Investigation of an experimental ejector refrigeration machine operating with refrigerant R245fa at design and off-design working conditions. Part 2. Theoretical and experimental results.

Investigation of an experimental ejector

refrigeration machine operating with refrigerant

R245fa at design and off-design working conditions.

Part 2. Theoretical and experimental results

K.O. Shestopalov

, B.J. Huang

, V.O. Petrenko

, O.S. Volovyk

a r t i c l e i n f o

Article history:

Received 15 November 2013 Received in revised form 29 December 2014 Accepted 9 February 2015 Available online 19 February 2015 Keywords:

Ejector

Ejector refrigeration machine R245fa

Test rig

Performance characteristics

a b s t r a c t

The main results of a theoretical and experimental investigation of the performance characteristics of an ejector and an ejector refrigeration machine (ERM) operating with refrigerant R245fa at design and off-design working conditions are presented. The ejector and ERM were explored theoretically using improved 1D model and the calculated results were validated experimentally on ejector test rig that has been assembled and operated at National Taiwan University. For typical cases, the performance characteristics variation with condensing, generating and evaporating temperatures along with performance maps are presented. The theoretical results are compared with the results of a set of experiments and good qualitative and quantitative agreement is observed.

© 2015 Elsevier Ltd and IIR. All rights reserved.

Une

etude du fonctionnement d

'un systeme frigorifique

exp

erimental de r

efrig

eration 

a

ejecteur avec le frigorig

ene

R245fa aux conditions de travail de conception et

hors-conception. 2

eme

partie -R

esultats th

eoriques et

exp

erimentaux

Mots cl
es : Ejecteur ; Machine frigorifique a
ejecteur ; R245fa ; Banc d'essai ; Caract
eristiques de performance

* Corresponding author. New Energy Center, Department of Mechanical Engineering, National Taiwan University, Taipei 106, Taiwan. Tel.:þ886 2 23634790.

E-mail address:[email protected](K.O. Shestopalov). w w w . i i fi i r . o r g

Available online at

www.sciencedirect.com

ScienceDirect

journal homepage: www.e lsevie r.com/locate/ijrefrig

http://dx.doi.org/10.1016/j.ijrefrig.2015.02.004

1.

Introduction

The ejector refrigeration machine (ERM) offers several ad-vantages over the other heat-driven refrigeration cycles, including simplicity in design and operation, high reliability and low installation cost, which enable their wide application in the production of cooling (Petrenko et al., 2011).

In Part 1 of this paper (Shestopalov et al., 2015) an improved 1-D mathematical model for determination of the entrain-ment ratiou and optimal design of ejectors with cylindrical mixing chamber (CMC) and conical-cylindrical mixing cham-ber (CCMC) are proposed.

Based on theoretical comparative analysis of ejector and ejector refrigeration cycle performance characteristics for eight low-pressure refrigerants, the refrigerant R245fa was selected as the most suitable working fluid for general pur-pose applications in the present study.

This paper provides the main results of joint research and development carried out in the period from 2008 to 2012 at New Energy Center of the Mechanical Engineering Depart-ment, National Taiwan University, Taipei, Taiwan, in coop-eration with the Odessa State Academy of Refrigcoop-eration, Ukraine, in the area of ejector refrigerating technologies that was based on the Global Research Partnership Award (GRP Award), granted by the King Abdullah University of Science and Technology (KAUST).

The ejector with advanced construction and improved flow profile and the design of ejector test rig are described in this paper. The testing technique of the experimental investiga-tion of ERM with maximum cooling capacity of 12 kW oper-ating at design and off-design conditions is presented.

Performance characteristics of ERM are determined; the influence of the ejector geometry and operating conditions on

machine performance and characteristics is shown. The comparison of experimental and theoretical data for ERM operating at design and off-design working conditions is presented. The validation of the mathematical modeling presented in the Part 1 is done.

2.

Description of experimental setup

The main components of experimental ejector refrigeration cycle include an ejector, a generator, an evaporator, a condenser, a receiver-subcooler, an expansion valve and a feed pump. The arrangement of these components is given in simplified diagram of experimental ERM (Fig. 1). The process Nomenclature

1, 2, 3 number of nozzle (Fig. 6)

1, 2, 3 number of conical-cylindrical mixing chamber (Fig. 8)

A, B, C number of cylindrical mixing chamber (Fig. 7)

A area (mm2)

COP coefficient of performance

d diameter (mm)

ERM ejector refrigeration machine

h specific enthalpy (kJ kg1)

k thermal conductivity (W m1K1)

l length (mm)

_

m mass flow rate (kg s1)

P pressure (bar)

