Energy 36 (2011) 6110 6120
`
`Contents lists available at ScienceDirect
`
`Energy
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e n e r g y
`
`The refrigerant R1234yf in air conditioning systems
`Claudio Zilio a, *, J. Steven Brown b, Giovanni Schiochet a, Alberto Cavallini a
`a Dipartimento di Fisica Tecnica, Università di Padova, Via Venezia, 1, I-35131 Padova, Italy
`b Department of Mechanical Engineering, Catholic University of America, Washington, DC 20064, USA
`
`a r t i c l e i n f o
`
`a b s t r a c t
`
`Article history:
`Received 1 March 2011
`Received in revised form
`2 August 2011
`Accepted 4 August 2011
`Available online 27 August 2011
`
`Keywords:
`Air conditioning
`Automobile
`R1234yf
`R134a
`
`1. Introduction
`
`Experiments were conducted for a typical R134a compact European automotive air conditioning system
`equipped with an internally controlled variable displacement compressor, minichannel condenser, TXV,
`and minichannel evaporator. A “drop in” R1234yf system was tested together with two modified R1234yf
`systems with the primary goal to document some laboratory results and their analyses which could
`prove useful in aiding manufacturers and researchers by indicating “minor” system modifications which
`could be implemented in existing air conditioning systems, with the aim to achieve with R1234yf similar
`capacity and efficiency as modern R134a systems. Since the experimental results indicate that, for a given
`cooling capacity, R1234yf systems present lower performance than the baseline R134a, numerical
`simulations were used to investigate the effects of “major” system modifications, such as, the use of an
`enhanced condenser and/or an enhanced evaporator.
`
`Ó 2011 Elsevier Ltd. All rights reserved.
`
`Over the last several years, much research and development
`effort has been focused on potential refrigerants possessing low
`Global Warming Potentials (GWPs). As discussed in [1], the catalyst
`for much of this effort can be attributed to European regulations
`regarding the use of R134a (GWP relative to CO2 based on a 100
`year time horizon, which is the reference that will be used
`throughout this paper, is 1430 [2]) in automotive applications. In
`particular, the European Union’s F gas regulations [3,4] specify that
`beginning on January 1, 2011 new models and on January 1, 2017
`new vehicles fitted with air conditioning cannot be manufactured
`with fluorinated greenhouse gases having global warming poten
`tials (GWP) greater than 150. Possible candidate refrigerants that
`possess GWP < 150, and that are being considered, include R 152a,
`R 744 (CO2), and R1234yf. R 152a has been investigated in the
`recent years in several applications [5], anyway, if used in air
`conditioning applications, would likely be implemented in
`a secondary loop system because of its flammability. R 744, if used,
`would likely be implemented in a transcritical cycle requiring
`significant modifications to the currently used automotive air
`conditioning systems [6] because of its considerably different
`thermophysical properties when compared to R134a. Among the
`fluorinated propene isomers which have normal boiling point
`
`* Corresponding author. Tel.: þ39 049 827 6893; fax: þ39 049 827 6896.
`E-mail address: claudio.zilio@unipd.it (C. Zilio).
`
`see front matter Ó 2011 Elsevier Ltd. All rights reserved.
`0360-5442/$
`doi:10.1016/j.energy.2011.08.002
`
`temperature data published in the public domain, several have low
`GWP and normal boiling temperatures relatively close to R134a
`(from about 3.7 C lower to about 8 C higher); however, among
`them, R1234yf is the one closest to commercialization (its normal
`boiling temperature is approximately 3.7 C lower than that of
`R134a). It has a GWP of 4 [7] and is being widely considered as
`a possible replacement for R134a in automotive applications or as
`a component in non azeotropic mixtures for heat pump applica
`tions [8].
