2 t_ihx_LP2 t_V1 t_CC Torque_1 Humidity_1 Humidity_1_temp Coaxial_in Coaxial_out
Position of sensor Grid: air temperature after evaporator Grid: air temperature after gas cooler Mass flow of CO2 after low pressure exit of IHX Mass flow of oil return line before compressor inlet High pressure after the compressor inside the oil return line High pressure before gas cooler High pressure after gas cooler High pressure after high pressure exit of IHX High pressure before the expansion valve Low pressure before the evaporator Low pressure after the evaporator Low pressure after low pressure exit of IHX Low pressure before compressor suction intake Crank case pressure of compressor before the control valve Temperature after the compressor inside the oil return line Temperature before gas cooler Temperature after gas cooler Temperature after high pressure exit of IHX Temperature before the expansion valve Temperature before the evaporator Temperature after the evaporator Temperature after low pressure exit of IHX Temperature before compressor suction intake Crank case temperature of compressor Torque of the compressor drive Humidity of air at air outlet of HVAC-box Temperature at humidity sensor Temp. of coolant inlet of secondary circuit heat exchanger Temp. of coolant outlet of secondary circuit heat exchanger Table 3: The sensors and their position
After the modification of the test bench to a system with a secondary circuit which will be explained later, the air temperature sensors on the grid following the evaporator (marked in red on Table 3) were replaced with temperature sensors at inlet and outlet of the new coaxial heat exchanger (marked in green on Table 3).
69
Chapter 3 The test bench
3.2.4.4 GRAPH module
The GRAPH module was used to create presentation graphics to document the test results. The data could be shown as tables or graphs which could also be modified as required.
3.3 Summary The test bench can be subdivided into two parts; the refrigerant circuit and the measuring system. The refrigerant circuit includes the three sub-circuits, namely the heat exchanger circuit, the oil return circuit and the compressor control circuit. The heat exchanger circuit is the main part of the system where all thermodynamic processes take place.
By-passing the oil return circuit which includes devices for filtering, mass flow measuring and regulating, the oil flows from the oil separator to the oil return port of the compressor.
The compressor control circuit connects the crank case of the compressor with the low pressure line by a control valve and a parallel arranged overflow valve. An opening of the control valve decreases the pressure inside the crank case and causes an increase in the compressor displacement. As the valve is closed the pressure inside the crank case rises and the displacement of the compressor decreases. The overflow valve protects the compressor against overpressure inside the crank case if the control valve is completely closed.
The measuring system of the test bench consists of hardware and software components.
70
Chapter 3 The test bench
The hardware includes the sensors, the signal conditioning unit and the personal computer. The sensor transforms the physical value into an electrical value and is connected to the signal conditioning unit. The test bench provides sensors for the measuring of temperature, pressure, relative humidity, volume flow and torque.
The signal conditioning unit amplifies the electrical signals and transforms them into digital values. By using the standard printer port the signal conditioning unit was connected to the personal computer.
As measuring software, DIAdem® was installed on the personal computer. It is a blockorientated tool for real time visualisation, recording and analysis of the measured data.
71
Chapter 4 The Flash Fogging problem
Chapter 4 The Flash Fogging problem This chapter is subdivided into three parts. Firstly, a detailed thermodynamic investigation of the Flash Fogging phenomenon is explained. Secondly, the different measures against the Flash Fogging problem are explained and discussed, and, finally, the selected model and construction thereof is described.
4.1 Investigation of Flash Fogging In order to understand Flash Fogging, it is important to analyse what occurs inside the AC box when the cycle is reversed. The humidity of the air that passes through the heat exchanger is increased when there is a change from the air-conditioning operation to the heat pump operation. In order to forecast if the increase in humidity will lead to a “steaming-up” of the vehicle’s windows, it is necessary to determine the dew point temperature of the air, which exits the AC box. On investigating the phenomenon of the Flash Fogging problem, it was found not to be exclusively a CO2 system problem, as it was never reported for conventional vehicle AC systems.
72
Chapter 4 The Flash Fogging problem
4.1.1 Incidents during the reversal of the cycle As mentioned previously, Flash Fogging can be described as the sudden “steaming-up” of the windscreen during the process of reversing the cycle from air-conditioning mode to heat pump operation mode. The Flash Fogging phenomenon can occur in every refrigerant cycle which can be changed from air-conditioning operation to heat pump operation. During the air-conditioning operation ambient air is cooled down over the evaporator’s surface. If the surface temperature of the evaporator is lower or equal to the dew point temperature of the ambient air, which can be defined as the temperature at which condensation begins when the air is cooled at constant pressure [16], some of the water contained in the ambient air is condensed. The condensate forms droplets of water on the evaporator’s surface. As the droplets flow down the exterior of the evaporator, they accumulate in the collection tray. There is a drainage opening in the lowest point of the collection tray where the condensate leaves the HVAC box.
Although the surface of the evaporator, especially the fins are optimised for condensate drainage, a significant amount of condensate adheres to the surface. The exterior design of an evaporator is always a compromise between thermodynamic effectiveness and condensate drainage. A large number of small fins generate a greater exterior surface, which improves thermodynamic effectiveness. A smaller number of large fins improves the condensate drainage but reduces the efficiency of the heat exchange process.
As the cycle is reversed to heat pump operation, the temperature of the evaporator rises very quickly and the portion of the condensate, adhering to the evaporator’s surface starts to vapourise. The remaining portion of the condensate in the air is carried away by the air flow through the heat exchanger. Both processes increase the absolute humidity of the air, 73
Chapter 4 The Flash Fogging problem
which is defined as the mass of water vapour present in a unit mass of dry air [16], exiting the HVAC box.
When the air leaving the HVAC box is blown onto the windscreen, which is below the dew point temperature, condensation of some of the water in the air takes place on the windscreen’s surface and this process is called, “Flash Fogging”.
4.1.2 Dew point temperature To determine the dew point temperature of humid air, it is necessary to measure the temperature and relative humidity of the air and from these two values, the dew point temperature can be calculated or determined graphically, using the correct psychrometric chart. 4.1.2.1 Graphical method
The psychometric chart is a helpful tool for analysing psychrometric processes and acquiring properties of moist air used for calculations. The basic features of the psychrometric chart are illustrated in Figure 35.
Figure 35: Features of a psychrometric chart
74
Chapter 4 The Flash Fogging problem
To determinate the dew point temperature from a psychrometric chart, firstly the state point has to be plotted. It is specified as the intersection point of the constant dry bulb temperature line and the constant relative humidity curve. Now, moving from the state point, along the constant absolute humidity (or humidity ratio) line down to intersect with the saturation curve, the dry bulb temperature can be read off, which represents the dew point temperature. Figure 36 illustrates an example of the graphical determination of the dew point temperature.
Figure 36: An example of the graphical determination of dew point temperature
In the example illustrated in Figure 36, the specified state point is at 30°C and 40% relative humidity. The value of absolute humidity at the intersection point is 11gwater/kgair. Now, following the line of constant absolute humidity down to the saturation curve, the dew point temperature can be read off from the intersection point and is approximately 14,8°C (by interpolation). 75
Chapter 4 The Flash Fogging problem
4.1.2.2 Calculation of the dew point temperature
Knowing the temperature and relative humidity, the determination of the dew point temperature can also be done by calculation. Firstly, the saturation pressure of the specified state point is calculated from the temperature. The saturation pressure is defined as the highest possible partial pressure of water vapour, contained in humid air at a certain temperature. t ps = 288,6 ⋅ 1,098 + 100
saturation pressure,
t
8, 2
Eq. 3 [17]
temperature of state point (°C)
30 p s = 288,6 ⋅ 1,098 + 100
8, 2
= 4239,0 Pa
From the measured relative humidity and the calculated saturation pressure, the absolute humidity can be calculated.
absolute humidity,
x = 0,622 ⋅
ϕ ⋅ ps p − ϕ ⋅ ps
ϕ
relative humidity (% RH·100-1)
ps
saturation pressure (Pa)
p
ambient (or total) pressure (Pa)
x = 0,622 ⋅
kg 0,4 ⋅ 4239 = 0,01059 water 101325 − 0,4 ⋅ 4239 kg air
76
Eq. 4 [17]
Chapter 4 The Flash Fogging problem
It is now necessary to determine the saturation pressure for the dew point temperature at the calculated absolute humidity, as the absolute humidity is constant for both conditions. This is done by modifying the equation for the absolute humidity [Eq. 4]. The value for the relative humidity is set to 1 (100%) and the subject of the equation is changed from x to ps. saturation
ps ,100% =
pressure for ϕ =1
x⋅ p x + 0,622
x
absolute humidity (kgwater·kgair-1)
p
ambient pressure (Pa)
p s ,100% =
Eq. 5
0,01059 ⋅101325 = 1695,3 Pa 0,01059 + 0,622
As the saturation pressure is known, the matching temperature can now be found. This is done by using equation 3 for the determination of the saturation pressure. The equation is modified by changing the subject from ps to t. The temperature, which is calculated is the dew point temperature.
dew point temperature
p t = 8,02 s ,100 % − 1,098 ⋅100 288,6
ps ,100% saturation pressure at 100% RH (Pa)
1695,3 − 1,098 ⋅100 = 14,9 °C t = 8,02 288,6
77
Eq. 6
Chapter 4 The Flash Fogging problem
The dew point temperature of air at a state of 30°C and 40% relative humidity, determined by calculation is 14,9°C.
4.1.3 Flash Fogging in conventional AC cycles? Discussing the Flash Fogging phenomenon, the question arises why the problems of reversing a refrigerant cycle from air-conditioning operation to heat pump operation seems to be a problem particular to CO2. In present R134a systems and previous R12 refrigerant systems, the additional use as a heat pump was never considered. The reasons for this were the different temperature ranges of the heat source and heat sink inside a vehicle’s refrigeration cycle, and the different system pressures of CO2 and conventional refrigerants. This is best explained in the following example.
The design temperature in Europe for a vehicle’s heating system is for an ambient condition of –20°C [18]. As the ambient air is used as a heat source, the evaporation temperature has to be lower than the ambient temperature. In order to achieve good heat transfer, the temperature difference is set at 10K. Thus the evaporation temperature has to be –30°C. At this condition, CO2 has an evaporation pressure of 14,3bar. If R134a would have been used, the evaporation pressure would have been below ambient pressure.
Using conventional compressors, a negative pressure difference between the refrigeration cycle and ambient conditions causes serious problems. The critical part is the shaft seal of the compressor. It is designed as a lip seal and therefore needs an internal “overpressure” to seat it properly. In this case, as the ambient pressure is higher than the system pressure, the seal starts leaking and the ambient air enters the system.
78
Chapter 4 The Flash Fogging problem
From an engineering point of view it is possible to redesign the shaft seal to withstand overpressure as well as vacuum, but “pressure” in the automotive industry regarding expenditure does not allow for additional costs for engineering and the seal itself. In a comparative study of air-conditioning and heat pump systems with CO2 and R134a carried out by three car manufacturers (Audi®, BMW®, DaimlerCrysler®) stated: “heat pump systems were just possible with R744”[19].