Q heat flow (kW)

q specific heat of evaporation (kJ kg1)

T temperature (C or K)

_

W power (kW)

Greek letters

b ejector area ratio

j angle u entrainment ratio y flow coefficient Subscripts actual actual b boiling c condenser cr critical e evaporator fp feed pump g generator mech mechanical max maximum n nozzle p primary s secondary sub subcooling suc suction t throat therm thermal

1, 2, 3, 4 cross-sections of the ejector (Tables 1, 2 and 4) 1, 2, 3…11 measuring points of temperature (Fig. 3)

Fig. 1e Diagram of experimental ejector refrigeration machine.

of a continuously operating ERM is characterized by points 1e7 inFig. 2, which is a diagram of an actual ejector refriger-ation cycle of this machine.

Usually the working fluid transported through the feed pump is close to the saturation conditions. Therefore a slight drop in the condensing pressure and pump inlet pressure may therefore cause cavitation, which in turn causes loss of pump work, decreasing in the coefficient of efficiency and steady operation of the system disorder. In order to increase the dependability and effectiveness of the system as a whole it is necessary to provide sufficient subcooling of suction liquid refrigerant.

In order to improve the system performance, special receiver-subcooler is added to the cycle as it is shown inFig. 1. The purpose of this component is to subcool the condensate prior to entering the feed pump and evaporator, thus increasing the operational reliability, specific cooling capacity qeand effectiveness of the ejector system.

To verify the theoretical analysis of the ejector geometry and performance characteristics of the ERM using refrigerant R245fa, an ejector test rig with a cooling capacity of 12 kW was designed and constructed. A schematic diagram and a photograph of the ejector test rig are shown inFigs. 3 and 4, respectively. The ejector test rig equipment includes the following nine major components: an experimental ejector, a generator, an evaporator, a condenser, a receiver-subcooler, a float regulating valve, a gear-type feed pump, a cooling tower and a control panel equipped with different measurement instrumentation. Locations of measurement points around the circuit are shown inFig. 3.

Fig. 5illustrates the structure of the ejectors with cylin-drical (a) and conical-cylincylin-drical (b) mixing chambers. A cone angle of the divergent part of the supersonic nozzles was J1¼ 6; a cone angle of the convergent part of the supersonic nozzles was J2¼ 30; the angle of the entering part of the conical-cylindrical mixing chambers was J3¼ 2 and of the diffusere J4¼ 8. The length of the mixing chamber lmchwas 6…10d3, and the length of the diffuser ldwas 10…12d3. Such

values of the cone sections angles and lengths of the ejector flow profile confirmed the high efficiency of the previously studied ejectors operated with refrigerants R141b, R142b, R236fa и R245fa (Petrenko, 1978; Huang et al., 1999; Eames et al., 2007, 2004).

A photograph of the experimental ejector assembly with two retrofract symmetrical suction inlet ducts and a suction manifold is shown inFig. 6. The assembly of the ejector con-sists of the following main components: a body, an axially movable supersonic nozzle, a mixing chamber made jointly with the diffuser, a frame and a mechanism to move the nozzle into optimal position with respect to the entrance of the mixing chamber. The ejector assembly is 510 mm in length, 60 mm in height and 130 mm in width.

The generator was designed with a cylindrical shape and had a glass level gauge for liquid level observation. The working fluid in the generator was heated by two 13 kW electric heaters those were separately controlled, as were the heat load, generating temperature and pressure. The evapo-rator was also designed in a cylinder shape with a glass level gauge for liquid level observation. Heat energy was directly transferred to the evaporator by two 6 kW electric heaters to simulate the evaporator cooling load. The electric heaters were also separately controlled, as were the cooling load, evaporating temperature and pressure. Electric energy inputs to the generator and evaporator were measured by electrical power meters. The generator and evaporator and all of their connecting lines were thoroughly thermally insulated. The condenser was a conventional shell-and-tube heat exchanger with a glass level gauge, cooled by water supplied from the cooling tower, with a rejected heat capacity of 52 kW. The condenser temperature and the ejector backpressure were automatically controlled by varying the water flow through the condenser. The receiver-subcooler was a specially designed shell-and-coil type vertical vessel cooled by water taken from the cooling tower. It was equipped with a level gauge and level transmitter for liquid level observation and control. The receiver-subcooler and all of its connecting lines were thermally insulated. A hydraulic gear-type pump driven by a three-phased variable speed electric motor was used as the generator feed pump.