`To date, researchers and manufacturers have focused their
`R1234yf research and development efforts primarily on charac
`terizing its flammability, toxicity, environmental impact, materials
`compatibility, oil compatibility, air conditioning system perfor
`mance, thermophysical property data, and in the developing of
`simple equations of state (EoS). Excluding thermophysical property
`data and EoS modeling, much of the remaining published work has
`been presented by chemical manufacturers and vehicle manufac
`turers through organizations such as VDA (German Association of
`the Automotive Industry), SAE (Society of Automotive Engineers),
`and JAMA (Japan Automobile Manufacturers Association). For
`example, VDA organizes an annual alternative refrigerant winter
`meeting where chemical manufacturers and vehicle manufacturers
`have presented some of their R1234yf work beginning with its
`February 2008 meeting. SAE [9] through its Cooperative Research
`Program CRP 1234 1, CRP 1234 2 and CRP 1234 3 has focused its
`efforts on safety and risk assessment, air conditioning system
`performance, and materials compatibility, with much of this work
`having been presented at an annual automotive alternate
`
`Page 1 of 11
`
`Arkema Exhibit 1103
`
`

`
`Nomenclature
`
`y
`
`C. Zilio et al. / Energy 36 (2011) 6110 6120
`
`6111
`
`Variable that represents either COP, Cooling Capacity,
`or _m
`
`
`
`
`COP
`CRP
`DISP
`DISP*
`EoS
`GWP
`_m
`N
`OCR
`P
`PAG
`POE
`Q
`R
`RH
`TXV
`T
`
`Coefficient of Performance
`Cooperative Research Program
`geometric displacement volume per revolution (cm3)
`displacement volume per revolution (cm3)
`Equation of State
`Global Warming Potential
`1)
`mass flow rate (kg s
`rotational speed (RPM)
`Oil Circulation Ratio (%)
`pressure (kPa, MPa)
`Polyalkylene Glycol
`Polyol Ester
`cooling capacity (kW)
`(hv DISP/DISPmax)  100
`Relative Humidity (%)
`Thermal Expansion Valve
`temperature (C)
`
`refrigerant systems symposium beginning with its June 2008
`meeting. In the previously mentioned work, the researchers, for the
`most part, have concluded that R1234yf appears to be a promising
`candidate to replace R134a in automotive applications, with SAE
`[10] concluding R1234yf “can be used as the global replacement
`refrigerant in future mobile air conditioning systems and it can be
`safely accommodated through established industry standards and
`practices for vehicle design, engineering, manufacturing, and
`service.” Notwithstanding the preceding statement, several issues
`are still being investigated through ongoing research and devel
`opment work. R1234yf, for example, has somewhat lower ther
`modynamic performance (e.g., Bang [11]) showed that the cooling
`capacity of R1234yf is 3e7% lower and its COP is 1e3% lower than
`R134a when used as a “drop in” replacement for R134a. However,
`Meyer [12] showed that with relatively simple system modifica
`tions the cooling capacity and COP of R1234yf could be made
`approximately equal to the baseline R134a values. Along these
`same lines, more recently, Petitjean and Benouali [13] experimen
`tally studied the effects of using an improved condenser, an
`improved evaporator, an adjusted (“tuned”) Thermal Expansion
`Valve (TXV), a modified compressor, and a liquid line/suction line
`heat exchanger (llsl hx) in an automotive air conditioning system.
`Regarding the conventional heat exchangers, they concluded that
`the condenser plays a much more important role than the evapo
`rator in the optimization of the R1234yf performance. This also can
`be seen from Cavallini et al. [14] who showed analytically and Del
`Col et al. [15] who showed experimentally that the condensation
`heat transfer coefficients for R1234yf can be up to 30% or so lower
`than R134a at high vapor qualities at the same mass flux and
`saturation temperature; however, Del Col et al. [15] showed
`experimentally that the condensation heat transfer coefficients for
`R1234yf are nearly the same as those of R134a for lower vapor
`qualities or for single phase liquid heat transfer. Finally, Petitjean
`and Benouali [13] also show that to optimize the R1234yf perfor
`mance, the TXV needs to be properly tuned, that the use of a llsl hx
`has a bigger impact for R1234yf than for R134a, and that the
`compressor capacity needs to be properly adjusted. In conclusion,
`they show that with properly designed and tuned components, the
`R1234yf performance can match or even better that of R134a.
