Accepted Manuscript

Characteristics and thickness effect of phase change material and

frost on heat transfer and thermal performance of conventional
refrigerator: Theoretical and experimental investigation

Mohand Berdja , Abdelkader Hamid , Osmann Sari

PII: S0140-7007(18)30387-6
DOI: https://doi.org/10.1016/j.ijrefrig.2018.10.003
Reference: JIJR 4124

To appear in: International Journal of Refrigeration

Received date: 27 May 2018

Revised date: 1 October 2018
Accepted date: 16 October 2018

Please cite this article as: Mohand Berdja , Abdelkader Hamid , Osmann Sari , Characteristics and
thickness effect of phase change material and frost on heat transfer and thermal performance of con-
ventional refrigerator: Theoretical and experimental investigation, International Journal of Refrigeration
(2018), doi: https://doi.org/10.1016/j.ijrefrig.2018.10.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service
to our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and
all legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Highlights
 PCM selection and the mechanism that permits extension of refrigerator downtime.
 Effect of PCMs on operating conditions and performance within refrigerator.
 Effect of PCM and frost thickness on heat transfer inside refrigerator.
 Frost affects heat transfer more significantly than PCM during the off cycle.

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Characteristics and thickness effect of phase change material and frost on heat transfer
and thermal performance of conventional refrigerator: Theoretical and experimental
investigation

Mohand Berdjaa,b,*, Abdelkader Hamida, Osmann Saric

a
Saad Dahleb University, 9000,Blida, Algeria
b
Unité de Développement des Equipements Solaires, UDES, Centre de Développement des Energies
Renouvelables, CDER, 42004, Tipaza, Algeria
c
Industrial Technologies Department, University of Applied Sciences Western Switzerland, CH 1400, Yverdon-
les-Bains, Switzerland

T
IP
ABSTRACT

CR
In order to design more energy-efficient refrigeration systems through means that have no
impact on the environment, the use of phase change materials (PCMs) coupled to the
refrigeration systems could be one of the most economically attractive solutions. Among the

US
many possible arrangements of the PCM within a refrigerator, PCM is often on the surface of
the evaporator, the location where frost forms. However, beyond a certain thickness, frost
would cause an excessive increase in the refrigerator air temperature and a reduction in the
AN
energy performance of the refrigerator. The aim of the present study is to clarify the nature
and characteristics of frost compared to water as phase change material (PCM), along with the
mechanisms with which they affect energy performance and heat transfer within the
refrigerator. This is based on a theoretical and experimental investigations conducted on a
M

conventional refrigerator equipped with a flat plate evaporator.

ED

Keywords: phase change material, frost, material characteristics, energy performance, heat
transfer, thermal storage.

*Corresponding author:
PT

E-mail: [email protected], [email protected] (M. Berdja)

CE

1. Introduction

The development of refrigeration and air conditioning is one of the most important factors
AC

influencing electricity peak consumption during extended periods of hot weather when power
generation and transmission systems often approach their capacity limits. Excessive use of
such equipment comes from the fact that the price of a kilowatt hour is relatively low in
Algeria, and that refrigeration equipment costs have become affordable.

Thus, the increase in energy consumption causes voluntary load shedding of the user's supply
in order to restore balance between their production and consumption network. This problem
prompts the government to acquire new and often expensive power plants, representing an
uneconomic investment. The use of a latent cold storage system is associated with more
efficient refrigeration systems and could become the solution to this problem in terms of
 ACCEPTED MANUSCRIPT

energy savings, depletion of fossil resources, and environmental impact [1, 2]. Many studies
were conducted on phase change materials for cold thermal storage, mostly focusing on the
integration of PCMs in buildings' structures, solar power systems, PCM properties
enhancements, air-conditioning and refrigeration equipment and many other applications.
Several interesting and significant studies and reviews that were carried out on the use of
PCMs and other thermal storage equipment can be found in literature [3–7].