The receiver-subcooler was equipped with level regulator (RLR) with electromagnetic level sensor. The signal from the electromagnetic level sensor was used for the automatic change of the motor speed of the feed pump by means of the inverter type regulator. Such a way the autocontrol of the output ca-pacity of the feed pump was realized. This ensured a reliable operation of the test rig and a steady maintenance of the generator operation as well as generating temperatures and pressures. Likewise the receiver-subcooler provided a positive suction head for the regular pump operation by subcooling of the liquid refrigerant and by geometrical static suction head.

In order to carry out experimental investigations in wide range of operating conditions Tg¼ 80e105C, Tc¼ 24e42C, Te¼ 4e20C and to obtain the values of COPtherm> 0.4, three supersonic nozzles N1, N2, N3 and three cylindrical mixing chambers A, B, C were designed and manufactured. Photo-graphs of the tested supersonic nozzles and cylindrical mixing chambers made jointly with the diffusers are shown inFig. 7 and Fig. 8, respectively. The specifications of the tested Fig. 2e Diagram of the actual ejector refrigeration cycle.

supersonic nozzles and cylindrical mixing chambers are listed inTables 1 and 2, respectively.

Table 3illustrates ejector specifications obtained by com-bination of tested three nozzles N1, N2 and N3 and tested three constant-area mixing chambers A, B and C.

Petrenko et al., 2011stated that the application of conical-cylindrical mixing chambers at the same operating conditions

allows for improvement up to 25e35% in u compared with

cylindrical mixing chambers. In Part 1 of the present series (Shestopalov et al., 2015) it has been shown theoretically that for different low-pressure refrigerants, this improvement is in the range from 0.7% to 23.6%. For refrigerant R245fa theoret-ical increase inu for ejectors with conical-cylindrical mixing

chambers is 19.6% for design conditions of Tg ¼ 95 C,

Tс¼ 32C and Te¼ 12C.

In order to compare the performance of ejectors for R245fa using cylindrical mixing chambers with ejectors using conical-cylindrical mixing chambers, three conical-cylindrical mixing chambers 1, 2 and 3 were designed and manufactured to carry out the experimental investigations at the same operating conditions as was mentioned before and to obtain the values of COPtherm> 0.6. These new mixing chambers were Fig. 3e Schematic diagram of the ejector test rig.

Fig. 4e Photograph of the ejector test rig.

Fig. 5e Structure of advanced ejector with cylindrical (a) and conical-cylindrical (b) mixing chambers.

made from cylindrical mixing chambers by boring the front part of the constant area section. Using this technology, all the dimensions of the new mixing chambers 1, 2 and 3 were the same compared to previous mixing chambers A, B and C. Only new conical section was created, which occupied part of the cylindrical section. Photograph of the manufactured conical-cylindrical mixing chambers made jointly with the diffusers

is shown in Fig. 9. Additional dimensions of new mixing

chambers are listed inTable 4.

3.

Testing technique

A standard procedure was used for operation of the test bench. Before each test run, refrigerant liquid levels within the generator, evaporator and receiver-subcooler vessels were set to predetermine values. After replacement of the nozzle or mixing chamber, a vacuum pump was used to evacuate the air from the suction line. For each test run, the generator heaters were switched on and set to the desired generating tempera-ture and generator heat load input. The flows of cooling water through the condenser and receiver-subcooler were adjusted by appropriate shut-off water valves and during the tests the cooling water flows were varied depending on the operating conditions. When the desired temperature in the generator was reached, the shut-off valves on primary flow line and secondary flow line were opened, and in the same time the feed pump and evaporator heaters were switched on. The pressure within the condenser (the back-pressure of the ejector) was controlled by automatic or manual adjusting the flow rate of cooling water through the condenser, and the pressures at the generator and evaporator were controlled by automatic or manual adjusting of power consumption of appropriate electric heaters.

Primary and secondary flow rates were measured by gear-type flow meters. The pressures of the primary and secondary flows, the back pressure at the condenser and the pressure after feed pump were measured using industrial direct-reading pressure gauges and pressure transmitters. Suction and condensing pressures were measured using high-precision Bourdon-tube gauges. Several K-type thermocou-ples were installed at appropriate places (T1-T11) for temper-ature measurement. RTD sensors were used to control the generating, evaporating and condensing temperatures. The

generator, evaporator and condenser were protected by pressure relief valves. The flow rates of the cooling water circulating through the condenser and receiver-subcooler were measured by glass flow meters. A control panel equip-ped with different instrumentation and other various stan-dard components of the refrigeration machine were also used in the construction of the ejector test rig. Specifications of the main measurement instrumentation of the ejector test rig are given inTable 5.

