Home

Open Issues in Control of Automotive R744 Air

1. 5000 y 286 r T One evaporator 4000 3 284 Rear evaporator 5B 3 mw Front evaporator 5 E 3 3000 a 2 282 F E E 2000 280 8 o 3 1000 S 278 o o o gt 0 i U 276 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 Time Time Figure 5 4 Cooling power distribution between two evaporators in comparison with one evaporator system Figure 5 5 shows the cooling power distribution against the ambient temperature At higher temperatures the pattern of distribution will change but it still provides acceptable cooling power for both evaporators 93 4000 y 284 r gt One evaporator 3500 p 1 5 282 Rear evaporator 3000 T Front evaporator i He 280 amp 2500 7 Hi Mig n n o e o g z 2000 278 5 e 1500 bu 5 5 S 276 1000 p potes gue 7 2 E 500 274 305 310 315 320 325 330 305 310 315 320 325 330 Ambient temperatue K Ambient temperatue K Figure 5 5 Cooling power distribution between two evaporators in comparison with one evaporator system Rear AirFlow Effect As we said earlier it is possible to change the air mass flow around the evaporator in order to change the cooling power The simulation has run in a limited range of airflow variation under two different cooling loads Figure 5 6 shows the change of the rear evaporator cooling power when the air mass flow is changed at 5000 second Figure 5 7 shows the cooling po
2. ION Optimized cycle I Optimized cycle cycle with two stage valve cycle with two stage valve 3 r r 7000 _ 6500 2 2 5 p 6000 1 a Q a 5500 E 2 o Oo 4500 p J 1 5 i 4000 i 300 305 310 315 320 325 300 305 310 315 320 325 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 12 Comparison of the two valves case 2 taii Optimized cycle I Optimized cycle cycle with two stage valve cycle with two stage valve 13 0 06 i E 0 055 ho i MON M MEME M O 1 2 peee i B n ee ere dad o BE E e 0 05 1 1 3 et 4 9 0 045 a es 3 amp 004 1 P E E s See xx p E 0 035 0 9 0 03 300 305 310 315 320 325 300 305 310 315 320 325 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 13 Comparison of the two valves case 2 38 As has been seen in this section for lower ambient temperatures COP of the cycle with two stage orifice is up to 4096 less than ideal cycle therefore it is suggested that in this range of ambient temperatures the evaporator temperature is kept as high as possible to improve the COP 4 3 2 Limit Cycle of Two stage Orifice In some operating points which result in a higher pressure difference than 73 bars as a consequence of the rising pressure the bypass starts to open and decreases the high pressure again the decrease in high pressure causes the closing of the bypass and this limit cycle continue until one of the inputs alter
3. Low pressure o Heat rejection capacity Balance point 12 Qer on the refrigerant side Qca on the air side Ph High pressure Figure 2 6 Balance of low pressure and high pressure The intercept of the Qer line in Figure 2 6 is changing by the inlet enthalpy which is a function of the high pressure The higher the high pressure the lower the refrig eration capacity on the refrigerant side Qer becomes Therefore depending on the high pressure balance point location the low pressure balance point location varies Figure 2 7 Therefore the pressure and capacity at the balance point in the refriger ation cycle are obtained when the high pressure low pressure and cooling power are balanced yor LE c I i Qea on the air side I I Cooling power Low pressure Heat rejection capacity Qca on the air side Qer on the refrigerant side High pressure Figure 2 7 Balance of refrigeration cycle 13 As indicated by the equation 2 3 6 air properties around the evaporator also affect the position of Q line in Figure 2 6 and balance point The effect of the airflow rate on low and high pressure balance is explained briefly here When the airflow rate around the evaporator increases the refrigeration capacity on the air side increases and this leads to higher low pressure on the refrigerant side and changes the balance point location as indicated in Figure 2 8 Therefore the refrig
4. o 4 3 Two stage Orifice Valve a det rem a EEN 4 3 1 Performance of the Valves go 4 2 e us Bs xum Bowe ded 4 3 2 Limit Cycle of Two stage Orifice AA Behr Alternative Controller cuisine RR Bee Rw aR a iv iv vi o 00 Q A A 11 14 16 18 18 23 25 5 Four Zone Air Conditioning 5 1 Two Evaporator System Model 52 MOS a A A doctr m NA d COP no ott 6 Summary and Conclusions References 49 51 92 55 57 Acknowledgements This thesis was carried out at Modelon AB in Lund Sweden Models of the prototype components have been provided by DaimlerChrysler Germany I should hereby thank the Modelon team for making my stay in Lund very pleas ant most of all I appreciate the great supervision and support of Dr Hubertus Tummescheit Besides my thesis he provided me with an opportunity to learn about object oriented modeling I would also like to thank Dr Par Samuelsson my local supervisor at Department of Electrical Engineering at Dalarna University and Dr Bj rn Sohlberg for their kind guidances during my studies in Sweden vi Chapter 1 Introduction Under the Kyoto protocol agreement by the year 2012 industrialized countries must reduce their collective emissions of greenhouse gas 5 below their 1990 levels Since the current refrigerant used in vehicles R134a has a GWP Global Warming Poten tial of 1410 R744 CO2 technology has been proposed
5. airin _evap l 7 1751 duration 1 64 433 creep COP 29 7 8 5 12 k init phiEvap Figure 3 6 A C library template for R744 refrigeration cycle Figure 3 6 shows a generic R744 refrigeration cycle whose template can be found in the A C library Chapter 4 Optimization and Control As mentioned in Chapter 1 4 a variety of control actions should been carried out in the A C unit to keep the passenger cabin in a comfortable condition and to maintain system efficiency This chapter deals with the control of COP and cooling power in the cooler unit The role of the cooler unit in the A C system is to provide maximum cooling power in order to cool down the air and dehumidify it before re heating and ventilation Its second target is to maximize COP in order to reduce the fuel consumption To increase cooling power at very high ambient temperature usually a lower COP is accepted However most of the operating times the optimization of COP is the dominant control target The control actions of the cooler unit are performed under the general logic of the A C control unit 4 1 Optimum High Pressure Control To achieve maximum COP in R744 systems a simple SISO control strategy has been proposed by Yang et al 2005 Its performance for the R744 system with the specifications stated in Chapter 1 is studied in this section They consider the high pressure as the main variable that affects the COP and cooling power Since t
6. the traditional object oriented programming language like C or Java Modelica concept of object orientation is different from them in some aspects Modelica uses declarative mathematical descriptions for its models A declarative representation of system behavior does not determine how something is calculated instead it de fines what it is Tummescheit 2001 so when you look at the Modelica code of a 18 19 system you will find similarities with the explicit mathematical description of that system Therefore the task of procedural description is derived by the compiler not by the programmer Classes Like any other object oriented language Modelica provides the notions of classes and instances Modelica classes can be classified in six major categories e Type All data objects in Modelica are instantiated either from basic data types Real String Boolean Integer or from enumeration types It is possible to define different attributes of a variable other than its value For example the attributes of Real type are predefined as value quantity unit display unit min max etc Therefore physical types e g pressure temperature mass can be derived from basic types but with their own special features All the well known SIunits are already instantiated from the basic types and exist in the Modelica standard library to ease declaring physical quantities and reduce the risk of errors in programming e Model It is a general term
7. R134 systems Vehicle Thermal Managment Systems 7 Conference and Exhibition SAE International 2023 Pfafferott and Schmitz 2002 Pfafferott T and Schmitz G 2002 Modeling and Simulation of Refrigeration Systems with the natural refrigerant CO2 Rasmussen 2002 Rasmussen B 2002 Control oriented modeling of transcritcal vapor compression systems Master s thesis Illinois University Urbana Shah et al 2003 Shah R Alleyne A and Ramussen B 2003 Application of multiple adaptive control to automotive air conditioning systems ASME Interna tional Tiller 2001 Tiller M 2001 Introduction to Physical Modeling with Modelica Kluwer Academic Publishers Tummescheit 2001 Tummescheit H 2001 Design and Implementation of Object Oriented Model Libraries using Modelica PhD thesis Lund Institute of Technology Tummescheit et al 2005 Tummescheit H Ebron J and Pr l K 2005 Air conditionig a modelica library for dynamic simulation of ac systems Proceedings of 4th International Modelica Conference Watanabe 2002 Watanabe S 2002 Automotive Air Conditioning DENSO corp Wiessler 2006 Wiessler P 2006 Air conditionig and global warming Automotive Engineering International SAE International 114 52 54 Yang et al 2005 Yang W Fartaj A and Ting D 2005 Co2 automative A C system optimum high pressure control Vehicle Thermal Managment Systems 7 Conferen
8. air sid v Qer e C AIT side on the refrigerant side 9 3 T VP PL Low pressure High pressure Figure 2 10 Balance change by the evaporator refrigerant flow rate 2 4 Control System The air conditioning control system consists of four basic parts e Sensors that detect the ambient air temperature the cabin condition and the A C system operating condition e A control panel that indicates the temperature the operating condition etc e The air conditioning Electronic Control Unit ECU which is responsible to run the logic of A C system control Air conditioning ECU calculates signals in order to control the outlet air temperature and the outlet airflow volume based on the various sensor signals and A C panel signals etc e The air condition unit that operates according to the signal from the ECU Control goals of an A C system cab be classified as below e Temperature Control 15 To provide the desired temperature the angle of the temperature blend door in Figure 2 1 is controlled by the ECU based on the outside environ ment and car cabin condition By this control the outlet air temperature varies between the outlet temperature just after the evaporator and the heater core temperature Airflow volume control With the airflow volume control ECU changes the speed of blower motor of A C unit See Figure 2 1 to increase the airflow volume when high cooling heating capacity is required and decrease it when the