`A second issue is the flammability of R1234yf, where testing has
`shown that, while flammable, a significantly larger amount of
`energy is required to ignite the refrigerant than for other common
`Class 2 refrigerants [16]. While Addendum w to ANSI/ASHRAE
`
`Greek Symbols
`yR134a
`yR1234yf
`yR134a
`volumetric efficiency (%)
`3)
`density (kg m
`
`3
`
`hV
`r
`
`Subscripts
`air
`air
`disch
`discharge
`evap
`evaporator
`in
`inlet
`max
`maximum
`out
`outlet
`suc
`suction
`sup
`superheat
`
`Standard 34 2007 [17] specifies R1234yf with a Safety Classification
`of A2, it is a likely candidate to receive a 2L Flammability Subclass
`defined in Addendum ak to ANSI/ASHRAE Standard 34 2007 [17].
`Note: the 2L Flammability Subclass includes Class 2 refrigerants
`1.
`with burning velocities 0.10 m s
`A third issue is the environmental performance of R1234yf.
`Koban [18] performed Life Cycle Climate Performance (LCCP)
`simulations and has shown under her assumptions that R1234yf
`results in reductions of 17e20% when compared to R134a.
`Finally, the compatibility of R1234yf with common materials
`and oils appears promising [16].
`Recently, several research groups have begun publishing
`R1234yf thermophysical property data. For example, Tanaka and
`Higashi [19] measured its critical properties, saturation pressure
`from 310 to 360 K, and surface tension from 273 to 340 K. Tanaka
`et al. [20] measured its liquid isobaric heat capacity and its liquid
`density from 310 to 360 K at pressures up to 5 MPa. Then, Akasaka
`et al. [21] developed both a Patel Teja EoS and an Extended Cor
`responding State EoS based on their research group’s data. Di Nicola
`et al. [22] measured saturation pressures from 224 to 366 K. Fedele
`et al. [23] measured saturation pressures for reduced temperatures
`ranging from 0.67 to 0.93. Brown et al. [24,25] developed a simple
`Peng Robinson (P R) EoS model coupled with group contribution
`estimation techniques to easily and relatively accurately predict
`thermophysical property data, which were then used to make
`performance predictions in a typical automotive air conditioning
`system [26]. Leck [27] used his employer’s data to develop
`a Martin Hou EoS and then used it to predict the performances for
`several refrigeration examples. Hulse et al. [28] measured its critical
`temperature, saturation pressure, liquid density, ideal gas heat
`capacity, liquid viscosity, and surface tension, and then developed
`an Extended Corresponding States EoS using their data.
`
`2. Motivation for this study
`
`Given the above, particularly the facts that (1) Zilio et al. [26],
`through a numerical approach, and Petitjean and Benouali [13],
`through an experimental approach, demonstrated that it is theo
`retically possible to achieve nearly the same cooling capacity and
`COP for R1234yf and R134a with some system modifications, (2)
`the deadlines established by European regulations are fast
`approaching, and (3) the sponsors of SAE CRP 1234 have concluded
`that R1234yf appears to be a promising global replacement
`
`Page 2 of 11
`
`

`
`6112
`
`C. Zilio et al. / Energy 36 (2011) 6110 6120
`
`refrigerant for future automotive systems, one of the principal
`questions that should interest automotive and component manu
`facturers, and which this paper will attempt to address, is if and
`what “minor” system modifications would allow R1234yf to have
`equal cooling capacity to R134a, while maintaining similar COP
`values between the two systems, particularly for vehicles smaller
`than ones considered in SAE CRP 1234.