A PC-based monitoring and control system was developed in the present study for the ejector test rig. The data were sampled by a data acquisition system every 10 s. Pressures, temperatures, primary and secondary flow rates, electric power consumption and other required data were recorded and the results were calculated. This enabled the main per-formance to be determined in a steady state condition of system operation.

Fig. 6e Photograph of the ejector assembly.

Fig. 7e Photograph of the tested supersonic nozzles 1, 2 and 3.

Fig. 8e Photograph of the tested cylindrical mixing chambers A, B and C.

During experiments the value of subcooling of liquid refrigerant in suction pipeline before feed pump was 2e4_C, the superheating of entrained vapor in suction pipeline of the ejector was 5e7_C.

For each ejector geometry, an independent nozzle exit position test was carried out. The optimal nozzle position, which provides the maximum ejector performance, was used to test each ejector under design and off-design operating conditions. For each series of tests for design and off-design operating conditions, only one parameter of the system was varied whilst keeping the others constant in order to deter-mine its influence on the system performance. When the ejector cooling system was operating in steady-state condi-tions for about 30e40 min, the pressures, temperatures, flow rates, and other required data were recorded and results were calculated. For each test the critical condensing temperature also was established and determined.

The entrainment ratiou was determined from Eq.(1):

u ¼m_s _

mР (1)

Primary and secondary flow rates were determined in two independent ways: directly (from the readings of the flow meters), and indirectly (from energy balance of generator and evaporator) fromEqs. 2e3:

_ mP¼ _ Wg h1 h6¼ _ Wg qg (2)

Table 1e Specifications of the tested supersonic nozzles.

Number of the primary nozzle N1 N2 N3

Operating conditions Te¼ 8C, Tg¼ 90C Te¼ 12C, Tg¼ 95C Te¼ 16C, Tg¼ 100C Throat diameter dt, mm 4.515 4.212 3.902 Throat area At, mm2 16.0 13.93 11.95 Exit diameter d1, mm 7.8 7.11 6.412 Exit area A1, mm2 47.76 39.68 32.27 Area ratio A1/At 2.985 2.85 2.70

Diverging angle at nozzle exit j1, deg 6.0 6.0 6.0

Converging angle at nozzle enter j2, deg 30 30 30

Diverging part length ldiv, mm 31.48 28.33 24.71

Table 2e Specifications of the tested cylindrical mixing chambers.

Number of the cylindrical mixing chamber A B C

Constant area section diameter d3, mm 12.155 13.020 14.010

Constant section area A3, mm2 115.98 133.07 154.08

Length of constant area mixing chamber lmch, mm 90.62 98.64 106.65

Length of diffuser ld, mm 142.9 135.75 128.65

Diverging angle of diffuser j4, deg 8.0 8.0 8.0

Exit diameter of diffuser d4, mm 32.0 32.0 32.0

Exit area of diffuser A4, mm2 803.84 803.84 803.84

Table 3e Specifications of tested ejectors with

constant-area mixing chambers.

Nozzles and mixing chambers combinations A3/At

Nozzle N1e mixing chamber A (Ejector 1-A) 7.25 Nozzle N1e mixing chamber B (Ejector 1-B) 8.32 Nozzle N1e mixing chamber C (Ejector 1-C) 9.63 Nozzle N2e mixing chamber A (Ejector 2-A) 8.33 Nozzle N2e mixing chamber B (Ejector 2-B) 9.55 Nozzle N2e mixing chamber C (Ejector 2-C) 11.06 Nozzle N3e mixing chamber A (Ejector 3-A) 9.71 Nozzle N3e mixing chamber B (Ejector 3-B) 11.14

_ ms¼ _ We h2 h7¼ _ We qe: (3) Cooling capacity of ERM without respect to the heat gains and losses was defined according to the power consumption of evaporator heaters, i.e. Qe¼ _We, and also from Eq.(4):

Qe¼ _ms$ðh2 h7Þ ¼ _ms$qe: (4)

Generator heat load without respect to the heat loss was determined by power consumption of generator heaters, i.e. Qg¼ _Wg, and also from Eq.(5):

Qg¼ _mp$ðh1 h6Þ ¼ _mp$qg: (5)

COPthermwas calculated from Eq.(6): COPtherm¼ Qe Qg¼ _ We _ Wg : (6)

Actual power consumed by feed pump _Wactual_fp , was deter-mined from direct measurement of power meter, which was inbuilt in variable speed driver. The values of _Wactualfp from experiment were used to calculate COPmechfrom Eq.(7): COPmech¼