4.2 Measures against Flash Fogging There are two different approaches in solving the Flash Fogging problem. One solution could be the prevention of an increase in humidity during the reversing of the refrigerant cycle. Another approach are strategies that allow an increase in humidity but prevent the condensation at the windows of the car.
In order to prevent the increase in humidity, there has to be one separate heat exchanger for both operation modes of the system. During AC operation the first heat exchanger is in use and the condensate accumulates between the fins. As the circuit is reversed, the second heat exchanger is put into operation. While the second heat exchanger is in use, the first heat exchanger is separated from the air and the refrigerant flow. Thus the condensate adhering to the surface of the first heat exchanger cannot be picked-up by the air flow. In implementing this idea, two different types of cycles are possible. Firstly, a direct refrigerant cycle and a system with a secondary coolant circuit. In a direct refrigerant cycle the standard CO2 system is extended by a second refrigerant heat exchanger inside the HVAC box. The second type of system consists of a refrigeration cycle and a coolant cycle. In a central heat exchanger the heating or refrigeration power is transferred to a flow
79
Chapter 4 The Flash Fogging problem
of coolant. The coolant feeds two separate liquid-to-air heat exchangers inside the HVAC box.
If the humidity increases during the process of reversing the cycle, two possibilities to avoid Flash Fogging remain. Either the windscreen is heated or the humid air flow is redirected in a way that the vehicle’s windows are not effected by it.
4.2.1 Electrically heated windscreen The process of condensation depends on the dew point temperature of the air and the temperature of the surface adjacent to which the air flows. If the surface temperature is increased above the dew point temperature, condensation will not occur. An electrically heated windscreen could be used to raise the surface temperature.
There are two different types of safety glass on the market. The method of heating the glass differs form type to type. In the case of a single pane toughened safety glass (TSG), heating conductors are applied to the window using the silk screen process, and are sintered in place during the prestressing process. Subsequent galvanisation increases the strength of the heating conductors, tempers them and protects them against environmental influences [20].
In the case of laminated safety glass (LSG), very thin, nearly invisible heating wires are applied in wavy lines to the plastic adhesive film. They are connected in series and/or parallel in order to achieve the intended electrical resistance. The thin wires provide even better heating coverage. The typical heating power of this kind of system is approximately 3 – 5W·dm-² [20].
80
Chapter 4 The Flash Fogging problem
4.2.2 Redirection of air flow during change of operation mode A different solution to prevent the humid air from causing Flash Fogging at the windows could be the automatic control of the air distribution flaps. During the change of operation mode of the refrigerant circuit only the foot level flaps should be opened. Following a definite time, by which the surface of the heat exchanger is considered to be dry, the flaps of the upper positions could be opened again.
Kampf [21] investigated the Flash Fogging for different CO2 heat pump systems. In his experiments the evaporator contained up to 200 grams of water after dehumidification. Under worst case conditions, this could cause the complete steaming-up of all the windows in the vehicle. It was not possible to prevent Flash Fogging by controlling the flaps of the AC box. In his opinion the only way to solve the problem was the use of an additional heat exchanger.
4.2.3 An additional HX in the refrigerant cycle The use of an additional heat exchanger in the refrigerant circuit prevents the increase in humidity of the air flow through the HVAC box during the reversing of the refrigerant cycle. Two separate heat exchangers were available, one for the air-conditioning mode, the other for the heat pump mode. They should be arranged parallel to each other inside the HVAC box. This kind of arrangement allows one heat exchanger to be completely separated from the air flow, as the operation mode changes. In addition, the parallel arrangement guarantees a homogeneous incoming flow to both heat exchangers. The HVAC boxes of conventional systems show a tandem arrangement of the heat exchangers which is not particularly suitable for reversible systems.
81
Chapter 4 The Flash Fogging problem
4.2.4 Use of a secondary (coolant) cycle In general, the system with a secondary cycle is similar to the system with two heat exchangers in the refrigerant cycle. Instead of the two refrigerant heat exchangers inside the HVAC box, there are two coolant-to-air heat exchangers. They are fed by a central heat exchanger which transfers the heating or refrigeration power to the coolant circuit. To circulate the coolant through the system, an additional pump is necessary.
The major problem with this solution, besides the arrangement of the heat exchanger inside the HVAC-box, was the availability of the required heat exchangers. There were no applicable liquid-to-air designs on the market. Commonly used heater cores are designed for large temperature differences and as a result can be manufactured to have relatively small heat transfer surface areas. This is ineffective when they are utilized as coolers since the temperature difference is not as great and this requires the core to have a relatively large heat transfer surface area. An even bigger problem was the refrigerant-to-liquid heat exchanger. Strictly speaking there was no suitable design on the market.
4.2.5 Discussion The investigation of Kampf [21] showed, that Flash Fogging could not be prevented by control of the flaps of the HVAC box. Also, a heated windscreen did not offer a reliable prevention from Flash Fogging as it needed some time to heat up and hence could not respond quickly to the stream of moist air. The continuous operation of the heating device increases the consumption of electric energy and is undesirable. Furthermore, Kampf found that the amount of condensate at the evaporator’s surface could cause a complete steaming-up of all the windows in the vehicle and if Flash Fogging could be prevented by merely heating, all windows would have to be heated. 82
Chapter 4 The Flash Fogging problem
Looking at these facts, an increase in humidity has to be prevented in order to eliminate Flash Fogging. Only the systems with separate heat exchangers for the different operation modes offered a solution.
The system with the secondary cycle has a great advantage in, that there is no longer a source of dangerous high pressure CO2 inside the cab as all the components of the refrigerant cycle are located under the bonnet with only a coolant circuit running through the vehicle’s interior. The components of the coolant circuit are cheap and the routing of the coolant hoses is much more flexible than that for refrigerant lines. Thus, the positioning of the whole HVAC box inside the interior is flexible. Furthermore, the system offers the possibility of being connected to the engine coolant circuit to form a total thermal management
system.
This
is
described
in
more
detail
in
“Chapter
7
Future Opportunities”. On the other hand, a disadvantage of the secondary cycle system would be the reduced efficiency caused by the loss of two heat exchanging processes (refrigerant-to-coolant, coolant-to-air) and the previously mentioned poor availability of the heat exchangers.
The system with two refrigerant heat exchangers shows a better efficiency, because of the direct heat exchange process of refrigerant and air. With two refrigerant heat exchangers inside the HVAC box, the routing of the line would be more difficult and the system would be less flexible than the secondary cycle system. With the direct cycle, the CO2 stays inside the cab and because of the second heat exchanger the risk of leakage would be increased.
83
Chapter 4 The Flash Fogging problem
A disadvantage of both systems with two separate heat exchangers would be that the reheat operation is not possible. In this circuit, cooled air coming from the air cooler would be heated by a subsequent heat exchanger which would provide warm, dehumidified air. In a system with two separate heat exchangers both heating and cooling are not possible at the same time.
The selected solution for the test bench was the secondary cycle system, because of the flexibility and cost effectiveness of the coolant cycle and it provided the best possibility for mass production. The poor availability of the heat exchanger would only pose a problem in the early stages of CO2 air-conditioning. On serial production, applicable refrigerant-tocoolant and coolant-to-air heat exchangers would be available. As adapted heat exchangers were used in this research, the loss of heat or refrigeration energy during the two heat exchange processes could be minimised.
84
Chapter 5 Secondary cycle system on the test bench
Chapter 5 Secondary cycle system on the test bench This section describes the implementation of the secondary cycle system into the test bench. The additional components and their arrangement are described. The refrigerant-tocoolant heat exchanger is emphasised because of its difficult dimensioning.
5.1 Additional components and their arrangement The components of the secondary cycle system were not designed to fit into a vehicle but the set-up of the cycle in the test bench serves to investigate Flash Fogging.
The basic component of the system is the refrigerant-to-coolant heat exchanger through which the coolant flows and removes heat or provides the refrigeration power, depending on the operation mode. Temperature sensors at the inlet and outlet of the heat exchanger were connected to the system using a clamping ring fitting. On leaving the heat exchanger the coolant is drawn in by the circulation pump which was a standard wet running pump used in central heating systems. Its characteristics were similar to a standard coolant pump used in a vehicle. The acceptable temperature range of the circulated liquid was between -10°C and +110°C. The pump feeds the coolant to a parallel arrangement of coolant-to-
85
Chapter 5 Secondary cycle system on the test bench
water heat exchangers, which are located inside the HVAC box. Figure 37 shows the two
heat exchangers.
Figure 37: HVAC box with coolant-to-water heat exchangers
Considering Figure 37 the heat exchanger in the horizontal position is the original heater core of the Mercedes A-Class. Normally the vertical position is occupied by the evaporator, but in this case it has been replaced by the heater core of a Volkswagen “New Beetle” which is identical to the Mercedes A-Class heater core with the only difference being the connecting tubes. As the original HVAC box was used, the heat exchanger (in the original evaporator position) had to serve as an air heater and the heat exchanger (in the original heater core position) as the air cooler. The reason for this was the arrangement of the flaps inside the HVAC box and only this configuration allowed for the separation of the condensate covered air cooler when the refrigeration circuit was reversed. Figure 38 illustrates the arrangement of the flaps and the heat exchangers inside the HVAC box. 86
Chapter 5 Secondary cycle system on the test bench
Figure 38: The different operation modes of the HVAC box
A ball valve was located in the coolant line before and after each heat exchanger which allowed a complete separation of the heat exchangers from the coolant circuit. Following the valves at the outlet of both heat exchangers, the parallel lines merge again and direct the flow to the refrigerant-to-coolant heat exchanger. An expansion tank was connected to a stub line at the heat exchanger’s inlet. The additional coolant inside the tank compensates for volumetric changes of the water/glycol mixture caused by the temperature variations of the different operation modes. Figure 39 illustrates the layout of the whole system.
87
Chapter 5 Secondary cycle system on the test bench
Figure 39: The layout of the secondary circuit
All the components of the system were linked with copper tubes whose dimensions were Ø18x1mm, and the tubes and fittings were joined by soft soldering.
5.2 The selection of the heat exchanger for the secondary cycle The core of the secondary circuit was the refrigerant-to-coolant heat exchanger. In airconditioning operation, the heat exchanger works as an evaporator and transfers the
88
Chapter 5 Secondary cycle system on the test bench
refrigeration power to the coolant. In the heat pump mode, the heat exchanger operates as a gas cooler and transfers heat energy to the coolant.
A secondary cycle heat exchanger for use in a vehicle has to be very compact. The small outer dimensions of the heat exchanger can only be achieved with designs which use multiport profiles. It was, however, not possible to implement such a design because the multiprofiles were not yet available on the market. Furthermore, for the manufacturing of heat exchangers made from multi-port profiles, a vacuum furnace was needed and this appliance could not be provided by the Fachhochschule or the project partner Valeo Klimasysteme® GmbH. The search for an adequate heat exchanger design was focused only on the demands of the test bench and the outer dimensions were not considered.