`It is worth mentioning that the transition from R134a to R1234yf
`and the accompanying system modifications most likely will not
`occur “instantaneously”, but will occur over several years similar to
`the way the transition from R12 to R134a happened. It is also worth
`remembering (see, e.g. [29]), that automotive manufacturers had to
`make several system changes to accommodate R134a, including:
`(1)
`introducing condensers with approximately 20% higher
`condensing capacity in order to maintain similar operating pres
`sures, and thus allowing for the same compressors to be used and
`(2) for TXV systems the valve settings had to be adjusted. Moreover,
`as is well known, the mineral lubricants used with R12 systems had
`to be replaced by synthetic POE or PAG oils (Note: it appears that
`R1234yf will not require such a “revolutionary” change in the
`lubricant). Other R134a system improvements have been intro
`duced over time, including the now widespread use of minichannel
`heat exchangers and variable displacement compressors.
`To date, SAE CRP 1234 1,2,3 are the only publically available
`studies from a large consortium of worldwide companies that had
`the purpose to assess the performance potential of R1234yf in
`existing automotive systems which have undergone only “minor”
`changes. In particular, the baseline system used in SAE CRP 1234
`was the same one considered in SAE CRP 150, namely, a Cadillac CTS
`with a fixed displacement compressor and a cross charged TXV.
`However for SAE CRP 1234, two “minor” changes were made to the
`baseline system in order to obtain the same cooling capacity as
`R134a, namely, (1) the TXV spring preload was adjusted to increase
`the bulb pressure to a corresponding temperature of 0 C and (2)
`the compressor suction line diameter was increased. The SAE
`studies do not report results for a purely “drop in” system. In
`a different study (SAE ARCRP [9]), the not so minor change of
`including a liquid line/suction line exchanger was considered for
`the R1234yf system.
`While the systems used in the previously mentioned SAE
`studies contained fixed displacement compressors, nowadays,
`many vehicles are equipped with variable displacement compres
`sors, where the stroke is modulated by the crankcase pressure.
`With this discussion in mind, in this paper, the performance
`potentials of R134a and R1234yf are compared in a typical Euro
`pean compact automobile equipped with an internally controlled
`variable displacement compressor and minichannel condenser and
`evaporator.
`In particular, experimental data from the baseline
`R134a system are compared to data from three systems: (i) the
`baseline system operating with an optimal charge of R1234yf with
`the baseline TXV setting, (ii) the system of i with an optimal charge
`of R1234yf with a TXV setting optimized for R1234yf, (iii) the
`system of ii with the variable displacement compressor control
`valve deactivated. The experimental part of the analysis outlines
`the limitations of systems i, ii, iii and indicates strategies for system
`performance enhancement through “minor” system modifications.
`Then, one possible enhanced system is further investigated using
`a simulation model which was validated using the experimental
`results.
`
`3. Experimental apparatus
`
`Fig. 1 shows a schematic of the experimental apparatus, which
`consists of four basic systems: (1) a closed loop air circuit for the
`evaporator, including centrifugal fan, electric resistance heater,
`
`Fig. 1. Schematic of experimental apparatus.
`
`humidification equipment, and associated controls, (2) a closed
`loop air circuit for the condenser,
`including centrifugal
`fan,
`chiller, electric resistance heater and associated controls, (3) the
`refrigerant circuit, including a minichannel evaporator, a mini
`channel condenser, a variable volume swash plate compressor, and
`a TXV, all components from the air conditioning system of a typical
`European compact automobile, and (4)
`the instrumentation
`system.