Qe _ Wactual_fp

(7) The material of isulation for the generator was high

effi-cient Flexible Rock Wool (Fibertex R120) with

k ¼ 0.036 W m1 _K1 _{and the thickness of 50 mm. The}

generating temperatures of refrigerant were 50e90_{C higher} than ambient temperature (depending from operating condi-tions). Therefore, average unwanted generator heat loss was assumed to be 5% of the electrical power input. The heat loss was calculated for the maximum generator heat load and maximum temperature difference between generating and ambient temperatures this respect to the thickness and thermal conductivity of the isolating material. Hence, the

generator heat load Qg from Eq. (5) was 5% lower than

measured power consumption of generator heaters _W. The evaporator was isolated with economical Flame

retarded NBR/PVC polymeric foam (K-FLEX ST) with

k¼ 0.033 W m1_K1_{with the thickness of 50 mm. The} evap-orating temperatures of refrigerant were 10e15_{C lower than} ambient temperature and sometimes even slightly higher. Thus, heat gains or losses in evaporator were negligible. Therefore, cooling capacity Qefrom Eq.(4)was the measured heat load of evaporator heaters.

During the tests, the two main errors are first of all from temperature, pressure, flow rate and heat load measurement accuracy, and secondlye from data logging and reading by the computer. A detailed error analysis indicates that the maximum uncertainty in the performance parameters is less than ±5.0%. Therefore, these errors are acceptable for the present study.

4.

Experimental results and discussion

4.1. Nozzles calibration

The first task of the experiments that used R245fa was to calibrate the supersonic nozzles under different generating temperatures and pressures (Fig. 10). In these tests, the rig was run with a closed suction line. The calibration tests for the

Table 5e Specifications of the measurement instrumentation of the ejector test rig.

Parameters Instruments Ranges Accuracy

Temperature Tg RTD sensor 0 ~ 200C ±0.12%

Temperatures Te, Tc RTD sensors 0 ~ 50C ±0.12%

Temperatures T1-T11 K-type thermocouples 200 ~ 320C ±1.5C

Pressures Pg, Pe, Pc, Pfp Industrial pressure gauges 1 ~ 30 bar ±1.5%

Pressure Psuc Bourdon-tube pressure gauge 1 ~ 1 bar ±0.25%

Pressure Pc Bourdon-tube pressure gauge 1 ~ 2 bar ±0.25%

Refrigerant flow rates Gear-type flow meters 0.1 ~ 6.0 l min1 ±3.0%

Electric energy input Electrical power meters 0 ~ 30 kW ±0.6%

Fig. 10e Theoretical and experimental mass flow rates of the primary nozzles 1, 2 and 3 against the generating temperature.

Table 4e Specification of conical-cylindrical mixing

chambers.

Number of the conical-cylindrical mixing chamber

1 2 3

Conical section entry diameter d2, mm 13.848 14.835 15.954

Area ratio b¼ A2/A3 1.3 1.3 1.3

Entry conical section angle j3, deg 2 2 2

Length of mixing chamber conical section lcs, mm

nozzles were repeated 4e6 times to ensure that no variation occurred with the data taken from the generator heaters.

The primary flow was determined from the energy balance of the generator and it was also measured using a flow meter. The theoretical mass flow rate through the supersonic nozzle can be expressed by (Volovyk, 2013):

_ mp¼ At$Pg g ffiffiffiffiffiffiffiffiffiffiffiffi 2$g gþ1$ Pg r1 q $  2 g þ 1 g g1 $y (8)

where g is heat capacity ratio, and ye stands for the flow coefficient.

The calculated values with flow coefficient y¼ 0.95 agree with the measured data perfectly.

4.2. Experimental determination of optimal position of

the nozzle

During further research for all ejectors with cylindrical mixing chambers (CMC) several experiments were carried out which allowed to determine optimal distance lnbetween exit cross-section of the nozzle and entrance cross-cross-section of the mix-ing chamber. The nozzle was moved in the line of mixmix-ing chamber and the efficiency of the system was determined.

Maximum entrainment ratiou ¼ uMAX_{corresponds to this}

optimal position of the nozzle for design operating conditions. Optimal values of ln, which were received from the ex-periments, and the values of the main geometrical parameter of A3/Atfor tested ejectors with CMC are shown inTable 6. All further researches, when the efficiency of the ejectors and ERM at design and off-design conditions was determined, were carried out at optimal values of the distance ln.

4.3. Experimental investigation of ejectors with

cylindrical mixing chambers

Once the supersonic nozzles were calibrated over a range of generating temperatures, the cooling machine was ready to conduct the full test for the designed ejector geometry over a range of design and off-design operating conditions.

The experimentation was carried out by varying the tem-peratures individually in the evaporator, condenser and generator. For each test the critical condensing temperature was established and determined.