The design of this heat exchanger had to be a compromise between the air-conditioning and heat pump mode. From a thermodynamic point of view, the air-conditioning operation was more critical because of the smaller temperature difference between the refrigerant and the coolant. The following section calculates the dimensions of the heat exchanger for the cooling operation only.
5.2.1 Considering a plate type heat exchanger Currently, plate type heat exchangers are the most effective coolant-to-refrigerant evaporators or condensers respectively. The plate-type heat exchanger is in principle constructed as a plate package of corrugated channel plates between front and rear coverplate packages. The number of the corrugated plates is variable. One heat exchanger can consist of up to 200 plates. In a stack of plates every second plate is turned around so that the edges of the herringbone pattern cross over each other. During the vacuum-brazing
89
Chapter 5 Secondary cycle system on the test bench
process every contact point between the plates is soldered together. In this way, two separate channels are formed where counter flow of the two media occur. The complex pattern of channels lead to a highly turbulent flow, which ensures a very high thermal conductivity. Figure 40 illustrates the principle of a plate type heat exchanger.
Figure 40: The principle of the plate-type heat exchanger
When using CO2 as refrigerant with this type of heat exchanger, two problems are experienced. The first problem is the maximum operation pressure, because plate type heat exchangers are designed for conventional refrigerants where, a maximum operating pressure of 30 bar is sufficient. If CO2 is used, the maximum operating pressure could be higher than 200bar. The problem can be overcome by using reinforcement plates made of steel. These reinforcement plates should support the upper and the underside of the heat exchanger. Four bolts press the reinforcement plates and the stack of heat exchanger plates together. It is sufficient to support only the top and the bottom of the plates of the stack because there is no pressure difference between the inner plates, which could damage the heat exchanger. Only the outer plates are exposed to the high pressure difference between
90
Chapter 5 Secondary cycle system on the test bench
the ambient condition at 1bar and the refrigerant circuit which has a working pressure of up to 140bar.
The other problem is the large cross-sectional area on the refrigerant side and, hence, the ratio between the heat exchange area of the refrigerant and the heat exchange area of the coolant which is too small. A large cross-sectional area results in a small velocity of flow of the CO2. Ideally for good thermal conductivity a velocity greater than 10 m·s-1 is required. With the standard plate type heat exchangers a velocity of less than 1 m·s-1 is possible.
Because of the high volumetric refrigeration power of CO2, a heat exchanger operated with this refrigerant requires more area on the coolant side than on the refrigerant side. The standard plate type heat exchangers offer a ratio of refrigerant area to coolant area of 50:50. This means that the refrigeration power produced in one unit volume on the refrigerant side cannot be completely transferred to the coolant and the effectiveness of the heat exchanger is reduced.
Despite all these restrictions, a standard plate type evaporator was investigated to establish whether it would be suitable for use with CO2 as a refrigerant. The calculations were conducted with in-house software used by the test centre of Valeo Klimasysteme® GmbH.
In the planned secondary loop system, the existing radiator in the HVAC box was used as an air cooler. It transferred the refrigeration power from the cooled water/glycol mixture to the air. The dimensioning of the refrigerant-to-coolant evaporator was dependent on the performance of the glycol-to-air heat exchanger. For this reason it was necessary to, firstly, 91
Chapter 5 Secondary cycle system on the test bench
calculate the cooling power of the radiator for the cooling operation. The heat exchanger was a standard radiator of the Mercedes® A-Class which was the Corion 15T14.3 manufactured by Valeo®. Table 4 summarises the heat exchanger characteristics. Description
Unit mm piece mm mm mm piece kg·h-1 kg·m-3 l·h-1 °C °C °C °C W kPa
Core depth number of tube pitches Finned length Finned height Fin pitch Quantity of tubes Air mass flow Air density Ratio of water/glycol mixture Volumetric flow of coolant Inlet temperature of air Outlet temperature of air Inlet temperature coolant Outlet temperature coolant Heating power Pressure drop (coolant)
Value 35 14 240,1 170,3 1,3 15 400 1,2 50/50 700 40 2,41 -6 0,01 -4205 6,81
Table 4: The heat exchanger characteristics for the Valeo® Corion 15T14.3 as an air cooler
The cooling power of the radiator was calculated as 4205W and, hence, the plate-type evaporator had to be designed to transfer at least this amount of energy.
The plate-type evaporator chosen was supplied by SWEP®. This company supplies the biggest variety of plate-type heat exchangers world-wide. Based on the above calculation, two different plate types with the smallest cross-sectional area of the refrigerant side were chosen to determine their suitability for use with CO2. The following table shows the technical data of the selected plates.
92
Chapter 5 Secondary cycle system on the test bench
Unit A B C D E Max. possible quantity of plates Area per plate Volume per channel Max. possible volumetric flow Weight of heat exchanger Material of heat exchanger Soldering metal Max. operating pressure Max. operating temperature Min. operating temperature
mm mm mm mm mm piece m2 l m3·h-1 kg bar °C °C
Plate type B 15 B8 466 312 72 72 432 278 40 40 9 + 2,3 x np 9 + 2,3 x np 60 60 0,036 0,023 0,051 0,034 4 4 1,3 + 0,106 x np 0,9 + 0,07 x np AISI 316, 1.4401 AISI 316, 1.4401 Copper 99,9% Copper 99,9% 30 30 185 185 -195 -195
Table 5: Technical data of the selected plates
Figure 41 illustrates the front and side views and the typical dimensions required for a plate type heat exchanger.
Figure 41: The typical dimensioning of the plate-type heat exchanger
93
Chapter 5 Secondary cycle system on the test bench
Based on the data in Table 5 and the experience of the Valeo Klimasysteme® GmbH test centre with heat exchangers in CO2 service, the design calculations were performed for two different types of corrugated plates and the results are summarised in Table 6.
Calculated values
Thermodynamic demands
Dimensions
Unit Length of plate (A) Width of plate (B) Height of stack (E) Height of single plate Quantity of plates Material thickness of plate Weight of heat exchanger Cross-section per plate (coolant) Cross-section per plate (CO2) Outer area (coolant) Ratio of outer and inner fins (Fo/Fi) Volumetric flow of coolant Inlet temperature of coolant Outlet temperature of coolant Inlet temperature of CO2 Inlet pressure of CO2 Enthalpy of CO2 Outlet density of CO2 (dry saturated) Mass flow of CO2 Velocity of flow of coolant Outer heat transfer coefficient (coolant) Velocity of flow of CO2 Inner heat transfer coefficient (CO2) Overall heat transfer coefficient Pressure before exp.-valve (CO2) Temperature before exp.-valve (CO2) Inlet enthalpy (CO2) Log. temperature difference (counterflow) Refrigeration power
mm mm mm mm piece mm kg mm2 mm2 m2 l·h-1 °C °C °C bar kJ·kg-1 kg·m-3 kg·h-1 m·s-1 W·m-2K-1 m·s-1 W·m-2K-1 W·m-2K-1 bar °C kJ·kg-1 K W
Plate type B 15 B8 466 312 72 72 57,5 110,4 2,3 2,3 25 48 0,4 0,4 3,91 4,66 70 70 70 70 0,9059 1,1645 1 1 700 700 -6,00 -6,00 0,10 0,10 -9,13 -9,13 27,00 27,00 64,12 64,12 72,67 72,67 112,00 112,00 0,22 0,12 1419,52 959,75 0,49 0,25 2000 2000 830 649 120,00 120,00 45,00 45,00 200,14 200,14 5,65 5,65 4250,58 4268,22
Table 6: The results of the calculation of the plate-type evaporator
The results show that both plates could form heat exchangers that would be able to meet the target of 4205 Watts of refrigeration power, but not both plates are ideally suited for use with CO2 as the cross-sectional areas of the refrigerant and the coolant sides are too large. By retaining the same surface area, the cross-sectional areas should be reduced by a
94
Chapter 5 Secondary cycle system on the test bench
factor of between 5 and 10. Comparing the two plates, the B15 plate is more suitable, because it shows a better ratio between cross-sectional area and heat transfer surface.
Based on results summarised in Table 6, the decision taken was to use a standard platetype heat exchanger with CO2 as the refrigerant, but the poor performance did not justify the effort of designing a suitable reinforcement package to increase the operating pressure of the heat exchanger.
5.2.2 Considering a coaxial type heat exchanger constructed from stainless steel (material number: 1.4571) The principle calculation for the evaporators’ tube length is as follows: surface area of the
AV =
evaporator
Q& C k ⋅ ∆t
Eq. 7 [22]
Q& C
cooling capacity required (kW)
k
overall heat transfer coefficient (kW·m-2·K-1)
∆t
temperature difference between the coolant and the wall of the tube: twater/glycol - twall (K)
overall heat transfer
k=
1
1
αo
coefficient
αo
+
δ
λstainl . steel
+
outer
1
Eq. 8 [23]
αi surface
heat
transfer
coefficient
(water/glycol) (W·m-2·K-1)
αi
inner thermal conductivity (CO2) (W·m-2·K-1)
δ
wall thickness = 0,001m
λstainl. steel
thermal conductivity = 14W·m-1·K-1 [24]
95
Chapter 5 Secondary cycle system on the test bench
The inner heat transfer coefficient, αi
The values for the heat transfer coefficient of evaporating carbon dioxide were provided by the Valeo Klimasysteme® GmbH test centre. A heat transfer coefficient of 2500W·m-2·K-1 was recommend and was based on experience for real conditions (oil wetting at the inner wall) with a velocity of flow greater than 10m·s-1 at the evaporator outlet. Therefore, as a first step it was necessary to find the matching inner tube diameter to achieve the required velocity of flow at the evaporator outlet.
volumetric flow
m& V& =
Eq. 9
ρ
m&
CO2 mass flow = 112kg·h-1
ρ
CO2 density at outlet, dry saturated = 72,67kg·m-3
112 V& = = 1,5412 m 3 ⋅ h −1 72,67
cross sectional area
A=
V& w
Eq. 10
V&
CO2 volumetric flow (m3·h-1)
w
required velocity of flow = 10m·s-1
A=
1,5412 = 4,2811⋅10 −5 m 2 10 ⋅ 3600
96
Chapter 5 Secondary cycle system on the test bench
radius of the tube
A
r=
π
=
4,2811⋅10 −5
π
= 3,6915 ⋅10 −3 m
The result of the calculation was an inner radius of 3,69mm. An inner radius of 3mm was chosen as it corresponded to the inner radius of an 8mm stainless steel tube with a wall thickness of 1mm. Now, recalculating, this produced a velocity of flow of 15,15m·s-1.
The outer heat transfer coefficient, αo
The calculation of the outer heat transfer coefficient was done using the method of calculation described in [25].