9. amplitude of the oscillations differ in these two different desired evaporator air outlet temperatures 42 Other output parameters which are correlated with the high pressure will also show this limit cycle The effect on the outlet air temperature is negligible less than 1 C in this case and it is seen in Figure 4 20 that the low pressure controller can remove the fluctuations Therefore passengers do not sense the oscillations of the temperature But the effect on the cooling capacity and COP is more considerable Desired Low Pressure 40 bar Desired Low Pressure 45 bar N 00 N N oo o Y E x o i o0 2 2 284 alas eo ees a ep a x dl A MEE e OLEI 278 2 580 3 3 5 5 2 BES DR i 2781 5 e 5 amp 276 8 CEA Semen Corer A server dal E H i a D 274 o epe wee 1 E S E E i m 272 300 305 310 315 320 325 300 305 310 315 320 325 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 20 Temperature and limit cycle Desired Low Pressure 40 bar Desired Low Pressure 45 bar 3 5 5 3 4 A 4 2 5 3 a a o 2 0 2 1 Oo Oo 15r tt 1 0 4 0 5 i i i i 1 i i i i 300 305 310 315 320 325 300 305 310 315 320 325 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 21 COP and limit cycle In the temperature interval where this phenomenon happens the highest deviation of the COP is about 50 less tha
10. as a natural alternative to current R134a based systems Main benefits of R744 as a refrigerant are e Energy efficient e Non toxic Non flammable No ozone depletion potential ODP 0 Low global warming potential GWP 1 Besides the environmental benefits mentioned above at some climate conditions using R744 as a refrigerant for Air Conditioning A C systems decreases the fuel consumption through less fuel needed to run the A C system Using R744 creates a need for major changes in the current A C system components DaimlerChrysler and its A C system supplier Behr GmbH have designed and de veloped the prototypes to be used in the future cars These prototypes have been modeled and validated by means of the Dymola Air Conditioning library This li brary is a Modelica package for dynamic simulation of A C systems which has been developed at Modelon AB Several issues with the R744 components design and system control remain unsolved or undecided Wiessler 2006 In this thesis some of these control issues will be studied From the control viewpoint two features of the A C systems are important to be considered 1 Comfort maintenance To provide passengers with a nice temperature an ac ceptable amount of air flow and low humidity in the cabin 2 Energy efficiency To reduce the fuel consumptions of the A C unit which is one of the major fuel consumers in the car The A C system consists of several different parts w
11. for a complete Modelica object which usually is a structured description of the physical systems A component is an instance of a model e Connector Interaction between objects components of well structured classes models is usually done through the connectors For example each electrical component should contain connectors named Pin which defines voltage and current in connection with other electrical components based on Kirchhoff s laws e Block Blocks have the well known internal semantics with known inputs at tached to input connector from which the outputs are computed e Function Similar to the block definition it defines the relation between its inputs and outputs but these are just variables not connectors Functions can be called inside an equation e Package A package refers to a collection of Modelica models which are meant to be used together Tiller 2001 20 Inheritance Similar to other object oriented languages the original class or the base class is extended to create a more specialized version of that class known as child class This derived class inherits the behavior and properties such as variable declara tions equations and other contents of the original classes Modelica also supports the concepts of abstract classes under the name of partial models These are not complete models and can not be instantiated only used in inheritance Equations Modelica is primarily an equation based language in co
12. subcritical heat rejection is just possible for refrigerants with the critical temperature much higher than the hot space temperature But the critical tempera ture of R744 is 31 C which for many the cases is less than the ambient temperature In such caess the heat rejection process does not happen in the subcritical region anymore and the refringent will enter the supercritical region to reach higher temper atures So the subcritical cooling system needs some modifications to be able to work for R744 Figure 2 5 illustrates this modified system In this system the condenser is replaced by a gas cooler Another component named internal heat exchanger is also added to the structure Refrigeration cycle stars at the point 1 when the vapor refrigerant enters the compressor then the compressor raises the refrigerant pressure temperature and enthalpy The refrigerant with high pressure leaves the compressor for the gas cooler Unlike the condensing process in subcritical systems the refriger ant remains in the gas phase and the heat rejection in the supercritical region does not occur isothermally Gas cooler Pressure gt IS Aw N Critical point Internal Heat Exchanger Liquid Evaporator Enthalpy Receiver Figure 2 5 Subcrtical refrigeration cycle To increase the system efficiency and reject more heat an additional heat exchanger is needed The inner part of this internal heat exchanger removes e
13. temperature becomes closer to the desired temperature and a large capacity is not required anymore Airflow distribution mode control Besides the temperature degree passengers comfort is also dependent on the direction of the airflow In some cars A C unit there is a air distribution door which directs the airflow either from face or head ventilation duct Figure 2 1 Air intake mode control This action is done to switch between fresh or recirculation air mode when it is required See Section 1 1 Humidity control The evaporator is responsible to absorb the humidity of air cabin by pro viding enough cooling power Capacity Control In section 2 3 1 we have seen that the cooling power and efficacy of the A C unit are variable quantities relating to operating conditions To have control on these quantities two components of the refrigeration cycle are made as controllable components expansion valve and compressor There fore control system can change the refrigerant flow rate by means of these components to achieve desired capacity 16 2 4 1 Controllable Components Expansion valve The expansion valve could be a linear proportional solenoid valve Usually the Kv of this valve which defines its flow area can be changed by pulse width modulation PWM techniques Compressor The compressor is a belt driven pump that most often is fastened to the car engine via a clutch There are primarily four types of compres
14. the pressure difference and 39 dd Optimized cycle m Optimized cycle cycle with two stage valve cycle with two stage valve 4 3200 3100 o 3 3000 oa o 5 2900 o Oo 2800 5 15 i 2700 295 300 305 310 315 320 295 300 305 310 315 320 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 14 Comparison of the two valves case 3 11 5 0 02 r a Optimized cycle 11 0 018 eyel with we Be valve 10 5 o 5 40 S 0 016 a o S 95r s l gt 004 i coe o 5 inis 9r E E 8 5 8 i i i 0 01 i i i A 295 300 305 310 315 320 295 300 305 310 315 320 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 15 Comparison of the two valves case 3 mass flow rate To observe the role of flow rate and pressure change in the limit cycle phenomenon directly all the boundary conditions are kept constant and the relative volume of the compressor is changed manually to provide the appropriate pressure difference and flow rate Figure 4 16 illustrates above explanations Figure 4 17 shows the portrait plot of the valve Kv against the pressure difference 40 1 05 to a High Pressure Pa eo to 0 85 i i i 1 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 1 Comprssor Relative Volume Figure 4 16 Limit cycle 0 03 T T T T T T T Valve Kv 0 015 Excess ett
15. 0 Time s Time s Time s Figure 4 25 Cycle bi stability Chapter 5 Four Zone Air Conditioning Modern luxury cars allow passengers to control a different climate in up to four climate zones This requires the presence of two evaporators to generate the cooling capacity for front and rear passengers The Electronic Control Unit ECU controls the position of the different temperature blend doors to provide the passengers with their desired temperature in different zones In the cooler unit the high pressure Condenser Gas cooler Expansion valve 4 Compressor Des Evaporator Sea DD S A Ad e Figure 5 1 Four Zone Air conditioning of Mercedes Benz S class Behr GmbH amp Co 49 50 refrigerant splits and flows from two different expansion devices to the front and rear evaporator Then the cooling capacity is divided between these evaporators But the amount of the division depends on the operating conditions and structure of the valves If a variable displacement compressor is used to control the front evaporator outlet air temperature and a two stage orifice valve to improve the COP then a fixed orifice can be used to pass the refrigerant to the rear evaporator In this case there is no direct control on the outlet air temperature of the rear evaporator To have full control authority on both evaporator temperature a model based controller is needed to be designed to control the compressor relati