Nine ejectors with cylindrical mixing chambers A, B and C were tested. The results of experimental investigations of ejectors with CMC operating at critical conditions are pre-sented inTable 7.

Some typical of the test results obtained for the ejector

cycle are shown in Fig. 11. These data were obtained for

ejector 2-B with the area ratio A3/At¼ 9.55 and optimal posi-tion of the nozzle ln¼ 20.6 mm in the front of the entry section of the cylindrical mixing chamber.

An ejector 2-B with cylindrical mixing chamber using R245fa was designed for operation at Tg¼ 95C and Te¼ 12C. Fig. 11demonstrates the effect of the condensing temperature Tcon the ejector 2-B and ERM performance for evaporating temperatures Tevarying from 8 to 16C and Tg¼ 95C. For each given Te, the entrainment ratiou, thermal coefficient of performance COPthermand cooling capacity Qeare indepen-dent of the ejector back pressure, i.e., the condensing tem-perature and pressure. However, when the condensing temperature Tcis higher than a certain value, known as the “critical condensing temperature” Tc*, thenu, COPthermand Qe will decrease suddenly and then fall to zero (Huang et al., 1985).

From obtained experimental results, it follows that the critical condensing temperature varies with generating and evaporating temperatures. This is therefore an important design criterion, which is determined by the highest temper-ature anticipated at the condenser heat sink. The optimum ejector and ERM performances can be obtained when the system operates at critical condensing temperatures. At a constant generating temperature, higher values ofu, COPtherm and Qecan be achieved when the evaporating temperature is allowed to rise. Therefore, for air conditioning applications, it is preferable to design an ejector for high evaporating tem-peratures. This then leads the ERM to operate at higher critical condensing temperatures as the condenser heat sink tem-perature increases.

Figs. 12 and 13illustrate the variations in theoretical trends and experimental critical values of A3/At,u*, COPtherm* and Qe* with Teand Tc*at Tg¼ 95C for ejectors 2-A, 2-B and 2-C.

It can be seen fromFig. 12that theoretical a¼ A3/Atfalls with increases in Tc*and decreases in Te. FromFig. 13it follows that the theoretical characteristicsu*_{, COP}

therm

* _{and Q}

e * _have

the same trend: they increase with increment of Te and

decrement of Tc*.

The theoretical results within experimental error are in good agreement with the corresponding experimental results. In practice, the actual operating temperatures Tc, Tgand Te vary with the surrounding conditions and usually are adjust-able (Huang et al., 1985). A theoretical method for the deter-mination of the optimal adjustment range subject to the requirements of the refrigeration user and concrete applica-tion condiapplica-tions is given byPetrenko and Volovyk (2007). This method enables one to select the combination of three

inde-pendent operating parameters Te, Tc and Tg to provide

maximum efficiency of the ERM at critical operating conditions.

The theoretical and experimental results were used to construct performance maps of all manufactured ejectors with cylindrical mixing chambers.

Fig. 14shows the theoretical (dotted lines) and measured (solid lines) variations inu*_{, COP}

therm

* _{and Q}

e

*_{with the critical} condensing temperatures over a range of evaporating and

Table 6e Experimental values of A3/Atand lnfor tested ejectors.

Ejector 1-А 1-B 1-C 2-A 2-B 2-C 3-A 3-B 3-C

A3/At 7.25 8.32 9.63 8.33 9.55 11.06 9.70 11.14 12.89

generating temperatures for ejector 2-B. From Fig. 14 for experimental data, it follows that at the given adjustment range of Te, which varied from 8 to 16C, in order to achieve the maximum performance of ERM with a variation of Tc*in the range of 29.5e39.0_{C, the temperature T}

gmust also be adjusted in the range of 88.0e102.0_{C. Under these conditions,} the characteristicsu*_{, COP}

therm

* _{and Q}

e

* _{will take on various} intermediate values in the following ranges:u*_{¼ 0.23e0.70,} COPtherm* ¼ 0.17e0.55 and Qe*¼ 3.3e7.8 kW.

Theoretical data show good coincidence with experimental results not only for the design points, but also for other off-design points. The maximum difference between the experi-mental results and the results of the proposed model falls within 5%.

The ERM performance maps provide useful assistance for the development of an automatic control system for the ERM

operating at off-design operating conditions. Obtained test results demonstrate that solar energy or low-grade heat can be used to drive efficient ERMs operating with R245fa for air conditioning and space cooling.

The patterns described above were also determined for all other tested ejectors (Table 3).

4.4. Experimental investigation of ejectors with

conical-cylindrical mixing chambers

Furthermore, nine ejectors with conical-cylindrical mixing chambers (CCMC) were also tested at the same operating conditions. The results of experimental investigations of ejectors with CCMC operating at critical conditions are pre-sented inTable 8.