`The inlet air temperatures for both the evaporator and the
`condenser are controlled using two closed loop air circuits, whose
`ductwork configurations provide uniform temperature and velocity
`profiles at the inlet faces of the heat exchangers. The volumetric air
`flow rates are measured using ISA 1932 nozzles per ISO 5167
`4:2003 [30] to accuracies of 0.8% of the measured values. A chiller
`supplies conditioned water to a cross flow heat exchanger to
`maintain the condenser inlet air temperature to near the desired
`value. Then, a PID controlled electric resistance heater located
`upstream from the condenser controls the inlet air temperature to
`the desired value. A separate, but similar, system controls the
`evaporator inlet dry bulb air temperature to the desired value. An
`industrial vapor generator located sufficiently far upstream of the
`evaporator to ensure uniform conditions at the evaporator inlet
`face controls the relative humidity. A grid of nine evenly spaced
`type T thermocouples for the condenser and six for the evaporator
`are used to measure the mean inlet air temperature for the
`respective heat exchanger. The air downstream from each heat
`exchanger is well mixed prior to its temperature being measured
`by separate grids of nine evenly spaced type T thermocouples. For
`both air circuits and for the refrigerant system, a Pt 100 platinum
`resistance thermometer with an accuracy of 0.02 C was used to
`calibrate the temperature measuring system (consisting of ther
`mocouple, electronic ice point and multimeter) to an accuracy of
`0.05 C. The relative humidity (RH) of the evaporator air circuit is
`measured in two ways: (1) using a capacity transducer with an
`accuracy of 3% RH and (2) using a load cell that measures the
`amount of water condensed in a specified time with an accuracy of
`0.5 g.
`
`Page 3 of 11
`
`

`
`C. Zilio et al. / Energy 36 (2011) 6110 6120
`
`6113
`
`Calibrated type T thermocouples measure the refrigerant
`temperature immediately before and after each major component.
`Diaphragm type differential pressure transducers measure the
`pressure drop across both the evaporator and the condenser to
`accuracies of 0.1 kPa. Strain gauge pressure transducers measure
`the absolute pressures in the evaporator to 0.8 kPa and in the
`condenser to 1.6 kPa.
`The refrigerant enthalpies are determined from measured
`temperature and pressure values, where the refrigerant quantities
`are calculated using a version of the P R EoS of Brown et al. [24,25].
`The predictions used in this paper are from 1% lower to þ1%
`greater when compared to the saturation data of Di Nicola et al. [22]
`over the temperature range from 220 to 360 K. The version used in
`this paper is based on using the normal boiling point temperature
`of Di Nicola et al. [22] and the critical state property data of Tanaka
`and Higashi [19].
`A Coriolis flow meter located in the liquid line measures the
`refrigerant mass flow rate to an accuracy of 0.1% of the measured
`value.
`An asynchronous electric motor drives the compressor. An
`optical system measures the compressor speed to an accuracy of
`0.01% of the measured value and a torsiometer located on the
`shaft that couples the electric motor with the compressor measures
`the torque to an accuracy of 0.2 N m. The expanded uncertainty
`on the mechanical power delivered to the refrigerant is less than 2%
`for compressor speeds of 900 and 2500 RPM and less than 3% for
`compressor speeds of 4000 RPM.
`A desiccant based drier and a sight glass are located between
`the condenser and the mass flow meter.
`The total expanded uncertainties for the airside heat transfer
`rate measurements are 1%. The total expanded uncertainties for
`the refrigerant side heat transfer rate measurements are 4% for
`the evaporator and 3% for the condenser, with the largest
`contributor being the uncertainties in the thermodynamic property
`estimates calculated using the P R EoS of Brown et al. [21].
`It is worth noting that while the tests were not conducted per
`SAE J 2765 2008 [31], they were carried out per ASHRAE and ISO
`standards that ensure accuracies as high as the ones required
`by [31].