16. Open Issues in Control of Automotive R744 Air Conditioning Systems Sanaz Karim 2007 Master Thesis Electrical Engineering Nr E3492E DEGREE PROJECT HOGSKOLAN Dalama Electrical Engineering Programme Reg number Extent Master of Science in Electrical Engineering E3492E 30 ECTS Name of student Year Month Day Sanaz Karim 2007 03 29 Supervisor Examiner Dr P r Samuelsson Dr Bj rn Sohlberg Company Department Supervisor at the Company Department Modelon AB Dr Hubertus Tummescheit Title Open Issues in Control of Automotive R744 A C Systems Keywords R744 Air Conditioning COP Cooling power optimization two stage orifice valve Two evaporator Abstract In this thesis one of the current control algorithms for the R744 cycle which tries to optimize the performance of the system by two SISO control loops is compared to a cost effective system with just one actuator The operation of a key component of this system a two stage orifice expansion valve is examined in a range of typical climate conditions One alternative control loop for this system which has been proposed by Behr group is also scrutinized The simulation results affirm the preference of using two control loops instead of one loop but refute advantages of the Behr alternate control approach against one loop control As far as the economic considerations of the A C unit are concerned using a two stage orifice expansion valve is desired by
17. ach speed Table 3 3 gives the exact value of this optimum pressure for our system In Yang et al 2005 it is also shown that other boundary conditions evaporator temperature air humidity and flow rate have negligible effect on the optimum high pressure As it stated in the first chapter a variable swash plate controller is going to be used as low pressure evaporator air outlet temperature controller and since any change in a angle of the swash plate will affect the pressure ratio as well as the compressor power it is expected that it changes the optimum high pressure as well Figure 4 4 shows that although COP and cooling power reacts to it considerably the value of 32 Table 4 3 Optimum high pressure and speed Compressor speed Hz Optimum high pressure at Tgc 35 C 16 97 26 108 36 111 the optimum high pressure does not vary too much 3 an T 5000 4000 F 25r a a 2 3000 a o 2 a o o 2000 o o 1 5 Q 1000 p 1 i 0 i i 0 6 0 8 1 1 2 1 4 0 6 0 8 1 1 2 1 4 High pressure Pa x10 High pressure Pa x10 Figure 4 4 Comparison of COP and cooling power with the change of relative volume Table 4 4 Optimum high pressure and relative volume Relative volume Optimum high pressure at 16 Hz bar 1 97 0 5 93 0 3 92 Therefore a high pressure regulator which controls the refrigerant flow based on
18. age valve 22 CE r 4600 2 s aset aspe ee ER 4400 o 4200 a 18 8 2 4000 r 1 6 3 8 3800 1 4 3600 i i i i 3400 i i i 295 300 305 310 315 320 295 300 305 310 315 320 Gas cooler air in temperature K Gas cooler air in temperature Figure 4 10 Comparison of the two valves case 1 x 10 11 5 T 0 024 Eat po 07 o peso Optimized cycle 11 selecta on a esee ea se 0 022 cycle with two stage valve E 10 5 K E S 0h e 0 02 o E os 2 a 95r R E 5 gt 0 018 o x 9 s 0 016 1 8 5 8 E i i i i 0 014 i i i i 295 300 305 310 315 320 295 300 305 310 315 320 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 11 Comparison of the two valves case 1 ambient temperature it cannot maximize the cooling power opposed to the optimized cycle For this cooling load the Kv shows a smaller deviation from the optimized one Evaporator temperature controlled low cooling load When the low pressure is controlled via the relative displacement of the com pressor the COP is improved for both cycles Since the pressure difference is low at the lower ambient temperatures the refrigerant passes through the fixed orifice of the two stage orifice valve and provides the high pressure that is needed for better COP The slope of the high pressure and valve Kv are closer to the optimized cycle
19. calculated by algebraic equations Efficiencies can be provided by non linear approximation functions or tabulated data Approximation functions are verified with data sets for many compressor types Displacement volume can be controlled by an external signal Besides these models of expansion devices pipes and volumes homogenous and in homogeneous air sources are also available The other important factor in the complete transient simulation of a vapor cycle is the refrigerant mass During different boundary conditions e g changing air temperature air massflow through heat exchanger the refrigerant mass is moving to different parts of the system and must be observed and then the optimal charge of the system and the changing process behavior during any variation of the compressor speed air massflow and temperature can be found out Limperich et al 2005 27 Cond_mair Cond_tair Cond_phi RV duration 2 _ ation 2 gascooler GD je BAAAAAA keintt phiCond AAAAAAA airln cond ihxout 1431936 FAAAAAAR 2971440 MIYZ YYN Legend z 3 h kJ kg p bar m a m g s T C A o s T C 5 rps Speed a JE Kv value 9 d 1 5 3 i w duration 2 i Y init E Relvol g F UE eva t A k 1 0 TI Time SEISSISIS receiver Evap_tair CIKIKIKIKIKINI R RISKISKISIC N 180 0 SSS i IP Do Do De Ber Cooling Power W LLLA A h Collected Water g 21 4
20. ce and Exhibition SAE International 2022
21. compressor provides work needed for the heat rejection In Figure 2 2 the location of these parts in Mercedes Benz S class is shown This A C unit has been designed by Behr GmbH and BHTC Condenser Gas cooler Expansion valve Compressor Figure 2 2 Air conditiong of Mercedes Benz S class Behr GmbH amp Co KG 2 2 Refrigeration Cycle In this section the cooling process will be explained by the help of pressure enthalpy PH diagram This diagram shows the phase changes of a fluid through different enthalpies and pressures Enthalpy is a measure of the useable energy content of a fluid Figure 2 3 shows the PH diagram of R134a which is one of the most in use refrigerant in vapor compression cycles In this figure point a which is located in the subcooled or liquid region shows the phase of the refrigerant when its temper ature is below its boiling point If heat is added to this refrigerant at the constant pressure its enthalpy will increase and make saturated liquid at point b with more heat it enters into the two phase or mixed vapor and liquid region Afterwards by adding more heat the vapor portion of refrigerant will increase and at point c it becomes saturated vapor The vapor fraction of the refrigerant is called quality with a value between 0 and 1 so at point b the quality of the refrigerant is 0 and at point c it equals to 1 The heat that is absorbed in the two phase region is called latent
22. compressors have been developed in to types swash and wobble plate In these compressors the required amount of refrigerant gas is sucked in and compressed by changing the pressure balance within the compressor The angle of the swash or wobble plate determines the length of the piston stroke In a variable displacement compressor that angle can be changed which changes the length of the pistons stroke and therefore the amount of the refrigerant displaced in each stroke Gordon 2005 Chapter 3 System Modeling and Simulation Tools The model of the air conditioning system which is used in our simulations is explained in this chapter This model is built based on the prototype components of Daimler Chrysler R744 A C system by means of the Air Conditioning Library This library has been designed and developed at Modelon AB as a commercial Modelica library for the steady state and transient simulation of automotive air conditioning systems 3 1 Modelica Language Modelica is a declarative object oriented modeling language which allows conve nient component based modeling of complex systems which could contain mechan ical electrical electronic hydraulic thermal control and electric components as well as multi domain structures It supports continuous discrete event and hybrid modeling approaches The key features of this language are described briefly here Object orientated Modeling Although it has many features in common with
23. eal capacitor extends TwoPin extends TwoPin parameter Resistance R parameter Capacitance C equation equation R p i v C der v p i end Resistor end Capacitor 23 3 2 Dymola Environment Dymola Dynamic Modeling Laboratory is a front end for Modelica which contains a symbolic translator to pre process the Modelica equations and then generate C code for simulation Symbolic manipulation can solve or simplify differential algebraic equations of any order at translation time in order to make less numerical calculations when running the simulation Fritzson 2003 Dymola has a graphic editor for composing Modelica models It can also import other data and graphics files The generated C code can be exported to Simulink and hardware in the loop platforms Scripts can be used to manage experiments and to perform calculations Figure 3 3 from Dymola 6 0 User Manual shows Dymola links with other environments A H g 2 E and Analysis b Visualization Figure 3 3 Architecture of Dymola Dymola has two kinds of windows Main window and Library window The Main window operates in one of two modes Modeling or Simulation The Modeling mode of the Main window is used to compose models and model components The Simu lation mode is used to make experiment on the model plot results and animate the behavior The Simulation mode also have a scripting subwindow for automation of experimentation and performing calcu
24. early with the pressure difference orifice Valve Ky Pressure Difference y Figure 4 9 Two stage orifice valve 4 3 1 Performance of the Valves To compare the operation of the controllable valve in an optimized cycle and a two stage valve all boundary conditions and the compressor speed are kept constant and the simulation runs three cases for both of them 1 Low cooling load and no control on evaporator outlet air temperature Fixed relative volume of the compressor The system with two stage valve has lower COP and higher cooling power than the optimized cycle even for the lower ambient temperature At higher tem perature losses decrease Figure 4 10 The high pressure with two stage orifice valve is kept fixed around 110 bars while the variable Kv valve allows the pres sure to change in a wider range The reasons is behind the internal mechanism of the two stage orifice valve which does not provide a very close Kv to the controlled Kv for most of this range Figure 4 11 2 High cooling load and no control on evaporator outlet air temperature Fixed relative volume of the compressor In comparison with the previous case at the higher loads the cycle with two stage orifice valve has a COP near to the optimum value but in the higher 37 pes Optimized cycle Um Optimized cycle cycle with two stage valve cycle with two st