Geometrical characteristics A3/At of tested ejectors with CCMC were the same as for ejectors with CMC, and for all ejectors with CCMC the value of b ¼ 1.3. Some typical test results obtained for the ejector 2e2 with area ratio A3/At¼ 9.55 and optimal position of the nozzle ln¼ 20.6 mm are shown in Fig. 15.

FromFig. 15it can be seen that at Tg¼ 88, 95 and 102C and Te ¼ 12C the behavior of u ¼ f(Tc), COPtherm ¼ f(Tc) and Qe¼ f(Tc) for ejector 2e2 with CCMC is the same as similar functions for ejector 2-B with CMC. For ejector 2e2 it can be found that the values ofu, COPthermand Qeare higher and Tc* are lower compared to ejector 2-B.

The comparison ofu*_{, COP} therm

* _{and Q}

e

* _{for ejectors with} nozzle 2 between cylindrical and conical-cylindrical mixing chambers at design operating conditions (seeFig. 16) shows that the application of the conical-cylindrical mixing cham-bers throughout considered range of critical condensing temperatures Tc* ¼ 31.3e35.6 C allows for the increasing of ejector and ERM efficiency up to 20% on the average.

Table 7e Experimental data for the nominal (critical) conditions of tested ejectors with CMC.

Ejector 1-А 1-B 1-C 2-A 2-B 2-C 3-A 3-B 3-C

Generating temperature Tg,C 90 90 90 95 95 95 100 100 100

Evaporating temperature Te,C 8 8 8 12 12 12 16 16 16

Critical condensing temperature Tc*,C 34.7 32.8 30.2 36.9 34.2 31.3 37.5 35.6 32.5

Critical entrainment ratiou* _{0.241 0.318 0.402 0.345 0.423 0.536 0.471 0.575 0.744}

Critical COPtherm* 0.186 0.243 0.309 0.265 0.323 0.411 0.358 0.440 0.570

Critical cooling capacity Qe*, kW 3.1 4.1 5.2 4.4 5.4 7.0 5.7 7.0 9.2

Fig. 11e Measured variation in u, COPthermand Qewith Tcat different Tefor ejector 2-B (with cylindrical mixing chamber).

Fig. 12e Variation in theoretical trends and experimental values of A3/Atwith Teand Tc*at Tg¼ 95C for different ejectors.

Fig. 17 shows the efficiency of ejector 2e2 in a form of performance maps, which were constructed from the theo-retical and experimental results to show the ejector perfor-mance characteristics, from which the design analysis of ejector refrigeration system could be carried out.

The efficiency comparison of ejector with cylindrical and conical-cylindrical mixing chambers is made. The results of this comparison are shown inFig. 18.

Average improvement in efficiency of ejectors with CCMC over ejectors with CMC is in the range from 23.5%: from 15% for ejector 3e1 compared to 3-A up to 54% for ejector 1e1 compared to 1-A. The critical condensing temperature Tc* de-creases 1.2e1.6_{C for ejectors with CCMC because their main} geometrical characteristic A3/At remained the same as for ejectors with CMC. The relative increasing in entrainment ratio corresponds to the higher condensing temperatures; this means that the application of CCMC is more preferable for high condensing temperatures.

The generating temperature for each compared ejectors was the same, resulting in similar flow core at the exit of the nozzle. Ejectors with conical-cylindrical mixing chambers had conical part of the mixing chamber at the entrance. This leads to the increase in flow area for entrained refrigerant vapor (increasing of entrainment ratio). But the diameter of the cy-lindrical part of the mixing chamber was the same as well as the other geometry. It means that the energy of the primary working stream was not enough to compress larger (compared to CMC) amount of entrained vapor up to the pressure corresponding to the critical condensing pressure

obtained for ejectors with CMC. That's why the critical

condensing temperature for ejectors with CCMC was lower. In order to equalize the critical condensing temperatures for such ejectors it is necessary to make smaller diameter of cy-lindrical part of mixing chamber of ejector with CCMC. But in the presented research the authors changed only one geometrical parameter.

Fig. 14e Measured and theoretical variation in u*_{, COP} therm

* _{and Q}

e *_{with T}

c* over a range of Teand Tgfor ejector 2-B (with cylindrical mixing chamber).

Table 8e Experimental data for the nominal (critical) conditions of tested ejectors with CCMC.