`
`4. Test conditions
`
`The system tested is a typical R134a based compact European
`automobile, which had a nominal cooling capacity of 5.8 kW at
`1. Four systems were
`a compressor volumetric flow rate of 7.8 m3 h
`tested: (i) the baseline R134a system, (ii) the same system as i
`operating with an optimal charge of R1234yf with the baseline TXV
`setting, (iii) the system of ii operating with a TXV setting optimized
`for R1234yf, and (iv) the system of iii operating with the variable
`displacement compressor control valve deactivated. While the
`operating conditions were not meant to exactly duplicate actual in
`vehicle conditions, they were chosen to cover typical European
`conditions. In particular, they covered three compressor speeds,
`three ambient temperatures (the inlet air temperature of both the
`evaporator and the condenser), and two relative humidity values
`for the evaporator inlet air stream. The evaporator and condenser
`fan speeds (air mass flow rates) were held constant throughout the
`tests (while this is not the case in a real vehicle, it reduced the
`complexity of the test matrix while still allowing for a comparison
`of the performance potential of R1234yf relative to R134a). Table 1
`provides the operating conditions. Note: in the figures that follow,
`the naming
`convention of Ambient
`Temperature/Relative
`Humidity/Compressor Speed is used. For example, 15/80/900
`represents the condition with an ambient temperature of 15 C, 80%
`relative humidity, and a compressor speed of 900 RPM.
`
`Table 1
`Operating conditions.
`
`Evaporator air inlet
`Evaporator air volumetric
`flow rate (m3 h 1)
`Condenser air inlet
`Condenser air volumetric
`flow rate (m3 h 1)
`Compressor speed (rpm) 900 2500 4000 900 2500 4000 900 2500 4000
`
`35 C & 40% RH 25 C & 80% RH 15 C & 80% RH
`400  3%
`400  3%
`400  3%
`35 C
`25 C
`15 C
`1580  3%
`1580  3%
`1580  3%
`
`5. Results and discussion
`
`More than 140 steady state tests were conducted. The energy
`balances on the airside and refrigerant side for both the evaporator
`and the condenser differed by a maximum of 4%.
`The R134a oil circulation ratio (OCR) was measured per ASHRAE
`41.4e1986 (R2006) [32]. Similar measurements were not per
`formed for R1234yf since the experimental procedure consumes
`considerable fluid, which was not available at the time of the tests.
`Therefore, the R1234yf OCR values were assumed to be those of
`R134a at the same compressor displacement, which are reported in
`Table 2. The refrigerant/Polyalkylene Glycol (PAG) oil enthalpies for
`both refrigerants were calculated per [31].
`The optimal refrigerant charge was determined per [31] for each
`of the four system configurations, where in each case the airside
`conditions were fixed at 35 C and 40% relative humidity for both
`the condenser and the evaporator. The air volumetric flow rates for
`1 and
`the condenser and the evaporator were 1600 m3 h
`1, respectively.
`400 m3 h
`In what follows in the next several paragraphs, particular
`attention will be paid to the variable displacement compressor
`since its internal control valve with factory settings regulates the
`compressor capacity based on R134a. In particular, the compres
`sor’s internal valve control action controls the displacement as
`a function of the difference between the suction and discharge
`pressures, while the compressor attempts to maintain the evapo
`ration pressure at a constant value. The suction pressure is simply
`the evaporator outlet pressure reduced by the suction line pressure
`drop. For fixed air inlet conditions and for a fixed TXV setting, the
`evaporator pressure is a function of the refrigerant’s thermophys
`ical properties and the refrigerant mass flux, which affect both the
`in tube heat transfer coefficient and the refrigerant side pressure
`drop (and thus also the local saturation temperature).
`For the internally controlled variable displacement compressor
`used in this study, the possible working conditions are the ones
`located in the upper part of Fig. 2, that is, operating conditions
`located above line BC.
`The position of line BC is fixed by the control valve setting:
`adjustment of the spring moves the line BC either up or down
`relative to its baseline position. Regardless, operating conditions
`below line BC are not possible. Note that without changing the
`valve setting, given that the R1234yf saturation temperature is
`slightly different from that of R134a, lines AB and CD will result in
`different refrigerant temperatures (since the pressure is fixed).
`Therefore, the first thing to be verified is that line CD will not result
`in a saturation temperature at the evaporator outlet below 0 C for
`the given refrigerant.
`
`Table 2
`Oil Circulation Ratio (OCR).