25. ed modeling of the same circuit Here the phys ical topology is lost Figure 3 2 Equivalent Simiulink model 22 In the equation section the command connect Pin1 Pin2 tells the compiler to com pute two equations Pinl v Pin2 v and Pinl i Pin2 i 0 The Pin connector definition is shown below Similar laws apply to flow rates in a piping network and to forces and torques in a mechanical system The sum to zero equations are generated when the prefix flow is used in the connector declarations partial model TwoPin Pin p n connector Pin Voltage v Voltage v equation flow Current i v p v n v p i n i 0 end Pin end TwoPin To model a resistor or any other simple electrical components it is useful to define a shell class TwoPin This class has two pins p and n and a quantity v that defines the voltage drop across the component See the code below The equations define common relations between quantities of a simple electrical component In order to be useful another complementary equation must be added The keyword partial indicates that this model class is incomplete To define a model for a resistor start from TwoPin and add a parameter for the resistance and Ohms law to define the behavior For the capacitor it just needs little change in the equation section Derivation is possible in Modelica and the expression der v means the time derivative of v model Resistor Ideal resistor model Capacitor Id
26. erant flow rate increases and the condenser can reject more heat Thus based on the balance principle for the condenser the high pressure corresponding to the refrigerant temperature increases Qer on the refrigerant side Balance point Qca Cooling power Qea on the air side P YP gt Low pressure Figure 2 8 Balance change by evaporator airflow rate When the airflow rate around the condenser decreases the heat rejection capacity on the air side decreases To compensate for this the high pressure increases to balance with the refrigeration capacity on the refrigerant side Figure 2 9 4 Qca on the air side Balance point Na Qca on the air side Heat rejection capacity Ph Pu High pressure Figure 2 9 Balance change by condenser airflow rate 14 Another parameter which influences the balance of the cycle is the compressor rotation speed When the speed increases the refrigerant flow rate also increases Since this increases the refrigeration capacity on the refrigerant side the balance point location is changed because the low pressure decreases to achieve the heat absorption capacity on the air side and the high pressure increases to achieve the heat rejection capacity of the condenser Watanabe 2002 Q er Qer the refrigerant side 4 Qca D on the air side 5 3 5 E 2 Y Yv Balance point o 2 5 Q cr on 5 amp Qea on 2 E the
27. erature is not a direct function of the high pressure the idea of measuring high pressure instead of low pressure causes some problems in the controller design that should be solved A feedforward compensator has been pro posed in Lochmahr et al 2005 to solve these problems In the following argument 46 Air mass flow Engine speed inlet air temperature Compressor MN Gas Cooler Temperature set point Internal Heat Exchanger Tracking signal Two stage orifice c valve Figure 4 24 Alternative control system for R744 cycle with two stage valve we will show that the idea of adding the feedforward compensator to the system be sides the temperature outer control loop will handle a part of these problems to some degree but it needs very accurate design and can not provide good robustness To investigate the relation between the high pressure and evaporator outlet air tem perature the system input and disturbances can be divided in two groups 1 Engine speed and any disturbances on the side of compressor and expansion valve 2 Boundary conditions and any disturbances on the air side of the gas cooler and evaporator If any items of the first group changes the system input power will change and the pressure lines on the PH diagram move in the opposite direction of each other In this case the high pressure is just a function of the input power and not a function of the low pre
28. flow properly in different parts of the car cabin by means of four major parts an air intake mode selector a blower cooler and heater The location of these parts differs in vehicles but they generally are installed in the instrument panel of the vehicle with the configuration of Figure 2 1 Air which is introduced to the blower might be fresh or reciruculation air Depending on the driving condition and the cooling load the air intake mode selector switches between reciruculation and fresh air In the recirculation mode inside air of the cabin is re circulated and in the fresh air mode fresh air from the outside is used for air conditioning In a place where it is full of exhaust gas or dirty ambient air or when high cooling power is needed the air mode selector is switched to re circulation air but normally fresh air is preferred The cooler is located after the blower In the cooler unit air is dehumidified and cooled and then by the means of a heater it is re heated to reach the desired temperature The air distribution unit controls the volume of air to be re heated in the heater part The cooler unit stores the evaporator and expansion valve Heat which is absorbed Recirculation air Fresh air Air intake mode selector Blower Evaporator Temperature blend door Heater Figure 2 1 Automative A C archetype by this unit is rejected to outside environment by a condenser which is located in the front of the car A
29. g system The refrigerant commonly used in this system is R134a At point 1 the refrigerant is sucked into a compressor at a low pressure and compressed adiabtically no heat is removed at a higher pressure and temperature higher than the critical temperature of fluid it enters a condenser at point 2 Since it has higher temperature than the hot space temperature the refrigerant is transferring its heat to the air and condensing It leaves the condenser as a subcooled fluid at point 3 Through an expansion device it expands and enters to the low pressure part of the system This expansion process moves the refrigerant into the two phase region at a low pressure and temperature Then in the evaporator the refrigerant with a low pressure and a temperature lower than the ambient temperature absorbs heat from the cold space to vaporize and then reach the point 1 as superheated vapor To be sure that just pure vapor enters the compressor and no liquid a component called receiver is fixed before the compressor which separates the liquid fraction of the refrigerant Another option is to place a receiver accumulator after the condenser to subcool the refrigerant thus the refrigerant entering the expansion device is always a saturated liquid Critical Condenser Pressure Expansion Compressor y valve Evaporator Receiver Enthalpy Figure 2 4 Subcrtical refrigeration cycle 2 2 2 Transcritical Cycle Apparently
30. gated either for this system Therefore it is not proved that using a two stage orifice valve is the best solution for this type of systems and future investigation should be considered Bibliography Astr m and H gglund 2006 str m K and H gglund T 2006 Advanced PID Control ISA The Instrumentation Systems and Automation Society Fritzson 2003 Fritzson P 2003 Principles of Object Oriented Modeling and Sim ulation with Modelica 2 1 A John Wiley amp Sons Inc Publication Gordon 2005 Gordon J 2005 Variable displacement A C compressor Motor Age 4 54 55 He and Liu 1998 He X D and Liu S 1998 Multivariable control of vapor com pression systems HVAC R Research Lemke et al 2005 Lemke N Tegethoff W Kohler J and Horstmann P 2005 Expansion devices for R744 MAC units Vehicle Thermal Managment Systems 7 Conference and Exhibition SAE International 2041 Limperich et al 2005 Limperich D Braun M Pr lf K and Schmitz G 2005 System simulation of automotive refrigeration cycles Proceedings of 4th Interna tional Modelica Conference Lochmahr et al 2005 Lochmahr K Baruschke W and Britsch Laudwein A 2005 Control system for R744 refrigerant circuits ATZ worldwide 9 796 799 97 98 McEnaney and Hrnjak 2005 McEnaney R and Hrnjak P 2005 Clutch cycling mode of compressor capacity control of transcritical R744 systems compared to
31. ged into the model and connected in an recognizable manner without considering the calculation order It also simplifies model maintenance as there is only one definition of each component model that needs to be maintained Tiller 2001 Algorithms In some cases where nondeclarative constructs are needed algorith mic statement can be written within the algorithm section of Modelica In the al gorithm section assignments if then else statements for loops and while loops are available to use 21 Functions are reusable algorithms with an fixed relation between its inputs and out puts but when a function is called inside an equation it can be used to calculate an input from a given output Example In Figure 3 1 Modelica code of a simple electrical circuit is shown In this model the components R1 and R2 have been instantiated from the Resistor class and so on about the other components Then in equation section the order of connections between different components has been defined model Circuit Resistor Ri R 10 R2 R 100 Capacitor C C 0 01 Inductor L L 0 1 CH 8 VsourceAC AC Ground G equation J connect AC p R1 p Capacitor circuit connect Ri n C p connect C n AC n connect Ri p R2 p Inductor circuit connect R2 n L p connect L n C n connect AC n G p end Circuit V 1 Figure 3 1 Modelica model of an electrical circuit Figure 3 2 illustrates the block orient
32. he heat rejection process of the R744 refrigeration cycle takes place in the supercritical region where the pressure is independent of the temperature the system efficiency is a nonlinear function of the working pressure and ambient temperature In the supercritical region the CO 28 29 working pressure typically ranges from 7500 to 13000 kPa By changing the valve opening to achieve this desired high pressure the simulation is conducted over this pressure range in order to study the effect of high pressure on the COP and cooling power in some typical European climate conditions First to study the effect of high pressure on COP of different gas cooler air inlet tem perature six other boundary condition are kept constant with the values of Table 4 1 Table 4 1 Boundary Conditions Parameter Value Gas cooler relative humidity 6096 Gas cooler air flow rate 0 6 kg s Evaporator air inlet temperature 25 C Evaporator relative humidity 40 Evaporator air flow rate 0 1 kg s Compressor speed 16Hz Figure 4 1 shows the variations of the COP and cooling power in the R744 system with respect to the high pressure variation at four different gas cooler air inlet tem peratures 30 C 35 C 40 C and 48 C Figure 4 1 indicates that e Higher ambient gas cooler air inlet temperature gives lower COP and cool ing power In lower ambient temperatures the sensitivity of the COP to the optimum high pre