Firstly it was necessary to calculate the velocity of flow inside the concentric gap. velocity of flow
w=
V& Ai , outside − Ao , inside
V&
Eq. 11
volumetric flow of the water/glycol mixture = 0,7m3·h-1
Ai , outside
inner cross-sectional area of the outer tube (m2)
Ao , inside
outer cross-sectional area of the inner tube (m2)
w=
0,7 = 3,0947 m ⋅ s −1 2 π ⋅ (0,006 − 0,004 ) ⋅ 3600 2
97
Chapter 5 Secondary cycle system on the test bench
Now, calculating the Reynolds number. Re =
Reynolds number
w
w⋅l
Eq. 12
υ
velocity of flow of the water/glycol mixture inside the concentric gap (m·s-1) characteristic length, in this case: inner radius
l
of the outer tube minus outer radius of the inner tube: l = (0,006-0,004)m = 0,002m
υ
kinematic viscosity of water/monoethyleneglycol 20%vol, = 2,3·10-6 m2·s-1 (at a mean temperature of 10°C, interpolated) [26]
Re =
3,0947 ⋅ 0,002 = 2691,0 2,3 ⋅10 − 6
Now, calculating the Nusselt number as described in [27]. The equation for rough calculations was used which is valid for 1,5 < Pr < 500 for turbulent flow.
Nusselt number
Num = 0,012 ⋅ (Re0,87 − 280) ⋅ Pr 0, 4 ⋅ [1 + (d i / l ) 2 / 3 ]
Eq. 13
Re
Reynolds number
Pr
Prandtl number of water/monoethylenglycol 20%vol, Pr = 18,29 (at a mean temperature of 10°C, interpolated) [26]
98
Chapter 5 Secondary cycle system on the test bench
di
inner diameter of the tube, in this case: hydraulic diameter = characteristic length = 0,002m length of the tube assumed to be 12m (to be
l
checked later)
Num = 0,012 ⋅ (2691,00,87 − 280) ⋅18,290, 4 ⋅ [1 +
0,002 ] = 26,25 12
The next equation considers the dependency of temperature on the material property values used in the previous equations [27]. Nusselt number, temperature adjusted
Nutube
Pr = Nu m ⋅ Prw
0 ,11
Eq. 14
Num
Nusselt number, not temperature adjusted
Pr
Prandtl number at a mean temperature of the medium (10°C)
Prw
Prandtl number at the wall of the tube for water/monoethylenglycol 20%vol, Pr (0°C) = 23,98 [26]
18,29 Nu = 26,25 ⋅ 23,98
99
0 ,11
= 25,48
Chapter 5 Secondary cycle system on the test bench
This Nusselt number was calculated for the flow through a tube and in the next step it has to be adjusted to reflect the special characteristics of the concentric gap. There are three possible cases for the calculation of the heat transfer inside the concentric gap [25]: 1. heat transfer at inner tube, outer tube insulated 2. heat transfer at outer tube, inner tube insulated 3. heat transfer at both tubes with the same wall temperature at inner and outer tube
The conditions inside the calculated coaxial heat exchanger are represented by case number 1. In this case Equation 15 is valid. Nusselt number, concentric gap
d Nui = 0,86 ⋅ o di
0 ,16
⋅ Nutube
do
inner diameter of outer tube (m)
di
outer diameter of inner tube (m)
Nutube
Nusselt number for flow through a tube
0,012 Nui = 0,86 ⋅ 0,008
Eq. 15
0 ,16
⋅ 25,48 = 23,38
Now, using the Nusselt number the outer heat transfer coefficient may be calculated. heat transfer coefficient
α=
Nu ⋅ λ dh
Eq. 16
Nu
Nusselt number, concentric gap
λ
thermal conductivity of water/monoethylenglycol 20%vol, = 0,507 W·m-1·K-1 (at a mean temperature of 10°C, interpolated) [26] 100
Chapter 5 Secondary cycle system on the test bench
dh
hydraulic diameter = characteristic length = (0,006-0,004)m = 0,002m
α=
overall heat transfer
23,38 ⋅ 0,507 = 5926,8 W ⋅ m − 2 ⋅ K −1 0,002
1
k=
1
αo
coefficient
k=
+
δ
λstainl . steel
+
1
αi
Eq. 8
1 1 0,001 1 + + 5926,8 14 2500
k = 1562,1 W ⋅ m −2 ⋅ K −1 )
Q& C k ⋅ ∆t
surface of evaporator
AV =
temperature difference
∆t = twater / glycol − t wall
Eq. 7
Eq. 17
∆t = (0 − (− 10 )) = 10 K
surface of evaporator
AV =
4,5 = 0,2881 m² 1,5621⋅10
length of the tube
ltube =
A π ⋅ dm
Eq. 7
Eq. 18
A
surface of evaporator (m2)
dm
mean diameter of tube (m) 101
Chapter 5 Secondary cycle system on the test bench
ltube =
0,2881 = 13,10 m π ⋅ 0,007
The calculated length of 13,1m for the coaxial type heat exchanger is close to the assumption of 12m that was made in the calculation for the Nusselt number in equation 13, thus, a recalculation is not necessary.
5.2.3 Consider a coaxial type heat exchanger constructed from copper (material number 2.0090) The coaxial stainless steel evaporator could not be manufactured in-house and, thus, it was necessary to investigate designs with alternative materials. A reasonable solution for inhouse manufacturing was the use of soft-annealed copper tube where the connections could be made by hard-soldering and the tube could be easily bent without any special tools.
There are only two standard size copper tubes which are able to withstand a pressure of more than 200bar and these were tubes with an outer diameter of 6mm and 4mm with wall thickness of 1mm for both cases. The 6mm diameter tube was selected in favour of a lower refrigerant pressure drop. In order to meet the required velocity of flow of 10m·s-1, three inner tubes were required.
The diameter of the outer tube was a compromise between a high velocity of flow of the water/glycol mixture and the manufaturability of the connecting part. On the one hand, a small diameter with a high velocity of flow results in a good heat transfer between the inner tubes and the water/glycol mixture, but, on the other hand, a small diameter leads to problems with manufacturing the connector block. The smaller the diameter of the
102
Chapter 5 Secondary cycle system on the test bench
connector block, the greater is the need for accuracy of the drilling through the connector block. An outer diameter of 18mm was the best possible compromise. Figure 42 illustrates the design of the connector piece.
Figure 42: A sectional view of the connector piece
The connector piece consists of an actual brass connector block with three 6mm diameter holes drilled through it and a standard copper 18mm “T”-piece. One of the opposite ends of the “T”-piece was soldered to the 18mm copper tube and the other end was soldered to the connector block. The three refrigerant lines of 6mm copper tube were located inside the bigger copper tube and pass through the “T”-piece and were soldered into the holes drilled into the connector block. Using two stainless steel clamping ring fittings (18/12mm and 12/8mm), the connector block was coupled to the refrigerant circuit. The remaining opening of the “T”-piece was connected to the coolant circuit. Figure 43 shows the drilled connector block in the “T”-piece and the complete assembly on the heat exchanger can be seen in Figure 45.
103
Chapter 5 Secondary cycle system on the test bench
Figure 43: Connector block with “T”-piece
The inner heat transfer coefficient, αi
As in the case of the stainless steel evaporator, the heat transfer coefficient can be assumed as 2500W·m-2·K-1 if the velocity of refrigerant is greater than 10m·s-1 (values provided by the Valeo Klimasysteme® GmbH test centre), thus, it is necessary to calculate the velocity of flow for this design. The volumetric flow of CO2 is known from the previous calculation using equation 9. velocity of flow (refrigerant)
w=
V& A
Eq. 19
V&
CO2 volumetric flow, = 1,5412m3·h-1
A
cross sectional area of the inner tubes (m2)
104
Chapter 5 Secondary cycle system on the test bench
w=
1,5412 = 11,36 m ⋅ s −1 2 π ⋅ 0,002 ⋅ 3 ⋅ 3600
This is higher than the required value of 10m·s-1 and ,thus, the dimension of the inner refrigerant tube previously calculated is still acceptable.
The outer heat transfer coefficient, αo
The procedure remains the same as in the previous section. Figure 44 shows the dimensions inside the concentric gap.
Figure 44: The cross-sectional view of coaxial type heat exchanger
Firstly, it is necessary to calculate the velocity of flow inside the concentric gap. velocity of flow (water/glycol)
w=
V&
V& Ai , outside − Aa , inside
Eq. 11
volumetric flow of the water/glycol mixture = 0,7 m3·h-1
105
Chapter 5 Secondary cycle system on the test bench
Ai , outside
inner cross-sectional area of the outer tube (m2)
Aa , inside
outer cross-sectional area of the inner tube (m2)
w=
0,7 = 1,6728 m ⋅ s −1 2 π ⋅ (0,008 − 0,003 ⋅ 3) ⋅ 3600s 2
By knowing the velocity of flow it is possible to calculate the Reynolds number. Re =
Reynolds number
w
w⋅l
Eq. 12
υ
velocity of flow of the water/glycol mixture inside the concentric gap
l
characteristic length, in this case: mean distance of the inner tubes to each other and to the inner wall of the outer tube, estimated: l = 0,0015m
υ
kinematic viscosity of water/monoethylenglycol 20%vol, = 2,3·10-6m2·s-1 (at a mean temperature of 10°C, interpolated) [26]
Re =
1,6728 ⋅ 0,0015 = 1091,0 2,3 ⋅10 − 6
106
Chapter 5 Secondary cycle system on the test bench
The flow inside the concentric gap is considered to be laminar because the Reynolds number is less than 2300. The assumed boundary conditions are heat transfer at constant wall temperatures and presence of hydrodynamic laminar flow. There are three possible cases for the calculation of the heat transfer inside the concentric gap [25]. The conditions inside the calculated coaxial heat exchanger are represented by case number 1: heat transfer at inner tube with outer tube insulated. In this case, Equation 23 for the calculation of a mean Nusselt number Num is valid. For the calculation of the mean Nusselt number Num it is necessary to determine the Nusselt numbers Nu1 (Equation 20) and Nu2 (Equation 21). Nusselt number 1
d Nu1 = 3,66 + 1,2 ⋅ i da
−0 ,8
di
outer diameter of inner tube (m)
da
inner diameter of outer tube (m)
0,006 Nu1 = 3,66 + 1,2 ⋅ 0,016
Nusselt number 2
Eq. 20
Nu2 = f g ⋅ 3 Re ⋅ Pr⋅
fg
−0 , 8
= 6,29
dh l
Eq. 21
factor for the differentiation of the three cases of heat transfer inside the concentric gap, it is calculated in Equation 22 [25]
Re
Reynolds number
107
Chapter 5 Secondary cycle system on the test bench
Pr
Prandtl number of water/monoethylenglycol 20%vol, Pr = 18,29 (at a mean temperature of 10°C, interpolated) [26]
dh
hydraulic diameter, equivalent to the characteristic length, dh = 1,5mm length of the tube, assumed to be 8m (to be
l
checked later)
For the calculation of the Nusselt number 2 the factor fg is needed which depends on the type of heat transfer inside the concentric gap. Three different cases of heat transfer are defined [25] with case number 1: heat transfer at inner tube, outer tube insulated, being suitable. In this case Equation 22 is valid.