`
`Compressor speed (RPM)
`900
`2500
`4000
`
`OCR (%)
`3.0
`4.2
`4.5
`
`Page 4 of 11
`
`

`
`6114
`
`C. Zilio et al. / Energy 36 (2011) 6110 6120
`
`Fig. 2. Relationship between the suction and discharge pressures for the internally
`controlled variable displacement compressor.
`
`Now, turning our attention to the compressor discharge, we
`note that the discharge pressure is simply the condenser inlet
`pressure increased by the discharge line pressure drop. The
`condenser includes an integrated liquid receiver in the manifold,
`separating the two phase region from the subcooling region, such
`that if the refrigerant charge is determined per [31], the tubes after
`the separator will be fed with saturated liquid. Cavallini et al. [14]
`showed theoretically, through an exergetic analysis, that for
`a given condenser design (with desuperheating and two phase
`regions) with fixed saturation temperature that the heat flux
`which can be rejected with R1234yf is lower than that of R134a.
`Also, Zilio et al. [26] showed for an ideal vapor compression cycle
`that the COP of R1234yf is lower than R134a, where the gap
`worsens if the actual compressor isentropic efficiencies are taken
`into account. Therefore, for constant cooling capacity for the
`R1234yf and R134a systems, the condensation heat rejection will be
`higher for the R1234yf system. Consequently, the R1234yf system
`will result in a greater saturation temperature at the condenser
`inlet. Therefore, based on the preceding discussion, the discharge
`pressures for R1234yf and R134a will be different, resulting in the
`compressor control valve reacting differently for the two fluids. The
`effect of the compressor’s internal control valve deactivation
`(dashed line in Fig. 2) will be discussed later in the paper.
`
`Fig. 4. Evaporator refrigerant superheat values. Note: for a given label along the
`abscissa, the first number is the ambient temperature in C, the second number is the
`relative humidity in percent, and the third number is the compressor speed in RPM.
`
`The experimental compressor suction pressures for systems i, ii,
`and iii described in Section 4 are shown in Fig. 3. Then, Fig. 4 shows
`the evaporator superheat values for the same systems. As can be
`seen, the “drop in” system has significantly higher superheat
`values than the R134a system. Thus, in order to obtain more
`reasonable superheat values for the R1234yf systems, the TXV was
`modified (system iii, referred to as “TXV tuned” in the figure) from
`its original setting of 0.20 MPa gauge at 0 C to 0.29 MPa gauge at
`0 C. Similar to the evaporator, the experimental compressor
`discharge pressures are shown for the same systems in Fig. 5.
`Fig. 6 shows that the experimental suction line pressure drop as
`a function of refrigerant mass flow rate is very similar for both
`R1234yf and R134a; whereas, Fig. 7 shows that the suction line
`pressure drop for R1234yf is greater than that of R134a for equal
`cooling capacity, which is due to the somewhat lower specific
`latent heat of vaporization for R1234yf.
`Because of the previously discussed issues, SAE ARCRP recom
`mended increasing the suction line diameter; however, for the
`compact automobile system considered herein the suction line
`pressure drop is smaller than the pressure drop of the compressor
`
`Fig. 3. Compressor refrigerant suction pressures. Note: for a given label along the
`abscissa, the first number is the ambient temperature in C, the second number is the
`relative humidity in percent, and the third number is the compressor speed in RPM.
`
`Fig. 5. Compressor refrigerant discharge pressures. Note: for a given label along the
`abscissa, the first number is the ambient temperature in C, the second number is the
`relative humidity in percent, and the third number is the compressor speed in RPM.
`
`Page 5 of 11
`
`

`
`C. Zilio et al. / Energy 36 (2011) 6110 6120
`
`6115
`
`Fig. 8. Relative cooling capacities. Note: for a given label along the abscissa, the first
`number is the ambient temperature in C, the second number is the relative humidity
`in percent, and the third number is the compressor speed in RPM.
`
`effects of this second approach can clearly be seen: the X marks are
`located in an area that is non reachable using the baseline setting.