Ejector 1e1 1e2 1e3 2e1 2e2 2e3 3e1 3e2 3e3

Generating temperature Tg,C 90 90 90 95 95 95 100 100 100

Evaporating temperature Te,C 8 8 8 12 12 12 16 16 16

Critical condensing temperature Tc*,C 34.7 31.7 29.0 35.6 33.0 29.7 36.7 33.9 30.8

Critical entrainment ratiou* _{0.355 0.437 0.529 0.449 0.544 0.673 0.587 0.727 0.892}

Critical COPtherm* 0.270 0.334 0.407 0.340 0.413 0.516 0.447 0.556 0.689

Critical cooling capacity Qe*, kW 4.5 5.6 7.1 5.6 6.8 8.7 7.0 8.9 11.4

Fig. 13e e Variation in theoretical trends and experimental critical values of u*, COPtherm* and Qe*with Teand Tc*at Tg¼ 95C for different ejectors.

Fig. 15e Measured variation in u, COPthermand Qewith Tcat different Tefor ejector 2e2 (with conical-cylindrical mixing chamber).

Fig. 16e Comparison of u*_{, COP} therm

* _{and Q}

e

*_{for design working conditions at T}

e¼ 12C and Tg¼ 95C.

Fig. 17e Measured and theoretical variation in u*_{, COP} therm

* _{and Q}

e *_{with T}

c* over a range of Teand Tgfor ejector 2e2 (with conical-cylindrical mixing chamber).

Fig. 18e Measured variation in u*_{, COP} therm

* _{and Q}

e *_{with T}

5.

Conclusions

The testing technique for the experimental investigation of ERM with maximum cooling capacity of 12.0 kW operating at design and off-design conditions is presented. The described testing technique is proposed to carry out the assessment of the suggested ERM. The main components of the system are specially designed and the instrumentation generally is classic for refrigeration test rig.

This paper presents the results of a theoretical and experimental investigation of an ejector and ejector refriger-ation machine operating with refrigerant R245fa.

The effect of the ejector cycle operating conditions on the ejector and ERM performance characteristics is studied in this part. A comparison of the test results and the model predictions demonstrates that the experimental perfor-mances at the design operating conditions are about 10% higher than the theoretical values. The main reason for this is that the advanced experimental ejector has lower process irreversibility and energy loss than was assumed in 1-D model.

The obtained test results demonstrate that low-grade heat or solar energy can be used to drive efficient ERMs operating with R245fa and designed for air conditioning and space cooling.

Acknowledgments

This publication is based on the work supported by Award No.KUK-C1-014-12, made by King Abdullah University of Sci-ence and Technology (KAUST), Saudi Arabia.

r e f e r e n c e s

Eames, I.W., Ablwaifa, A.E., Petrenko, V.O., 2007. Results of an experimental study of an advanced jet-pump refrigerator operating with R245fa. Appl. Therm. Eng. 27, 2833e2840. Eames, I.W., Petrenko, V.O., Ablwaifa, A.E., 2004. Design and

experimental investigation of a jet pump refrigerator. In: Proc. 3rd International Conference on Heat Powered Cycles, HPC (Larnaca, Cyprus).

Huang, B.J., Jiang, C.B., Hu, F.L., 1985. Ejector performance characteristics and design analysis of jet refrigeration system. J. Eng. Gas Turb. Power 107, 792e802.

Huang, B.J., Chang, J.M., Wang, C.P., Petrenko, V.O., 1999. A 1-D analysis of ejector performance. Int. J. Refrigeration 22 (5), 354e364.

Petrenko, V.O., Volovyk, O.S., 2007. Analysis of performance characteristics of ejector refrigeration machine operating at design and off-design conditions. Refrig. Eng. Technol. 2 (106), 20e26.

Petrenko, V.O., Huang, B.J., Shestopalov, K.O., Ierin, V.O., Volovyk, O.S., 2011. An advanced solar-assisted cascade ejector cooling/CO2sub-critical mechanical compression

refrigeration system. In: ISES Solar World Congress 2011, 28 Auguste 2 September, Kassel, Germany.

Petrenko, V.O., 1978. Investigation of Ejector Cooling Machine Operating with Refrigerant R142b. Odessa Technological Institute of Refrigeration Industry, Ukraine. Ph.D. thesis. Shestopalov, K.O., Huang, B.J., Petrenko, V.O., Volovyk, O.S., 2015.

Investigation of an experimental ejector refrigeration machine operating with refrigerant R245fa at design and off-design working conditions. Part 1. Theoretical analysis. Int. J. Refrigeration.

Volovyk, O.S., 2013. Improvement of Characteristics and Performances of an Ejector Refrigeration Machine Operating with Low-boiling Working Fluids. Odessa National Academy of Food Technologies, Ukraine. Ph.D. thesis.