factor fg
−0 , 5 di f g = 1,615 ⋅ 1 + 0,14 ⋅ d o
Eq. 22
di
outer diameter of inner tube (m)
do
inner diameter of outer tube (m)
−0 , 5 0,006 f g = 1,615 ⋅ 1 + 0,14 ⋅ = 1,9842 0,016
Nusselt number 2
Nu2 = 1,9842 ⋅ 3 1091,0 ⋅ 18,29 ⋅
0,0015 = 3,08 8
From Nusselt number 1 and 2, the mean Nusselt number can be calculated: 108
Eq. 21
Chapter 5 Secondary cycle system on the test bench
(
mean Nusselt number
3
Num = Nu1 + Nu2
(
)
1 3 3
1 3 3
Num = 6,29 + 3,08 3
)
Eq. 23
= 6,53
The next equation considers the dependency of the temperature on the property values used in the previous equations [25]. Nusselt number, temperature adjusted
Pr Nu = Num ⋅ Prw
0 ,11
Eq. 14
Num
Nusselt number, not temperature adjusted
Pr
Prandtl number at mean temperature of 10°C
Prw
Prandtl number at the wall of the tube for water/monoethylenglycol 20%vol, Pr (0°C) = 23,98 [26]
18,29 Nu = 6,53 ⋅ 23,98
0 ,11
= 6,34
Now, using the Nusselt number to calculate the outer heat transfer coefficient. heat transfer coefficient
α=
Nu ⋅ λ dh
Eq. 16
Nu
Nusselt number, concentric gap
λ
thermal conductivity of water/monoethylenglycol 20%vol, = 0,507W·m-1·K-1 (at a mean temperature of 10°C, interpolated) [26]
109
Chapter 5 Secondary cycle system on the test bench
dh
hydraulic diameter = characteristic length = 0,0015m
α=
overall heat transfer
6,34 ⋅ 0,507 = 2142,9 W ⋅ m − 2 ⋅ K −1 0,0015
1
k=
1
αo
coefficient
k=
+
1 δ + λcu α i
Eq. 8
1 1 0,001 1 + + 2142,9 372 2500
k = 1150,3 W ⋅ m −2 ⋅ K −1
Q& C k ⋅ ∆t
surface of evaporator
AV =
temperature difference
∆t = twater / glycol − t wall
Eq. 7
Eq. 17
∆t = (0 − (− 10 ))K = 10 K
surface of evaporator
AV =
4,5 = 0,3912 m² 1,1503 ⋅10
length of tube
ltube =
A π ⋅ dm ⋅ z
A
surface of evaporator (m2) 110
Eq. 7
Eq. 18
Chapter 5 Secondary cycle system on the test bench
dm
mean diameter of inner tubes (m)
z
number of inner tubes
ltube =
0,3912 = 8,30 m π ⋅ 0,005 ⋅ 3
The calculated length of the copper coaxial type heat exchanger was 8,30m which is close to the assumption that was used for the calculation of the Nusselt number using Equation 21. Thus, a recalculation is not necessary.
The heat exchanger was constructed to the calculated dimensions. Figure 45 illustrates the coaxial heat exchanger as it was installed on the test bench.
Figure 45: The refrigerant-to-coolant heat exchanger
111
Chapter 6 The tests conducted and results obtained
Chapter 6 The tests conducted and results obtained This chapter explains how the tests were performed and provides interpretations of the measured data.
As mentioned before, the compressor was a delicate prototype. The manufacturer estimated the life-cycle to be only 1000 hours and so in order to avoid any possibility of an early failure the manufacturer suggested an operating speed of not more than 2000 rpm.
Prior to the commencement of the test sessions, the refrigerant charge for the system was estimated to be 400g of CO2. The project partner, Valeo Klimasysteme® GmbH, developed a prototype CO2 air-conditioning system for a Mercedes® A-Class using the same set of heat exchangers as those on the test-bench and they reported a optimum charge level of 300g. The influence of the long refrigerant lines on the test-bench were estimated to be 100g and this amount of refrigerant was added to the original charge level to provide a total of 400g. The subsequent tests showed a good performance of the system at all possible operating points. As expected, the system layout with the low pressure accumulator turned out to be very tolerant to different charge levels, thus a time consuming charge determination test was not necessary. 112
Chapter 6 The tests conducted and results obtained
6.1 The control curves In order to control the CO2 refrigerant cycle in the tests, the effect of different adjustments made with the control devices on the system’s behaviour had to be investigated.
With adjustments to the compressor control valve it was important to observe the changes of the refrigerant mass flow, the low pressure and the crank case pressure. Changing the pressure ratio across the valve would result in a change in the refrigerant mass flow and with respect to the control characteristic of the expansion valve, it controlled the high pressure and had to be monitored.
Ideally for the investigation of this control characteristic, the data acquisition procedure, during a constant stepwise opening and closing of the valves, was not always possible. During some settings, the pressure and the temperature at the compressor discharge port reached levels that would cause damage to the compressor and, in those cases, large valve adjustments were needed.
6.1.1 The compressor control valve The test was started with the compressor control valve completely closed, i.e. the compressor works with minimum displacement. Every two minutes the control valve was opened at increments of 1/32 of a turn until the full displacement was reached following which the valve was closed. The graph displaying the result of the test is illustrated in Figure 46.
113
Chapter 6 The tests conducted and results obtained
Figure 46: Control curves: compressor control valve
114
Chapter 6 The tests conducted and results obtained
The graph illustrates four curves: the black curve shows the low pressure (p_V1), the red curve shows the high pressure (p_V2), the green curve shows the crank case pressure (p_CC) and the blue curve shows the refrigerant mass flow (Mass_CO2). All pressure curves refer to the left hand axis and the mass flow curve refers to the right hand axis.
On opening the valve, marked as position 1 on the graph, the system shows a slight increase in high pressure and CO2 mass flow. The crank case pressure and the low pressure start to show a characteristic sine curve. In comparison to the crank case pressure, the amplitude of the low pressure curve is smaller and the positions of minimum and maximum are opposed.
The sine curve presents the two position control characteristic of the overflow valve. As the opening of the control valve is very small, the continuous high pressure blow-by increases the pressure inside the crank case. At a certain pressure difference between the crank case pressure and low pressure, the overflow valve opens the crank case to the low pressure side and the crank case pressure decreases.
The decreasing crank case pressure causes a slight increase of pressure on the low pressure side and this happens at a pressure difference of 21bar (p_CC ≈ 50bar; p_V1 ≈ 29bar). The pressure difference than falls and the overflow valve closes again. In this case there was a pressure difference of 14bar (p_CC ≈ 45bar; p_V1 ≈ 31bar). As the overflow valve was closed again, the high pressure blow-by starts to increase the crank case pressure and at the same time the low pressure shows a slight decrease because it is no longer influenced by the higher crank case pressure.
115
Chapter 6 The tests conducted and results obtained
Following a further 1/32 turn of the control valve at position 2, the crank case pressure decreases and there is no longer a sine curve characteristic of the crank case pressure and low pressure. The low pressure remains constant at the same pressure level. This further opening of the valve caused only a slight increase in high pressure and refrigerant mass flow.
Following a further 1/32 turn of the control valve at position 3, the high pressure and the refrigerant mass flow show a slight increase, and, although there is a significant decrease in the crank case pressure (∆p ≈ 13bar), the low pressure remained constant.
The valve opening at position 4 caused a moderate decrease in crank case pressure. In contrast, the changes in high pressure and refrigerant mass flow were significant. The previous openings of the control valve caused a change in the refrigerant flow of approximately 2m³·h-1, but this time a difference of approximately 10m³·h-1 was realised and, in the case of the high pressure, an increase in pressure rise from 1bar to 13bar. Furthermore, there was a significant drop on the low pressure side, which, up to this point had remained constant.
The next steps marked with 5, 6 and 7 caused similar changes. At position 7, the low pressure and crank case pressure are equal, i.e. the full displacement of the compressor was reached with the refrigerant flow at 90m³·h-1.
At position 8, the control valve was closed by a 1/32 turn. The system appeared to have a hysteresis, with the high pressure and refrigerant mass flow remaining almost constant.
116
Chapter 6 The tests conducted and results obtained
At position 9, the control valve was closed completely, in order to protect the compressor from the high pressure and temperature at the compressor outlet. As the valve was closed completely, the curves for the crank case pressure and low pressure again take on a sine wave characteristic.
This test confirmed that there was no linear relationship between the control valve opening and the change in crank case pressure. In addition, there was no linear relationship between the crank case pressure and the change in refrigerant flow. The changes in refrigerant mass flow were analogous to the changes in high pressure. With the initial three turns (a small opening) of the control valve, the low pressure was not effected by the changes but at a defined valve opening (point 4), a decrease in the crank case pressure also caused a decrease in the low pressure.
6.1.2 The expansion valve The test was started with the expansion valve opened to an initial 1/8 turn with the compressor control valve completely closed and the compressor ran with minimum displacement. At intervals of two minutes the setting of the expansion valve was changed. Initially, the valve opening was increased in three steps, with each step being a 1/16 turn, following which the valve was then closed in three steps. Then, once again from the starting point, the compressor control valve was opened in two steps of 1/32 turns. Subsequently, the expansion valve was opened in two steps, with each step being a 1/16 turn. The graph displaying the result of the test is illustrated in Figure 47. The legend for the graph is the same as that for the control characteristic of the compressor control valve.
117
Chapter 6 The tests conducted and results obtained
Figure 47: Control curves: expansion valve
118
Chapter 6 The tests conducted and results obtained
The first opening steps marked as 1, 2 and 3 on the graph, caused a noticeable stepwise increase in the refrigerant flow with a stepwise decrease in the high pressure with the change also progressively decreasing. On the third adjustment of the valve, the high pressure remained constant.
During the first three valve opening steps, the crank case pressure showed a continuously decreasing sine wave characteristic, as was established for the compressor control valve. This effect was caused by the slight stepwise increase in the low pressure from position 1 to position 3. As the crank case pressure remained constant, the increasing low pressure caused a smaller pressure difference at the overflow valve and the amplitude of the sine curve was reduced.
The stepwise closing of the expansion valve from position 4 to position 6 caused the exact opposite curve progressions than during the opening of the valve from position 1 to position 3 and the control characteristic showed no hysteresis.
The two opening steps of the compressor control valve (position 7 and 8) caused obvious changes in the crank case pressure and the high pressure. There was an increase in the high pressure and a decrease in the crank case pressure. There was a fluctuation and some instability in the flow, but does show an increase. The opening step (position 7), shows no effect on the low pressure but with the next step, caused a slight decrease in the low pressure.
At position 9, the expansion valve was opened a 1/16 turn. This adjustment had a major effect on the refrigerant flow as it rose from 50m³·h-1 to 85m³·h-1. This change was greater 119
Chapter 6 The tests conducted and results obtained
than after the adjustment at position 1 because of the greater high pressure prior to the adjustment. Now, at point 9, the refrigerant flow increases and the high pressure decreases. This was caused by a shift of refrigerant from the high pressure side to the low pressure side and the system reacts like two refrigerant reservoirs connected by a valve. As the valve is opened refrigerant flows from the high pressure reservoir to the low pressure reservoir with the result that the pressure in the high pressure reservoir decreases, while the pressure in the low pressure reservoir increases. The low pressure showed a slight increase following the opening of the expansion valve and the crank case pressure remained relatively constant.