33. heat of evaporation because this heat does not increase the temperature and just increase internal energy or enthalpy Therefore the evaporation of the refrigerant occurs isothermally in the two phase region With further heat increase the refriger ant state moves into the superheated vapor region and temperature will increase in this region On the PH diagram it is seen that the saturated liquid and vapor curves unite at a point are called critical point At this point the refrigerant temperature and pressure called critical temperature and critical pressure respectively The Su percritical region is located above the critical pressure where the refrigerant state does not undergo distinct phase transitions by heat addition or reduction Constant Temperature Lines AE Temperature increases to the right Supercritical Region Critical Point X Mixed Liquid Superheated 1 Vapor Region Vapor Region Pressure i I __ Saturated Liquid Line Saturated m Vapor Line Enthalpy Figure 2 3 Pressure Enthalpy Diagram To absorb heat from the cold space car cabin and discharge it into the hot space outside environment a specific cycle of phase transitions needs to take place which will be explained in the next section 2 2 1 Subcritical Cycle Figure 2 4 illustrates a subcritical refrigeration cycle with its related coolin
34. hich cooperate with each other to satisfy the above goals but the task of the heat absorbtion from the car cabin is mostly carried out by one evaporator for ordinary vehicles or two evaporators for luxury or very big vehicles The absorbed heat of the evaporator is rejected to the outside environment from a condenser by aid of a compressor Since the A C system is a multi input multi output process for a multiple perfor mance control the model based multivariable control strategies needs to be applied For R134a systems such strategies have already been developed by Shah et al 2003 and He and Liu 1998 However such sophisticated controllers are not popular cost effective solutions in the opinion of the automotive industry Therefore decoupled Single Input Single Output SISO feedback loop techniques must be developed by considering the strong cross coupling among the various actuating inputs and outputs Shah et al 2003 Variable displacement compressors electric expansion valves and variable speed fans for air flow over the heat exchangers are the available controllable components which can be used for the purpose of multiple SISO control Apparently using several SISO control loops results in increase of control authority but again cost reduction consid erations require as few as possible controllers sensors and controllable components One of the current multiple SISO control algorithms for R744 cycle tries to optimize the perfor
35. i enterado bs dee te cae eta dote 4 4 5 5 5 5 6 6 5 7 7 5 8 8 5 Pressure Difference Pa x 10 Figure 4 17 Limit cycle Because of this phenomenon the COP is subject to fluctuations its highest deviation is about 20 less than its average value in this range Therefore the limit cycle occurrence will reduce the expected average COP 41 3 5 L j 1 i 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 1 Comprssor Relative Volume Figure 4 18 Limit cycle However this limit cycle is not a consistent behavior and its characteristic differ in different circumstances The following observations demonstrate this statement Assuming a low pressure controlled cycle the ambient temperature varies in the range of 30 C to 45 C and other operating conditions are kept constant Figure 4 19 shows the portrait plot of two different cases when the limit cycle takes place One of them happens when the desired low pressure is 40 bar and the other one at 45 bar Desired Low Pressure 45 bar Desired Low Pressure 40 bar 0 03 T 0 03 r 0 025 0 025 S 0 02 S 002p f o o 2 S 0 015 S 0 015 0 01 0 01 0 005 0 005 1 06 1 08 1 1 1 12 1 14 1 06 1 08 1 1 1 12 1 14 High Pressure Pa x10 High Pressure Pa x 107 Figure 4 19 Portrait plot of the valves Kv against the pressure difference As depicted in Figure 4 19 the number of limit cycle orbits and the
36. lations MA Anmann Commands DIEZ Sus e ETE EIA eX Bees ees e a O Medals Fabero ure Unnered Sedis KR Uses Guide 5 patentes 2 este 8 frena ut s 2exezaed P LI 2d dinem dS Cabn Mesa 3 Congressos 3 Conrolerindi a ed i js Drag and drop Check Diagram classes S Test1 Modelon Blocks AutoPID Test Test1 Na cd C Kome P000 Internal Modelon plot co2Cycle DC szwummary P evaporator 3 simulateModel Modelon Blocks RutoPID Test Testi lt i Figure 3 4 Modeling and Simulation mode of Dymola 24 25 3 3 Modelon Air Conditioning Library The aim of the modeling is to create a library with physical based models of the A C system components Such a library with models of these components and of additional components for testing like air sinks and sources can be used for investigations of both single components and complete refrigeration cycles Furthermore it is of great interest to make dynamic simulation as well as steady state simulation of refrigeration cycle and single components especially heat exchanger Pfafferott and Schmitz 2002 The Air Conditioning Library provides component models for automotive A C sys tems This library has been derived from the Modelica library ThermoFluid and the ACLib library Traditionally A C system level models are only used as steady state models with the exception of very sim
37. mance of the system and satisfy comfort requirements by the help of two controllable devices the variable displacement compressor and the electronic expan sion valve It would be much more agreeable if one of these controllable device can be replaced by a simpler and cheaper one Therefore a mechanical expansion device the two stage orifice expansion valve has been built to replace with the electronic expansion valve However the operation of this device in various climate conditions has to be studied more carefully for concerns of unit efficiency and comfort One of the major A C component suppliers Behr GmbH amp Co has designed its second generation of the R744 A C system by accepting this valve and claimed at minimum additional cost their new system achieves the same level of steady state cooling performance as present R134a systems Lochmahr et al 2005 This claim is also going to be examined precisely The problem of the decoupling is more complicated for two evaporator systems be cause of their numerous coupled inputs It should be verified whether the same sim plified control approach of one evaporator systems is applicable for two evaporator systems According to this concise introduction the main goals of this thesis can be classified as follows Goals e Verification of the current multiple SISO loop control algorithm for different cooling loads and in different climate conditions e Comparison between the operati
38. me Ca Specific heat of air Gea Air mass flow rate h4 Enthalpy of the refrigerant at the evaporator inlet h Enthalpy of the refrigerant at the evaporator outlet Ter Temperature of inlet air Tea Temperature of the refrigerant in the evaporator From the 2 3 6 it is seen that the lower the refrigerant temperature Ter the higher capacity on the air side Qea can be obtained In other words the refrigerant tem perature changes the intercept of the Qea line in Figure 2 6 On the other hand the lower refrigerant temperature Ter results in the lower capacity on the refrigerant side Qe and changes the slope of the Qe in Figure 2 7 because the lower refriger ant temperature means the lower low pressure as it is seen from the PH diagram in Figure 2 3 and due to the fact that the specific volume of gas is less at the lower pressures so according to equation 2 3 7 this decreases the capacity on the refrigerant side Thus the refrigerant flows at a low pressure which satisfies Qer Qea We call it balance point for the low pressure Similar to the evaporator and low pressure the high pressure balance point is defined by the air and refrigerant side properties of the condenser as indicated in Figure 2 6a The only difference is that the high pressure variation does not change the refrigerant flow rate considerably a 4 Qer 5 Balance point on the refrigerant side Z o y a 20 E i E i Qea O i on the air side l l l P gt
39. mperature dynamic characteristics are close enough to the low pressure dynamics Figure 4 23 shows this fact by illustration the relation between the low pressure and evaporator air outlet temperature in an uncontrolled system when the compressor speed is changing very fast Therefore for our system using cascade controller gives no significant advantage and one of these control loops can be eliminated from the cascade loop 45 2 8 Low Pressure Pa N i Evap air out temerature K 6 i 500 600 580 600 Time Time Low Pressure Pa N co T 267 268 269 270 271 272 273 Evaporator air outlet temperature K Figure 4 23 Evaporator temperature and low pressure In the second proposed controller instead of the controlled valve the two stage orifice valve is used therefore there is no direct control of the COP They also suggest to enter a high pressure controller in place of the low pressure one in the inner loop because the refrigerant high pressure sensor required for controlling the high pressure is already present for monitoring and protection functions in today s R134a circuits so no additional sensors are needed and this plan remove the cost of low pressure sensor The configuration of the system with this controller is shown in Figure 4 24 The first rule of the cascade control is no more true here Due to the fact that the evaporator outlet air temp