`Now, turning our attention more in detail to the compressor: for
`a reciprocating positive displacement compressor, the displace
`ment volume is given by:
`
`Fig. 6. Suction line refrigerant pressure drops vs. refrigerant mass flow rates.
`
`manifold. Therefore, modification of the suction line would only
`marginally reduce the overall pressure drop for this system. Also,
`note that modification of the suction manifold would not be
`considered a “minor” modification and thus is not considered in
`this paper.
`Now turning our attention to cooling capacity and COP, Fig. 8
`reports the cooling capacity deviations for systems (i) and (ii)
`relative to the R134a baseline system, and Fig. 9 reports the COP
`deviations for the same systems. In the figures, the deviation is
`ðyR1234yf
`yR134a=yR134aÞ, where y is the cooling
`defined as 3
`capacity, the COP, or the refrigerant mass flow rate.
`As Fig. 8 shows, the R1234yf systemeeven with a tuned
`TXVeseverely lacks cooling capacity. In order to overcome this
`deficit, the compressor displacement internal control valve setting
`was modified using the model of Dou et al. [33]. In particular, their
`model consists of a mechanical fluid dynamic representation of the
`internally driven control valve, where the crankcase pressure and
`the spring force represent the actuation signal for displacement
`control. The signal is a function of Pdisch, Psuc, and of the flow and
`design characteristics of the suction and discharge ports. Thus, one
`possible approach to adjusting the valve control for R1234yf is to
`modify the port characteristics (i.e. flow coefficients and port
`sections). A second possibility is to simply change the valve spring
`preload, which is what the dashed line in Fig. 2 represents. The
`
`(1)
`
`DISP*
`
`hv DISP
`
`60 _m
`rsuc N
`where DISP is the geometric displacement, N is the rotation speed
`in RPM, hv is the volumetric efficiency, rsuc is the suction density,
`and _m is the refrigerant mass flow rate.
`(hv DISP/DISPmax)  100 as a function
`Fig. 10 plots the term R
`of the difference between the discharge and suction pressures,
`where the lines are simply interpolations of the experimental data.
`126 cm3.
`Note: the compressor used in this study has a DISPmax
`Note that at 900 RPM the new valve setting results in the
`maximum displacement possible (it achieves a value of nearly
`100%; recall that DISP* is reduced by the volumetric efficiency).
`
`Fig. 7. Suction line refrigerant pressure drops vs. cooling capacities.
`
`Fig. 9. Relative COPs. Note: for a given label along the abscissa, the first number is the
`ambient temperature in C, the second number is the relative humidity in percent, and
`the third number is the compressor speed in RPM.
`
`Page 6 of 11
`
`

`
`6116
`
`C. Zilio et al. / Energy 36 (2011) 6110 6120
`
`Fig. 10. Relative compressor displacement as a function of the difference between the
`suction and discharge pressures.
`
`Fig. 11 shows the cooling capacity data, including results for the
`modified compressor valve setting. Despite this modification, the
`system is still not able to provide the same cooling capacity as
`R134a at an ambient temperature of 35 C. For an ambient
`temperature of 25 C the same cooling capacity as the R134a system
`is achievable for the 2500 RPM condition but not for the other
`compressor speeds. Finally, for an ambient temperature of 15 C the
`system achieves greater cooling capacity than the R134a baseline
`system for all compressor speeds.
`Fig. 12 shows a large mass flow rate increase for all conditions
`for the modified R1234yf systems due to the combined effects of
`increased DISP* and the greater R1234yf suction density [26].
`The greater cooling capacity for R1234yf at an ambient
`temperature of 15 C results in much lower evaporator inlet
`temperatures for R1234yf (see Fig. 13), where the mismatch
`decreases with increasing ambient temperature. Thus,
`for an
`ambient temperature of 15 C, the R1234yf evaporator outlet
`pressures (see Fig. 14) and temperatures (see Fig. 15) are lower than
`R134a.
`It is worth noting that because of the regulating action of the
`compressor internal control valve, even at ambient temperatures of
`15 C,
`the compressor did not cycle since t