Further opening of the expansion valve at the position 10 had the same effect as the previous adjustment but that the changes were not as pronounced.
6.1.3 Conclusion to the controls In theory, the high pressure inside a CO2 refrigeration circuit is controlled by the expansion valve and the refrigerant flow by the compressor control unit. The control curves showed that the control of a real cycle is much more complicated as both control devices were influencing each other. An adjustment of the compressor control valve did not effect the refrigerant flow only, but at the same time, a change in the high pressure was also created. The adjustment of the expansion valve was intended to change the high pressure, but it also influenced the refrigerant flow. The following table provides an overview of how the different parameters were effected by the adjustments of the two control devices during the tests. The colours of the arrows refer to the colour of the curves on the graphs.
120
Chapter 6 The tests conducted and results obtained
Compressor Control Valve Expansion Valve
opening closing opening closing
HP
Mass CO2
pCC
LP
↑ ↓ ↓ ↑
↑ ↓ ↑ ↓
↓ ↑ ↔ ↔
↓ (↔) ↑ (↔) ↑ ↓
Table 7: An abridged overview of the control characteristics for the system
Legend:
↑
increase
↓
decrease
↔
no change
↑ (↔) slight increase ↓ (↔) slight decrease
6.2 The air-conditioning mode In this section, the behaviour of the system in air-conditioning operation is described and discussed. The trends of the most important variables are illustrated in three different graphs. The first graph, Figure 48, shows the characteristics of the main system pressures versus time, and the second, Figure 49, the temperatures versus time. Both graphs include the refrigerant volume flow rate. The third graph, Figure 50, shows the temperatures at the internal heat exchanger.
The tests were performed at a compressor speed of 1100rpm with the compressor control valve fully opened and the expansion valve opened to 1/8 of a turn. The ambient temperature was 23°C and the blower inside the HVAC box discharged a volumetric air flow of 222m³·h-1. 121
Chapter 6 The tests conducted and results obtained
Figure 48: Air-conditioning operation: pressure and volume flow
122
Chapter 6 The tests conducted and results obtained
Figure 49: Air-conditioning operation: system temperatures and volume flow
123
Chapter 6 The tests conducted and results obtained
In Figure 48, from start-up, the curves quickly reach full operational condition. The low and the crank case pressures achieved the same pressure level, which indicates that the compressor control valve was fully opened and the compressor was at full displacement. By the fourth minute the curves seem to have settled, except for the high pressure curve which was still rising. At 6,5 minutes, the decrease in high pressure and increase in refrigerant flow indicated a further opening of the expansion valve which was done in order to keep the high pressure under 100bar.This had to be repeated at 11 and 12 minutes. With each opening, the low pressure and the crank case pressure increased slightly.
In Figure 49, the temperature curves refer to the primary y-axis and the mass flow curve refer to the secondary y-axis. The red curve displays the temperature at the gas cooler inlet, the yellow curve shows the temperature characteristic at the outlet. The temperature in the evaporator is illustrated by the black curve, and the air temperature at the outlet of the HVAC box by the orange curve. As in the previous graph, the blue curve shows the refrigerant flow. The temperature curves show a similar characteristic as that of the pressure curves in Figure 48. At 5 minutes the system had settled, except for the gas cooler inlet temperature which was still rising. The three opening steps (1/16 turn each) of the expansion valve caused only slight changes in the temperatures. After twenty minutes the system showed a steady-state condition. A CO2 volume flow of 86m³·h-1 at an evaporation temperature of –2°C created an air outlet temperature of 6°C. In the gas cooler, the high pressure refrigerant was cooled from 105°C to 30°C. With a temperature difference of 7K between the gas cooler outlet temperature (30°C) and the ambient temperature (23°C) which indicated an adequate effectiveness of the gas cooler as good gas cooler designs should reach a temperature difference between 5K and 7K.
124
Chapter 6 The tests conducted and results obtained
Figure 50: Air-conditioning operation: temperatures at internal HX
125
Chapter 6 The tests conducted and results obtained
The third graph is illustrated in Figure 50 and shows the temperatures at the four ports of the internal heat exchanger. The yellow curve shows the high pressure inlet, the red one the outlet, the green curve displays the low pressure inlet and the blue curve the outlet.
All the temperatures reach the operating level very quickly and at 3 minutes all the temperatures have nearly settled, but for a slight rising. The opening steps of the expansions valve caused only minor changes in the temperature curves. Both high pressure temperatures increased and the low pressure inlet temperature decreased and the low pressure outlet temperature remained unaffected. At 15 minutes, a steady-state condition was reached. The high pressure refrigerant flow through the internal heat exchanger was cooled from 29°C to 27°C by increasing the temperature at the low pressure side from 13°C to 23°C.
The decrease in temperature on the high pressure side of only 2K was very little. The reason for this was not the poor performance of the heat exchanger, as a temperature difference of only 4K after the heat exchanging process indicated a good effectiveness. So, the main reason for the marginal cooling of the high pressure refrigerant was the fairly high inlet temperature of the low pressure refrigerant. At an evaporation temperature of –2°C, an evaporator outlet temperature of 13°C indicated a high degree of superheat of the refrigerant. By increasing the speed of the compressor it was possible to deliver a higher refrigerant flow to the evaporator and hence, reduce the degree of superheat.
126
Chapter 6 The tests conducted and results obtained
6.3 The heat pump mode In this section, the characteristics of the variables during the heat pump operation are illustrated and explained. As in the previous section the most significant variables are displayed on two different graphs. One graph shows the characteristics of the main system pressures and the other the temperatures versus time. However, in this case the characteristics of the temperatures at the internal heat exchanger are disregarded because in heat pump operation the internal heat exchanger is of no consequence to the operation of the system.
This test was performed at a compressor speed of 1500rpm with the compressor control valve fully open. At the start of the test, the expansion valve was opened to 1/16 of a turn and then during the first three minutes of the test it was opened a further four steps until an opening of 1/8 of a turn was reached. The ambient temperature in this test was 21°C and the blower inside the HVAC box discharged a volumetric air flow of 222m³·h-1.
Figure 51 shows all the pressure versus time curves on the primary y-axis and the refrigerant flow versus time curve on the secondary y-axis. The position of sensor p_o2 in air-conditioning operation is the outlet port of the evaporator. During heat pump operation, the position of sensor p_o2 is the inlet of the same heat exchanger which now serves as a gas cooler. The low pressure at the compressor inlet is represented by the black curve and the crank case pressure by the green curve. The positions of these sensors do not change when the cycle is reversed. The refrigerant volume flow is shown as a blue curve. Only from a time period of 10 minutes does the graph displays curves and, therefore, the curves do not start at a uniform level.
127
Chapter 6 The tests conducted and results obtained
Figure 51: Heat Pump operation: pressures and volume flow
128
Chapter 6 The tests conducted and results obtained
Figure 52: Heat Pump operation: system temperatures and volume flow
129
Chapter 6 The tests conducted and results obtained
From the start of the test at 10 minutes until the 14th minute, the curve of the refrigerant flow showed the characteristic stepwise increase which indicates the opening steps of the expansion valve. With every opening step, the high pressure curve shows a sudden drop in pressure followed by a recovery. The low and the crank case pressures are at the same levels which indicates that the compressor control valve was fully opened.
With every opening step of the expansion valve, the low pressure increases to a new level. The crank case pressure shows the same characteristic as the low pressure because both sensor’s positions are directly connected as the compressor control valve is completely opened.
At 20 minutes a steady-state condition was reached. The low pressure and the crank case pressure were at 27bar and the high pressure was at 105bar. The refrigerant flow reached a value of 80m³·h-1.
In Figure 52, the temperature curves refer to the primary y-axis and the refrigerant flow curve refers to the secondary y-axis. The red curve displays the temperature at the evaporator outlet, which is now the gas cooler inlet, and the yellow curve shows the temperature characteristic at the evaporator inlet, which is now the gas cooler outlet. The temperature of the gas cooler outlet which is now the evaporator inlet, is illustrated by the black curve. The purple curve displays the temperature of the gas cooler inlet, which is now the evaporator outlet. As was the case in the previous graph the orange curve shows the air temperature at the outlet of the HVAC box and the blue curve shows the refrigerant flow.
130
Chapter 6 The tests conducted and results obtained
Besides the evaporator outlet temperature, all the temperature curves show a moderate increase from the 11th minute to the 14th minute. These temperatures increased in a stepping characteristic, corresponding to the opening steps of the expansion valve but for the evaporator outlet temperature which remained constant for the whole period. At 20 minutes, the system had completely settled. A refrigerant flow of 80m³·h-1 was cooled in the gas cooler from 127°C to 50°C while the ambient air flowing through the HVAC box was heated from 21°C to 44°C. Simultaneously, the CO2 evaporates at 0°C and leaves the heat exchanger at 20°C.
The heat exchanger which was used for the evaporation provided a very good performance under these conditions. At the outlet, the difference of refrigerant temperature to the ambient temperature was only 1K (20°C – 21°C) and the heat exchanger inside the HVAC box also showed good efficiency as the temperature difference between the refrigerant outlet and the air flow was only 6K (50°C – 44°C). In order to increase the outlet temperature of the air flow through the HVAC box, the compressor outlet temperature or the refrigerant mass flow has to be increased. Higher temperatures at the compressor outlet were not possible because a further increase would damage the compressor and the compressor was already set at full displacement and the mass flow could only be increased by a higher compressor speed.
6.4 Reversing from AC to HP mode This section describes the comparison of the system before and after the modification was made to address “Flash Fogging” as the cycle is reversed. Chapter 4 explained that the “Flash Fogging” phenomenon depends on the dew point temperature of the air which was
131
Chapter 6 The tests conducted and results obtained
blown onto the windscreen. During the tests the temperature and the relative humidity of the discharged air flow were recorded and the dew point temperature was calculated from these values.
To ensure that the surface of the heat exchanger inside the HVAC box was covered with condensate, the system was left running in air-conditioning operation until water could be seen leaving the condensate drain of the HVAC box. The compressor was switched off and the 4-way valve was switched to heat pump position and the expansion valve for the airconditioning operation was fully opened, and the expansion valve for the heat pump mode was opened to 1/8 of a turn, and the compressor was switched on.
6.4.1 Reversing without the secondary cycle This test was performed at a compressor speed of 2000rpm and a volumetric air flow through the HVAC box of 222m³·h-1. The condition of the ambient air was 25°C and 44%RH, with a corresponding calculated dew point of 12°C.
Figure 53 illustrates the reversing of the cycle after 10 minutes in air-conditioning operation. The graph shows the characteristic of the air temperature (orange curve), the relative humidity (blue curve), and the inlet (black curve) and outlet (red curve) temperatures of the evaporator.