40. n d 1 1 i i i i 2500 i i i i 295 300 305 310 315 320 295 300 305 310 315 320 Gas cooler air in temperature K Gas cooler air in temperature K Figure 4 7 COP and cooling power in the low pressure controlled cycle 1T Low pressure controlled Fixed relative volume 278 T T T T T T T T T nani n nf e i P E e p nd npn 1 272 eed ee E Pc e E 268 Evaporator air outlet temp K 66 i L j 1 1 fi 298 300 302 304 306 308 310 312 314 316 318 Gas cooler air in temperature K Figure 4 8 Evaporator air outlet temperature 4 3 Two stage Orifice Valve To remove the costs of the high pressure controller and gas cooler temperature mea surement device a two stage orifice expansion valve has been produced by Egelhof GmbH and modeled by DaimlerChrysler This model has been modified by Modelon AB and is used in our simulations This valve has an internal mechanism to change the Kv value of the valve based on the pressure difference between the low and high pressure side Lemke et al 2005 It consists of a standard orifice and a bypass as it is shown in the Figure 4 9 at pressure differences below 23 bar the refrigerant flows only through the orifice The bypass starts to open at 73 bar with a jump and for 36 higher pressures the Kv value rises linearly If this increase results in pressure dif ference higher than 73 bar the bypass will start to open again and the Kv increases lin
41. n expected average value 43 4 4 Behr Alternative Controller Behr GmbH amp Co KG Stuttgart is a systems partner for the international automo tive industry who has developed different components as well as new control methods for R744 systems Lochmahr Barushke and Britsch Laudwein in Behr GmbH have introduced two types of improved controllers for the R744 system Lochmahr et al 2005 The first con troller has been designed for the optimized system with a controllable valve and the second one for the system with a two stage orifice valve The advantages of these modified version in comparison with simple controllers that we have seen in the previous sections are not mentioned in the article Therefor according to cascade control methodology recommended by str m and H gglund 2006 the probable advantages of having these controllers will be verified in this section Compressor MN Gas Cooler Temperature set point Y Internal Heat Exchanger Evaporator Controllable valve o Evaporator temperature control COP control Figure 4 22 Control System for R744 cycle 44 In the first method they use the same COP optimization approach mentioned in section 4 1 but instead of the low pressure or temperature control they propose a tighter control by nesting both the low pressure and evaporator outlet air temperature control loops in a cascade loop Figure 4 22 show
42. ng different inputs to fulfil two control tasks optimization and temperature control have been verified Decoupling between the COP optimization loop and the evaporator outlet air temperature control loop is possible for a wide range of different climate conditions and cooling loads by ignoring the effect of speed disturbances Since at higher compressor speeds the role of the high pressure is less significant on the COP optimization those two control loops should be designed at a low speed to provide better efficiency Secondly the operation of the two stage orifice valve has been studied and compared with a controllable valve If a high COP is desired by using this valve it is necessary to increase the setpoint of the temperature control loop especially at lower cooling loads If we do not need to use the full cooling power the behavior of the cycle with two stage valve is close to the optimized cycle However it can never provide the same efficiency One other drawback of using this valve is the risk of trapping in a limit cycle at some operating conditions This limit cycle has no significant effect on the sensed temperature by passengers but can reduce the average COP In the next step the performance of the first and second generation controllers pro posed by Behr GmbH have been studied It has been shown that in a first alternative 95 56 one of the unnecessary cascade control loops can be removed Design problems of the high p
43. ngine speed Since the compressor of the automotive A C unit acquires its driving force from the engine its power is a function of the engine speed which is a very fluctuating variable thus some kind of control on compressor capacity is needed to compensate engine speed disturbances to satisfy the comfort requirements and avoid temperature variations Furthermore it is needed especially at higher speeds which bring about an undesirable power of the compressor and a very cold evaporator Among the other methods proposed to control the compressor capacity using a vari able displacement compressor McEnaney and Hrnjak 2005 is the most attractive one As indicated by the PH diagram and the discussion of section 2 3 1 cooling power is a function of suction pressure Due to the fact that in the sub critical region 34 where the heat rejection takes place isothermally evaporator refrigerant and air tem perature are linear functions of the low pressure so the swash plate control makes it possible to control cooling power and the evaporator temperature Changing the inclination of the swash plate relative displacement or relative volume of the compressor causes a change of the pressure ratio namely high pressure as well as low pressure Figure 4 6 but the effect of the expansion valve on the high pressure is dominant Figure4 5 thus the relative volume is regarded as a low pressure con troller Concerning the previous section at a constan
44. ntrast to ordi nary programming languages where assignment statements build the body of the code Assignment fix the computational causality but equations have no pre defined causality For example in a assignment statement like z x y the left hand side of the statement is assigned a value calculated from the expression on the right hand side and z has no effect on the value of x and y but an equation may have expressions on both its right and left hand sides for example x y z just describes an equality The translator and analyzer of the simulation engine have to manipulate and sort the equations according to data flow dependencies to determine their order of execution and which components in the equation are inputs and which are outputs Acausal Physical Modeling Modelica supports both two common approaches to modeling in engineering causal or block oriented and acausal component based modeling formalism This differs from other general purpose modeling packages such as Simulink that just use causal modeling methods by means of block diagrams By the help of component and connector definition in Modelica it is possible to built system models without considering the order in which the variables need to be calculated This enables models to be defined in a more general way The key advantage of acausal modeling is that it speeds up the model development process as it simplifies what the user must do Component models are simply drag
45. on of a system with two controllable devices and system with a two stage orifice valve for different cooling loads and in different climate conditions e Examination of the Behr alternative controllers e Simulation of the two evaporator system in different climate conditions and cooling loads in order to investigate the problems of the SISO control for this system Thesis Outline Chapter 1 gives an overview of different parts of A C systems The thermodynamical principle of the process can also be found in this chapter In order to figure out the control operation it is necessary to understand these fundamentals As mentioned earlier modeling of the system was carried out with the Dymola A C library The modeling principles of this language and properties of the library are explained in chapter 2 In chapter 3 control algorithms are discussed and examined through several simulations Chapter 4 talks about two evaporator structure in the four zone A C system and cooling power distribution between them In the last part a summary of the report and conclusions can be found Chapter 2 Overview of Automotive Air Conditioning 2 1 Air Conditioning Unit To provide the passengers in a car with a comfortable cabin the automotive air conditioning A C unit should handle unpleasant effects of temperature humidity airflow and heat radiation To perform this task the A C unit needs to cool and heat air dehumidify and distribute the air
46. plistic often linear models for control design ThermoFluid provided accurate dynamic models but could not be used for steady state tasks Air Conditioning bridges that gap and is suited both for dynamic and steady state design computations eliminating the need for multiple platforms and models Tummescheit et al 2005 Three levels of models are available in the A C library e Ready to run templates for components and cycles for example template for standard R744 refrigeration cycle e Component models for drag and drop e Base classes for fully user defined models Users can connect component as they desire which makes it easy to build also non standard configuration such as two evaporator internal heat exchange and losses at any place in the cycle Among the others the standard model of these main components are available in A C the library Heat exchangers The heat exchanger model is composed from two fluid objects air and refrigerant and one wall element The wall mass is determined from detailed geometry input data and therefore reflects distributed capacities Heat conduction in the solid material is modeled one dimensional and perpendicular to both fluids Heat is transferred between wall and fluid using a heat connector class Figure 3 3 26 Wall e ES i refngerant Figure 3 5 Heat exchanger modeling Compressors To model different types of A C compressors mass flow and change of enthalpy are
47. r uses fresh air for ventilation and the rear compartment has just one zone The outlet air of the front evaporator enters the car cabin it is mixed with recirculation air of the rear compartment and then enters the rear evaporator for the second phase of cooling Figure 5 3 shows the structure of this model AirTemper Relol_Co duration 2000 NI Cond_tair duration 2 Cond_mair duration 2 Cond_phi duration 5 airlnitphiCond MOLISSELI Y 00 uogeanp s uoneanp o TA Figure 5 3 Model of the two evaporator system airlnit hd or uoenp MO JSSEJLI Sumuadosajea ooz 52 5 2 Simulations Cooling Power Distribution To compare the cooling power of the one evaporator system with the two evaporator one both systems are simulated under the same cool ing load and at the same operating conditions typical European climate conditions of Table 4 1 Figure 5 4 illustrates that the summation of the capacity of the front and rear evaporator is equal to the capacity of one evaporator system in this condi tion It also shows that the outlet air temperature of the front evaporator is same for both cases With a perfect model which includes the corresponding effects of the rear compartment on the front one this distribution scheme may change a little and more compressor work will be needed to keep the front evaporator temperature constant