132
Chapter 6 The tests conducted and results obtained
Figure 53: Reversing the cycle without the secondary cycle
133
Chapter 6 The tests conducted and results obtained
The dew point temperature was calculated from the values “Humidity_temp_1” and “Humidity_1”. The result is illustrated in Figure 54.
Figure 54: Dew point temperature without the secondary cycle
During the air-conditioning operation the dew point temperature was around 4°C. After the compressor was stopped (at 10 minutes) the dew point temperature increased to 16°C. When all the valves were adjusted and the compressor was restarted (at 11 minutes), the dew point temperature further increased to a maximum of 25°C. In this state the dew point of the discharged air was 13K higher than that of the ambient air drawn in. The dew point temperature of the air discharged then decreased and settled at 13°C. The total time taken to reduce the dew point temperature to 13°C, was 3,25 minutes.
Assuming that the windscreen temperature was at the ambient temperature, this characteristic of the dew point temperature would cause Flash Fogging because at the
134
Chapter 6 The tests conducted and results obtained
maximum point of the curve the dew point temperature and the ambient temperature are equal.
6.4.2 Reversing with the secondary cycle As the system was equipped with a secondary cycle the HVAC box contains two heater cores which can be used either for cooling or for heating. Figure 55 illustrates the arrangement with both heat exchangers inside the HVAC box.
Figure 55: Cross-sectional view of the HVAC box
As mentioned in section 5.1 the most reasonable operating method to avoid Flash Fogging was to use the heat exchanger in the heater core position for cooling and the other for heating and, thus, the heat exchanger covered with condensate can be separated from the air flow during the heat pump operation. In the following text, this operating method was called “heat exchanger arrangement 1”. In addition to this method, the opposite arrangement of the heat exchangers was also investigated and this operation was called “heat exchanger arrangement 2”.
135
Chapter 6 The tests conducted and results obtained
6.4.2.1 Heat exchanger arrangement 1
This test was performed at a compressor speed of 2000rpm and a volumetric air flow through the HVAC box of 113m³·h-1. The condition of the ambient air was 35,5°C and 30%RH with the calculated dew point being 15°C.
Figure 56 illustrates the process and it consists of four curves. The blue curve shows the relative humidity at the air outlet of the HVAC box, and the orange curve displays the temperature at the same position. The coolant temperature at the inlet and the outlet of the coaxial heat exchanger are represented by the black and red curves, respectively.
The test commenced in air-conditioning operation, and the heat exchanger in the heater core position was used for cooling. Due to the poor performance of this heat exchanger the air flow through the HVAC box was maintained at 113m³·h-1 because at higher flow rates unsufficient cooling of the air occurred and no condensate was produced. An indication of the poor heat exchanger performance was the large temperature difference of 15,5K between the coolant which leaves the heat exchanger and the air discharged. The coolant was heated from –5,5°C to 2,5°C and the air temperature was reduced from 35,5°C to 18°C.
Following 20 minutes in air-conditioning operation droplets of water at the condensate drain of the HVAC box indicated that the surface of the heat exchanger in the heater core position was covered with condensate.
136
Chapter 6 The tests conducted and results obtained
Figure 56: Reversing of the cycle with a secondary cycle to arrangement 1
137
Chapter 6 The tests conducted and results obtained
Immediately after the compressor was switched off, the flaps inside the HVAC box were closed in order to separate the wet heat exchanger from the air flow and all the valves were switched to heat pump operation, and the compressor was restarted at 21,5 minutes. At 22,5 minutes, the valves of the coolant circuit were switched to connect the heat exchanger in the evaporator position to the system as the coolant temperature was already at 30°C. This was indicated on the graph as a short peak of the inlet temperature of the coaxial heat exchanger. Figure 57 illustrates the change in operation mode on the dew point temperature of the discharged air.
Figure 57: Dew point temperature with secondary cycle arrangement 1
On review of Figure 57, during the air-conditioning operation, the dew point temperature was at 14,2°C, which was slightly below the dew point temperature of the ambient air, which was 15°C, and indicates that a dehumidification of the air took place. After the compressor was stopped the curve showed a short peak to 15°C and then decreased to 13°C, from which the dew point temperature increased to 14,7°C at 22,75 minutes. A half 138
Chapter 6 The tests conducted and results obtained
minute before this moment the heat exchanger in the evaporator position was connected to the coolant cycle and as a consequence at 23 minutes the ambient temperature began to rise. This procedure caused a short drop in the dew point temperature curve but at 24 minutes it began to increase and settled at 18,5°C after 32 minutes.
This characteristic of the dew point temperature during the process of reversing the cycle would not cause Flash Fogging. At the maximum dew point temperature there was still a difference of 17K between the dew point and ambient temperature. Although the wet heat exchanger in the heater core position was separated from the air flow on reversing the cycle, the dew point temperature of the air increased from 15°C to 19°C by passing the HVAC box. This indicated that some moisture was picked-up by the air obviously because of insufficient sealing of the flaps inside the HVAC box. 6.4.2.2 Heat exchanger arrangement 2
This test was performed at a compressor speed of 2000rpm and a volumetric air flow through the HVAC box of 146m³·h-1. The condition of the ambient air was 35°C and 32%RH with the calculated dew point being 16°C. Figure 58 illustrates the process and the legend is the same as the graph for arrangement 1 in the previous section.
In this arrangement of the heat exchangers inside the HVAC box, the heat exchanger in the evaporator position was used for cooling during the air-conditioning operation. In contrast to the opposite arrangement, the performance of the heat exchanger during the airconditioning operation was better and the air flow rate could be increased to 146m³·h-1.
139
Chapter 6 The tests conducted and results obtained
Figure 58: Reversing of the cycle with secondary cycle to arrangement 2
140
Chapter 6 The tests conducted and results obtained
In this case the graph showed a smaller temperature difference of only 10K between the coolant which leaves the heat exchanger and the air discharged. As the coolant was heated from -2,5°C to 10°C, the air temperature was reduced from ambient condition which was 35°C to 14°C at the air outlet of the HVAC box.
At 20 minutes, after some condensate was drained, the compressor was stopped and the cycle was reversed. At 21 minutes, the compressor was restarted, and one minute later the heat exchanger in the evaporator position was disconnected from the coolant circuit and the heat exchanger in the heater core position was connected. This was indicated on the graph as a short peak in the inlet temperature of the coaxial heat exchanger. Figure 59 illustrates the effect of the change in operation mode on the dew point temperature of the air discharged.
Figure 59: Dew point temperature with secondary cycle arrangement 2
141
Chapter 6 The tests conducted and results obtained
During the air-conditioning operation, the dew point temperature was 11,7°C. The difference in the dew point temperature of the ambient air of 16°C, was greater than in the previous test and indicates a better dehumidification of the air. On stopping the compressor the curve showed a short peak to 14°C followed by a drop to 10,3°C, from which the dew point temperature shows a steep rise to 20,3°C at 22 minutes. At this moment the heat exchanger in the heater core position was connected to the coolant circuit and, as a consequence, the ambient temperature started to rise. This procedure caused a short decrease in the dew point temperature curve to 18,6°C and at 23 minutes it fell again to 17,5°C. At 26 minutes the dew point shows a gradual increase and settles at 31 minutes at 19,5°C.
The biggest difference between this characteristic and the one in the previous test was the steep increase in the dew point temperature between stopping the compressor and the moment when the heat exchanger in the heater core position was connected to the coolant circuit. At the peak dew point temperature of the discharged air, it was still 14,7K below the ambient temperature and, hence, this arrangement of the heat exchangers would not cause a Flash Fogging problem when the cycle is reversed. When passing through the HVAC box, the dew point temperature of the air flow increased from ambient condition at 16°C to a maximum of 20,3°C at the outlet of the HVAC box. 6.4.2.3 Conclusion to the various arrangements and Flash Fogging
The performance of the secondary cycle system was expected to be limited. The reason for this was one additional heat exchange process and the energy loss from the coolant line to ambient air. However, the most limiting factor was the efficiency of both coolant-to-air heat exchangers in the HVAC box during the air-conditioning operation as they both suffer
142
Chapter 6 The tests conducted and results obtained
from a characteristic small temperature difference when used for cooling. They were designed for heating with a temperature difference twice or thrice as much as that for cooling. Despite both heat exchangers being identical, the one in the evaporator position showed a better performance. The final difference was their positioning inside the HVAC box and, therefore, unequal approaching air flows. The design of the HVAC box had a negative effect on the heat exchanger performance. If this HVAC box was used in a conventional system, this constraint would have to be compensated for with a higher temperature difference across the heat exchanger. Obviously, the negative effect on the approaching air flow to the heater was accepted in favour of a more compact design of the HVAC box.
An unexpected result from the tests was the small difference concerning the dew point increase during the process of reversing the cycle between arrangement 1 and arrangement 2. From a theoretical point of view, arrangement 1 appeared to be a superior solution in order to achieve a low dew point temperature of the air discharged because the wet heat exchanger could be separated from the air flow. The characteristic of the dew point temperature during the reverse process shows that an increase in absolute humidity of the air flowing through the HVAC box cannot be obviated, only delayed. The reason for this is that the flaps inside the HVAC box do not close tightly enough to separate the condensate completely from the air flow. In a completely sealed system the dew point temperature of the air at inlet and outlet of the HVAC box should be the same. Even in arrangement 1 there was an increase of the dew point temperature and it was obvious that condensate was re-evaporated into the air flow.
143
Chapter 6 The tests conducted and results obtained
The characteristic of the dew point temperature of the test for arrangement 2, showed a different trend to that in arrangement 1, but the effect concerning the prevention of Flash Fogging was very similar. The prevention of Flash Fogging lies in the difference between the peak dew point temperature and ambient temperature. A large temperature difference means increased safety against Flash Fogging while a small difference translates to reduced safety. Normally, Flash Fogging occurs as the dew point temperature reaches or exceeds the ambient temperature. In some cases, Flash Fogging can occur below the ambient temperature level as the windscreen is cooled by rain or the airstream and for this reason a factor of safety in the form of a substantial temperature difference between dew point temperature and ambient temperature should be provided. In case of arrangement 1, this difference at peak level of the dew point temperature was 17K. Despite the wet heat exchanger being positioned directly in the air flow, the difference between the dew point temperature and ambient temperature in arrangement 2 was still only 14,7K.
This observation shows that the decisive measure against Flash Fogging was not the separation of the wet heat exchanger but the use of different heat exchangers for heating and cooling.
144
Chapter 7 Future Opportunities
Chapter 7 Future Opportunities Despite the compromises regarding the performance of the coolant-to-air heat exchanger and the limitations of the serial type HVAC box, the tests showed that the secondary cycle system performed very well concerning the prevention of the Flash Fogging phenomenon. When it comes to serial production some detailed problems have to be solved. First of all, a more compact design for the refrigerant-to-coolant heat exchanger has to be found. As mentioned before, this is possible by using multi-port profiles. Another weakness of the tested system was the performance of the coolant-to-air heat exchangers in the cooling operation. More capable designs have to be developed. Furthermore, the design of the HVAC box regarding the approaching flow to the heat exchangers and the sealing of the flaps has to be improved.
As these problems are solved, one major problem remains: the reheat operation. The secondary cycle system suffers from the fact that only cooling or heating can be achieved. Cooling of the air with subsequent heating was not possible. In order to solve this problem the secondary coolant cycle has to be connected to the engine coolant circuit. In reheat operation, one heat exchanger inside the HVAC box is fed by cold coolant provided by the
145
Chapter 7 Future Opportunities
refrigerant-to-coolant heat exchanger and the other one with warm coolant from the combustion engine.
In a further developmental step these two cycles could be combined to form a total thermal management system. The core of this system would be a CO2 refrigerant cycle with two refrigerant-to-coolant heat exchangers, one working as a gas cooler and the other as an evaporator. In the HVAC box, two coolant-to-air heat exchangers as in the tested system are needed. The radiator in front of the car has to be separated in a high and a low temperature section. The widely ramified cycle is driven by two electrical pumps. The coolant flow through the cycle is directed and controlled by five 3-way solenoid valves, one 2-way solenoid valve and two 3-way variable control valves.
The function of the system is best explained by describing the sequence of processes taking place inside the system during different operation modes. There are draft system layouts in the different operation modes, in the appendices. All components through which the coolant flows during the described operation mode are coloured, with all others in black. Red indicates the coolant flow coming from the combustion engine, the coolant flow through the gas cooler is marked in orange and the one through the evaporator is shown in green.
Heater only operation (Appendix A1)
In this operation mode the refrigerant cycle is not running. In order to prevent undercooling of the engine the coolant flow from the engine (marked in red) is split using a 3-way variable control valve. One portion of the warm coolant flows back to the engine while the other portion is directed to heat exchanger 1 inside the HVAC box and the 146
Chapter 7 Future Opportunities
coolant heats the air flowing through the HVAC box and thereafter flows back to the engine.
Maximum cooling operation (Appendix A2)
During this operation mode, the cold coolant flow from the evaporator is directed to heat exchanger 1 inside the HVAC box. A 3-way solenoid valve connects the outlet of heat exchanger 1 with the inlet of heat exchanger 2, with both heat exchangers used to cool the air flow through the HVAC box. The coolant flow from the gas cooler is directed to the radiator and at the same time, this component is used to cool the coolant from the engine. The coolant which is returned to the engine leaves the radiator at the high temperature outlet. The coolant which is intended to cool the gas cooler leaves the radiator at the low temperature outlet. Separation of the radiator is required because the gas cooler has to be cooled to a much lower temperature than the combustion engine.
Reheat (Appendix A3)
In this operation mode, heat exchanger 1 inside the HVAC box is connected to the engine coolant flow. The surplus of hot coolant is split at the 3-way valve and is cooled inside the hot temperature section of the radiator. Heat exchanger 2 is fed by the coolant from the evaporator which is used to cool and dehumidify the air flow through the HVAC box, following which the air flow is heated again by crossing over heat exchanger 1. The hot coolant from the gas cooler and the coolant from the engine use the radiator as a heat sink. As explained in the previous example, the coolant to the engine leaves the radiator at the high temperature outlet and the coolant to the gas cooler at the low temperature outlet.
147
Chapter 7 Future Opportunities
Heat pump with engine coolant as the heat source (Appendix A4)
In this operation mode, the coolant from the gas cooler is directed to heat exchanger 2, which heats the air flow through the HVAC box and then flows back to the gas cooler. The evaporator is heated by the warm coolant from the engine and as the coolant leaves the evaporator at a lower temperature it is directed back to the engine.
Heat pump with ambient air as the heat source (Appendix A5)
As in the previous example, heat exchanger 2 is connected to the gas cooler and heating of the air is achieved through the HVAC box. However, in this case the evaporator is not connected to the engine coolant circuit but to the radiator. The coolant flowing through this component picks up heat energy from the ambient air and transfers it to the evaporator. In this operation mode, the engine coolant is not connected to the system.
Heater and heat pump with ambient air as the heat source (Appendix A6)
This operation mode is similar to the previous example with the evaporator and gas cooler connected in the same manner as before. In this case, the engine coolant cycle is connected to heat exchanger 1 which provides additional heating of the air flow through the HVAC box. From the heat exchanger the coolant is returned to the engine to be heated again.
Conclusion
With a suitable control unit controlling the combination of the heat sources and heat sinks inside the car, the introduced system is able to provide a maximum of passenger comfort level irrespective of the ambient condition and the state of operation (temperature, load, speed) of the combustion engine. The system combines the advantages of a heat pump system (quicker heat up at low ambient conditions) with the advantages of a conventional 148
Chapter 7 Future Opportunities
system (reheat operation possible). Furthermore, by using two heat exchangers for maximum cooling it compensates for the lack of performance simple secondary cycle systems like installed on the test-bench have in this operation mode.
As a secondary effect, a systems like this reduces attrition and exhaust gas emission of the combustion engine because, during the heat up phase of the engine no heat is transferred to the cabin and, hence, it reaches the optimum operation temperature in a shorter period of time (with ambient air as heat source of the heat pump).
149
List of references
List of references
[1] Schwarz, W. 2001. Emissionen des Kältemittels R134a aus mobilen Klimaanlagen (Emissions of refrigerant R134a from mobile air-conditioning systems), Research for the German Department of the Environment, September 2001.
[2] Meeting Summary of the SAE Automotive Alternative Refrigerant Systems Symposium 1999, pp. 1 – 5.
[3] Konz, M., Holdack-Janssen, H. 2000. Ersatzkältemittel für Fahrzeugklimaanlagen (Alternative Refrigerants for Automotive AC Systems), lecture during DKV-conference Bremen.
[4] Goedhart, J. 2000. Assertions about car air conditioners by Dieckmann and Magid, Technical report, Greenchill Technology Association Inc., Australia, February 2000.
[5] Fröhling, J. 2000. CO2 as Refrigerant for A/C and Heat Pump Operation, SAE Automotive Alternate Refrigerant Systems Symposium.
150
List of references
[6] Neitzel, A. 2002. Untersuchung von inneren Wärmetauschern für Kfz-Klimaanlagen mit dem Kältemittel R744 (Investigation of different internal heat exchangers for automotive air-conditioning systems with R744 as refrigerant); Diploma Thesis Fachhochschule Braunschweig/Wolfenbüttel, p. 4.
[7] Werthenbach, J., Maue, J. 1996. Klimakälteanlagen mit CO2 im Pkw, lecture during Steinbeis-conference about air conditioning of vehicles with natural refrigerants; Karlsruhe Germany, March 8th, 1996.
[8] Köhler, J., Sonnekalb, M., Lauterbach, B. 1998. Kohlendioxid als Kältemittel für BusKlimaanlagen (Carbon dioxide as refrigerant for AC systems in motor-coaches); Ki Luftund Kältetechnik, no. 34, pp. 194 – 197.
[9] Liao, S., Jakobsen, A. 1998. Optimal heat rejection pressure in trans-critical carbon dioxide air-conditioning and heat pump systems, IIF-IIR Oslo, pp. 301 – 310.
[10] Heyl, P. 2003. Der effektive Einsatz des inneren Wärmeübertragers in transkritischen CO2-Prozessen (The effective use of the internal heat exchanger in transcritcal CO2 processes); Ki Luft- und Kältetechnik, no. 8, pp. 344 – 348.
[11] Reichelt, J. 2001. Design of a Liquid–Line Receiver, Meeting summary of the conference “Basics of Automotive AC systems with CO2”, July 9th – 11th 2001, pp. 5 - 10
151
List of references
[12] Burkhardt, C. 2002. Flexible Metal Hoses for the Guidance of CO2 in AirConditioning Systems of Passenger Cars, VDA Alternate Refrigerant Wintermeeting.
[13] Meeting summary of the 7th “LuK Kolloquium”, April 11th/12th 2002, pp. 127 – 128.
[14] Hoffmann, J. 1998. Taschenbuch der Messtechnik (Pocketbook of measurement technology); Leipzig: Fachbuchverlag Leipzig, pp. 132 – 135.
[15] Storm, R., Kolahi, K., Röck, R. 2001. Model based correction of Coriolis mass flow meters; lecture during Instrumentation and Measurement Technology Conference; Budapest Hungary, May 21th – 23th, 2001.
[16] Cengel, Y., Boles, M. 1998. Thermodynamics – an engineering approach; 3rd edition, Boston: WCB/McGraw-Hill, pp. 725 – 728.
[17] Baumgarth, S., Hörner, B., Recker, J. 2000. Handbuch der Klimatechnik (Handbook of Air-Conditioning Technology); 4th edition, Heidelberg: C.F. Müller Verlag, p. 195.
[18] Volkswagen AG, 2000. Self Study Program No. 208: “Klimaanlagen im Kraftfahrzeug (Air-Conditioning for Vehicles)“, p. 4.
[19] Mager, R., Hammer, H., Wertenbach, J. 2002. Comparative study of AC- and HPsystems using the refrigerants R134a and R744, lecture during VDA Alternate Refrigerant Wintermeeting 30th – 31st 2002, Saalfelden, Austria. 152
List of references
[20] Robert Bosch GmbH, 2000. Automotive Handbook; 5th edition, Stuttgart: Robert Bosch GmbH, pp. 774 – 775.
[21] Kampf, H., Krauß, H.-J., Feuerecker, G., Walter, C., Parsch, W., Rinne, F. 2001. CO2 als alternatives Kältemittel (CO2 as alternative Refrigerant), PKW-Klimatisierung II, Essen: Expert Verlag. pp. 52 – 81.
[22] Cerbe,G., Hoffmann, H-J. 1994. Einführung in die Thermodynamik (Introduction into thermodynamics), München: Carl Hanser Verlag. p. 355.
[23]Cerbe,G., Hoffmann, H-J. 1994. Einführung in die Thermodynamik (Introduction into thermodynamics), München: Carl Hanser Verlag. p. 354.
[24]VDI Wärmeatlas, 7th edition, chapter Dea 11 “Material properties of pure metals and alloys”.
[25] VDI Wärmeatlas, 7th edition, chapter Gd “Heat Transfer in a Concentric Gap”.
[26] VDI Wärmeatlas, 7th edition, page Dd 17 “Properties of Technical Heat Transfer Mediums”.
[27] VDI Wärmeatlas, 7th edition, page Gb7 “Heat Transfer During a Flow Through a Tube”.
153
Appendices
Appendices
154
Appendices
A1 Improved cycle: heater only operation
155
Appendices
A2 Improved cycle: maximum cooling operation
156
Appendices
A3 Improved cycle: reheat operation
157
Appendices
A4 Improved cycle: heat pump (engine coolant as the heat source)
158
Appendices
A5 Improved cycle: heat pump (ambient air as the heat source)
159
Appendices
A6 Improved cycle: heater & heat pump
160