48. re at 16 Hz bar 30 90 35 97 40 106 48 118 Another parameter which is expected to influence the optimum high pressure is the compressor speed because the power of the compressor and the low pressure hence the COP and cooling power are nonlinear functions of the speed Figure 4 2 Figure 4 3 illustrates significant effect of the speed on the optimum high pressure Based on Figure 4 3 it can be concluded that e Both the COP and cooling power are strong function of the compressor speed e Higher compressor speed gives lower COP but higher cooling power 31 3000 3 95 3 9 _ 2500 S Y 3 85 un E o a B 2000 ri o 38 o o C oO o E 3 3 75 1500 3 71 1000 n i 3 65 i 10 20 30 40 10 20 30 40 Compressor speed Hz Compressor speed Hz Figure 4 2 Effect of compressor speed on shaft power and low pressure COPs at Tgc 35 C Cooling power at Tgc 35 C 2 r 6000 1 8 F 5000 1 6 9 4000 a 14 f E S I T E e EIL E 3000 eem dee P Qe p 8 jg 1 P eee 1 8 POS E 2060 keresne A a fte 16 Hz 0 8 I eem dee s i eee 26 Hz 36 Hz i i 1000 0 6 0 8 1 1 2 1 4 0 6 0 8 1 1 2 1 4 High pressure Pa x10 High pressure Pa x10 Figure 4 3 Comparison of COP and cooling power with the change of speed e The sensitivity of the COP to the high pressure variations is larger at low speeds and there exists a different optimum pressure for e
49. ressure control loop in the second alternative have also been detected especially the drawbacks of the feedforward compensator design in the controller structure At last a brief description of the two evaporator system has been given then through some simulations it has been concluded that the same approach of SISO control for the one evaporator system is also possible to apply for this system and in order to change the capacity of the rear evaporator a variable speed fan can be used Future works Since the A C system with a two stage orifice valve is highly nonlinear proper tuning of the PID controller which is used for the capacity control is another issue which can be studied It is expected that a systematic PID parameter optimization method increases the average COP for the whole driving cycle The structure and internal mechanism of the two stage orifice valve can be modified to achieve better performance For example to make the COP closer to the optimized cycle COP it is required that the valve opens at different pressures depending on different climate conditions A tunable valve is suggested to make this possible The problem of the limit cycle event can be also solved by changing the bypass opening mechanism For the two evaporator system the possibility of the COP optimization by using controllable valve has not been studied in this thesis The drawbacks of the limit cycle problem of the two stage orifice valve are not investi
50. s structure of this controller Basic rules to have a reasonable cascade control suggested by str m and H gglund are examined for this controller e Rule 1 There should be a well defined relation between the primary and secondary measured variables In the subcritical two phase region the low pressure and refrigerant tempera ture have such a relation Since the evaporator outlet air temperature sensor is inserted into the evaporator the relation between evaporator outlet air and refrigerant temperature is also linear Therefore such a well defined relation exists between the low pressure and evaporator outlet air temperature e Rule 2 Essential disturbances should act in the inner loop Since the low pressure and temperature are correlated according to balance principle in section 2 3 1 it is not possible to distinguish between the effect of disturbances on the inner and outer loop except for engine speed disturbances which acts directly on the compressor power and changes the low pressure firstly Therefore this rule just holds true on the effect of engine speed disturbances e Rule 3 The inner loop should be faster than the outer loop The typical rule of thumb is that the average residence times should have a ratio of at least five The thermal conductivity of R744 is very high and the design of the evaporator facilitates the system with a high heat transfer rate in the subcritical region as a result of this fact the air te
51. sors used in the automotive A C reciprocating scroll screw and centrifugal The most preferred one in R744 systems is the variable displacement swash plate compressor which belongs to the reciprocating category In reciprocating compressors the refrigerant vapor is compressed by a piston located inside a cylinder The piston is connected to the crankshaft by a rod As the crankshaft rotates it causes the piston to travel back and forth inside the cylinder A suction valve and a discharge valve are used to trap the refrigerant vapor within the cylinder during this process The piston travels away from the discharge valve and creates a vacuum effect Reduction in the pressure within the cylinder to below suction pressure forces the suction valve to open and the refrigerant vapor is drawn into the cylinder The piston reverses its direction and travels toward the discharge valve compressing the refrigerant vapor the suction valve is then closed and traps the refrigerant vapor inside the cylinder As the piston continues to travel toward the discharge valve the refrigerant vapor is compressed The discharge valve is forced to open and the compressed refrigerant vapor leaves the cylinder To have control on the capacity of the compressor variable displacement Crankshaft Shoe Shaft pressure Intake pressure Cylinder CY Lug plate Swash plate Piston Control valve Figure 2 11 Variable displacement compressor by Toyota Industry 17
52. ssure For example any increase of the compressor speed causes higher high pressure and lower low pressure which in turn decreases the evaporator temperature Therefore by the high pressure control the input power of the cycle is maintained and 47 the effects of the first group of disturbances on the temperature can be suppressed so the two first rules of cascade control are satisfied But if any items of the second group changes according to the balance principle of the refrigeration cycle in section 2 3 1 both the low and high pressure will vary In this case the high pressure variation is not a result of the input power variation so when the controller tries to maintain the input power based on the high pressure measuring the evaporator temperature will vary in a wrong direction Hence the high pressure controller set point needs to be corrected to give the proper control action The role of the feedforward compensator in this configuration is to correct the high pressure set point in proportion of the second group of parameters variations In the Lochmahr et al 2005 proposed controller four parameters are entered to feedforward block engine speed temperature set point inlet air temperature and mass flow rate It is investigated that this design has these problems e Since the high pressure controller inner loop is fast enough to compensate the effect of speed disturbances insertion of the speed measured value in the feedfor
53. ssure is more pronounced e For each ambient gas cooler air inlet temperature there is an optimum high pressure which results in the maximum COP With the increase of the ambient temperature the optimum pressure increases The exact optimum values are shown in Table 3 2 e The cooling power increases with high pressure but there exists a maximum capacity Since the condenser gas cooler in R744 system is located in the front of the car and exchanging the heat directly with the environment the gas cooler air inlet temperature is assumed to be equal to the ambient temperature However in some particular conditions such as an idle car which is placed in exposure of sun radiation these two temperatures might differ a lot 30 COPs at 16 Hz compressor speed Cooling powers at 16 Hz compressor speed l r 5000 r r 45 4000 l Ej 3000 o 8 E 2000 o 3 0 5 1000 0 i i 0 l 0 6 0 8 1 1 2 1 4 0 6 0 8 1 1 2 1 4 High pressure Pa x10 High pressure Pa x10 Figure 4 1 Comparison of COP and cooling power with the change of high pressure e The optimum high pressure which brings about the maximum COP can not provide the maximum cooling power but the cooling power at this pressure is close to its maximum value especially at higher ambient gas cooler air inlet temperatures Table 4 2 Optimum high pressure Gas Cooler temperature C Optimum high pressu
54. t speed it is acceptable to DON Low Pressure High Pressure Pressure Pa 0 4 0 5 0 6 0 7 0 8 0 9 1 Relative Volume Figure 4 6 Effect of Compressor relative volume on pressures neglect the cross coupling between the first SISO loop which tries to maximize COP by high pressure control and the second one which aims to control the low pressure evaporator air outlet temperature but it is not satisfactory to decouple these loops in the case of speed changes Assuming constant speed control of the evaporator air outlet temperature in the case of low cooling load can improve the COP significantly Figure 4 7 shows the advantage of this control at a low speed and low cooling load Figure 4 8 shows that in these operating conditions without any control on evapo rator air outlet temperature control it goes a couple of degrees below zero which is not desired for low cooling loads but by the low pressure control it remains in an acceptable range As mentioned earlier in the case of speed changes to control the evaporator air outlet temperature and optimize the cycle at the same time a control strategy is needed which considers the cross coupling between these two loops 35 AL Low pressure controlled TTrrm Low pressure controlled Fixed relative volume Fixed relative volume 3 5 4500 4000 o a Q 3500 O Oo 8 O 3000 Mi
55. the ambient temperature and compressor speed is suggested For the purpose of simpli fication the effect of speed is neglected and the controller is reduced to a controller which works just based on the ambient or gas cooler temperature and is designed at a low speed since at higher speeds of the compressor the role of the optimum high pressure is less significant 33 An electronic expansion valve like a PWM valve can be used as an actuator to change the flow rate to achieve desired the high pressure The model of a typical variable Kv valve has been described in section 3 3 Figure 4 5 shows the effect of variation of Kv value of this valve on both pressures at a constant speed and temperature 2 E Ro 00 cCORC o m dumme Low Pressure High Pressure Pressure Pa T 0 5F 0 0 02 0 04 0 06 0 08 0 1 0 12 Valve Kv Figure 4 5 Pressures against the valve Kv 4 2 Evaporator Temperature Control So far besides COP considerations getting maximum cooling power has been the main purpose of control of the cooler unit of the A C system especially in extreme conditions such as very high ambient temperature or high cooling loads however in some other conditions it is necessary to control the compressor power to keep the cooling power in the acceptable range for the A C system and not let it reach its maximum possible capacity These conditions are low cooling load and or high e
56. the automotive industry thus based on the experiment results an improved logic for control of this system is proposed In the second part it is investigated whether the one actuator control approach is applicable to a system consisting of two parallel evaporators to allow passengers to control different climate zones The simulation results show that in the case of using a two stage orifice valve for the front evaporator and a fixed expansion valve for the rear one a proper distribution of the cooling power between the front and rear compartment is possible for a broad range of climate conditions i 111 Table of Contents Table of Contents Acknowledgements 1 Introduction 2 Overview of Automotive Air Conditioning 2 1 Air Conditioning Unit AS ARA a ed 2 2 Refrigeration Cycle ola mas 2 2 Subcritical Cycle ti ia Aa e 2 2 2 Tra ns ritical Cycle ed o a ew Ue ae Ib e 2 3 Refrigeration Capacity and Power 0 2 3 1 Balance of Refrigeration Cycle o 24 Control Systemi ta gi ids rA sea doc a bo ee Oo ted e eu dy od 2 4 1 Controllable Components eur bre odode Be eke 3 System Modeling and Simulation Tools 3 1 Modelica Language pra ead CLS ae A a ee oe S 3 2 Dymola Environment 20000002 eee 3 3 Modelon Air Conditioning Library 4 Optimization and Control 4 1 Optimum High Pressure Control o 4 2 Evaporator Temperature Control
57. the compressor relative vol ume is limited and the integral action should be used in both the inner and outer control loops therefore it is necessary to have a scheme to avoid inte gral windup str m and H gglund 2006 This requires being able to inject a tracking signal into outer loop The connection place of this tracking signal is shown in Figure 4 24 which should be added to the original configuration e In the case of high pressure control and in the vicinity of 73 bar pressure difference when the two stage orifice valve changes its flow path bi stability phenomenon takes place In this case any disturbances which leads to small variance in the pressure difference cause the valve to jump to the alternate path while the high pressure is kept constant by the controller Therefore the system is able to exist in either of two steady states while the high pressure is fixed Figure 4 25 shows that a small disturbances of the pressure pushes the system to another steady state and causes significant change in the cooling power Although this will be compensated by the outer loop later on it is another situation where the high pressure loop acts against the main purpose of control 110 5 1 4000 T 110 09 2 o 109 5 loci gt 0 8 3500 o 2 S 409 eir 07 7 E 3000 a 108 5 a 8 0 6 lt C D o T 108 A eei 5 0 5 O 2500 107 5 50 4 O 107 2000 o 1000 2000 o 1000 2000 o 1000 200
58. the enthalpy at the outlet Since the adiabatic compression increases the enthalpy the compressor work per formed on the fluid can be represented as Boss AG 2 3 4 Where hg is the discharge enthalpy The efficiency of the cycle is defined by a quan tity named the Coefficient Of Performance COP It is the ratio of the absorbed heat from the cooled space to the amount of compressor work required for this absorption Des Po Higher COP values indicate that more heat is removed for a given amount of work COP 2 3 5 The COP is not only a function of the system features but also a function of the operating conditions such as the car cabin and environment temperature 11 2 3 1 Balance of Refrigeration Cycle According to the heat transfer principle the amount of the heat which is removed from air by the evaporator Qea should be equal to the evaporator heat absorbtion capacity on the refrigerant side Qer 2 3 3 The absorbed heat on the air side Qea is proportional to the difference between the refrigerant temperature Ter and the air temperature Ter By air here we mean warm air which is entering the evaporator inlet air not cooled air after the evaporator outlet air Qea 73 Qe Ca CealTea Ter 2 3 6 Substituting G from the 2 3 1 in the equation 2 3 3 the evaporator capacity on the refrigerant side is 6 E Us Qe ha hi 2 3 7 Qe Evaporator temperature efficacy V Compressor suction volu
59. ve volume based on the measured value of the outlet air temperature of both evaporators and two desired values Figure 5 2 Gas cooler Front Evaporator Temp c E Controller Temp Sensor RJ gt ear Evaporator Fixed orifice Figure 5 2 Two evaporator temperature control The easier way to control the cooling capacity of the rear evaporator is to use a variable speed fan and change the air flow around the evaporator while the temperature of the front evaporator is controlled with the compressor relative volume variation The same SISO approach as for the one evaporator system Based on the balance principle in section 2 3 1 this will change the balance point of the rear evaporator low pressure and this in turn changes the front evaporator low pressure Since the temperature controller tries to fix the front evaporator low pressure it is expected that under certain cooling loads this control plan develops limit cycle behavior In this section some simulations will be done to investigate whether this phenomenon happens or not 51 5 1 Two Evaporator System Model To model the four zone system perfectly a car cabin model is needed which includes the air interchanging among the different zones Since we do not have such a com plete model yet the simple car cabin model of the A C library is used to model the air mixing of the front and rear evaporator In this model it is supposed that the front evaporato
60. ward compensator is not required Furthermore according to the multi variable modeling approach in Rasmussen 2002 high pressure is a function of speed and four other states besides this if the nonlinear behavior of the valve in this system is considered the relation between speed and high pressure is highly nonlinear therefore adding a simplified feedforward transfer function for the speed may tune the high pressure in a limited range of operating conditions but increases the risk of overcompensating in other conditions e The gas cooler boundary conditions are not measured in this scheme and not entered in the feedforward compensator consequently before the outer loop reacts to the resultant temperature variation and compensate the effect of these disturbances because of the fixed set point the high pressure controller acts falsely For example if the ambient temperature increases then as a result of the balance principle for the gas cooler the high pressure will increase therefore the high pressure controller decreases the compressor power and the cooling power while in this case higher power is needed to maintain cabin temperature e Humidity of the cabin is also ignored in this design Due to the fact that the outlet temperature is a nonlinear function of the air temperature and humid ity such a feedforward compensator can not cover a wide range of different humidities 48 e The output signal of the secondary controller i e
61. wer variation against the air mass flow variation under a high and a low cooling load High Load Low Load 5000 3000 2 Rear evaporator 4000 S IH T RENS EYED rc eee ES 2500 Front evaporator o i 3 3000 ee al 2000 a 1 o g 20001 ie TM M UH 1500 o 1000 ERN CE 1000 0 500 0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000 Time Time Figure 5 6 Rear evaporator air flow change It is seen that the rear cooling power is changed while the front one is almost kept constant At lower cooling loads the rear evaporator capacity is more sensitive to the air mass flow can change 54 High Load Low Load 5000 r 3000 Rear evaporator 4000 p 0 A J 2500 H Eron evaporator a 3000 MER eat od 2000 c dass dius k 8 EL a 2000 E i mec 4500 ada e o o o 1000 A EN o J 1000 A ee 0 500 0 1 0 15 0 2 0 25 0 1 0 15 0 2 0 25 Rear air mass flow Kg s Rear air mass flow Kg s Figure 5 7 Rear evaporator air flow change Therefore at these conditions using a two stage valve besides the front evaporator temperature control is possible while the capacity of the rear evaporator is controlled by means of air mass flow changing Chapter 6 Summary and Conclusions In this thesis some control issues of R744 A C system have been studied First of all the problem of decoupling amo
62. xcess heat from gas leaving the gas cooler and the outer part which contains the refrigerant coming from the evaporator will absorb this excess heat After the internal heat exchanger the refrigerant leaves the high pressure section by passing through an expansion device and at a low pressure and temperature enters the evaporator to absorb heat from the cooled space Finally in the inner part of the internal heat exchanger it absorb more heat then the superheated refrigerant is ready to become compressed again 10 2 3 Refrigeration Capacity and Power The PH diagram is depicted for refrigerant per unit weight to obtain actual capacities and powers the refrigerant charge circulating in the cycle must be known The mass flow rate of a fluid G Kg h that is circulating in a cycle by power of a compressor is given as Gr Ve 2 3 1 2 3 1 Where V m h is the compressor suction volume and v m kg is the specific vol ume of the fluid and the compressor suction volume is obtained as below Watanabe 2002 V x N x 60 poems 2 9 2 106 x n V Compressor cylinder volume cc N Compressor speed rpm w Volumetric efficiency The cooling power can be represented as the change of enthalpy of the refrigerant when the evaporator absorbed heat from the cooled space car cabin Hence for the subcritical system it is computed by Qer E h4 P hi G 2 3 3 Where h is the enthalpy at the evaporator inlet and h is

Open Issues in Control of Automotive R744 Air

Contents

Download Pdf Manuals

Related Search

Related Contents