An interesting study performed on the integration of PCMs in buildings has been presented by
Stritih et al. [8]. Authors conducted an experimental and numerical analysis of a phase change
material (PCM) integrated into ceiling panels and doors for cooling a building, using the free-

T
cooling principle. Results from measurements under real conditions show that PCMs reduce
the temperature fluctuations, bringing more thermal comfort, and reducing the need for

IP
additional cooling and air-conditioning in the building. Also, the authors concluded that
proper selection of the PCM is one of the main complications with the free-cooling principle,

CR
pertaining to obtainment of a complete solidification and melting process of the PCM.
Following this, Tzivanidis et al. [9] conducted a parametric analysis of a ceiling cooling
system coupled with a PCM. Authors integrated cooling pipes into the PCM layer, instead of

US
them being embedded in the ceiling slab as per usual for ceiling cooling systems. Based on
cold water flowing during the night, this PCM cooling system offers some advantages, such as
a high cold storage capacity, a night-time operation at a low price, a shifting daily peak load,
AN
and an elimination of leaking from the ceiling. The analysis concerned the effect of the
parameters such as pipe spacing, PCM layer thickness, and PCM properties on the system.
Authors concluded that a PCM-based system provides a higher level of thermal comfort as the
M

indoor air temperature fluctuation is lower than that of a concrete-based system. Also, the
performance of a ceiling cooling system is improved by increasing the difference between the
inlet temperatures of the heat transfer water and the PCM's phase change temperature, with a
ED

recommended difference value of 6°C. Kousksou et al. [10] developed a numerical code to
study the thermal behavior of building material containing microencapsulated PCM. The
study investigated the temperature regulation and heating/cooling rate effects resulting from
PT

the incorporation of microencapsulated phase change materials in cement based mortar. The
analyze of different parameters such as the heating/cooling rate and the mass fraction of the
CE

PCM on the thermal behavior of the PCM allowed to provide a tool to select the adequate
PCM in a building materials, for a maximal thermal storage rate.

The use of PCMs as means of thermal storage for solar power systems has attracted a lot of
AC

interest in recent years, because of fluctuations in the solar energy input. Several studies in
this topic have been mostly focused on heating and air conditioning for building. One
reference is the work of Esen and Ayhan [11] who studied theoretically the performance of
solar assisted cylindrical thermal energy storage for domestic heating system. A numerical
model has been developed for simulating the transient behavior of phase change thermal
energy storage in short term. The storage system consists of cylindrical tubes filled up with a
PCM and a heat transfer fluid (HTF) flowing parallel to the tubes. The enthalpic method has
been used to solve the phase change problem, as well as the Gauss-Seidel iteration method to
solve the resulting set of equations. Based on their numerical results, the authors noticed that a
 ACCEPTED MANUSCRIPT

high temperature and mass flow rate of the HTF doesn’t affect substantially the PCM melting
time. In order to increase the thermal energy storage performance, the authors suggested using
a smaller cylinder size, selecting a larger material thermal conductivity of the cylinder, and
increasing the effective thermal conductivity through means of fins or matrix structures to be
integrated into PCM. Following this, Esen et al. [12] simulated the transient behavior of two
different thermal energy storage configurations with four types of PCMs, in order to optimize
its performance. The first configuration consists of cylinders filled up with a PCM and a heat
transfer fluid (HTF) flowing parallel to it. The second one is a storage tank filled up with
PCM and containing cylindrical tubes where a HTF flowing inside. A parametric study has
also been conducted to define optimal geometric design of the thermal energy storage. The

T
authors noticed that the determination of PCM melting time is the most important criteria for

IP
selecting the PCM and for designing the storage tank. Melting time should be total so that the
PCM can store the maximum of thermal energy equivalent to the sunshine period. This

CR
parameter is mostly affected by the type of PCM and configuration. It has also been suggested
that the second configuration is more suitable for solar application, as it presents a shorter
melting time which is correlated to the thickness of the PCM mass. Later, Esen [13]

US
performed experiments to validate the numerical model that has been conducted for the
thermal energy storage of the first configuration. The experimental setup consists of 30m2 flat
plate solar collectors coupled to thermal energy storage containing 1090 Kg of encapsulated
AN
PCM. The material used is CaCl2 .6H2O salt. Numerical and experimental results have been
compared and have shown a reasonable agreement. The authors suggested reducing the
thickness of the tube walls if lots of tubes with small radii are used in order to reduce the
sensible energy stored in the tube material, and to promote latent heat thermal storage in the
M

PCM. They also proposed practical recommendations for a proper operating of such storage
systems.
ED

PCMs should possess desirable thermo-physical, kinetic, economical, and chemical properties
for an adequate thermal storage system. Although a large substances that have been defined as
PCMs, majority of PCMs do not satisfy the criteria required. Beside its advantages, PCMs
PT

show some undesirable properties such as a low thermal conductivity, supercooling, corrosion
and incongruent melting. The study of PCM’s properties has gained extensive attention and is
a matter of several investigations, which concerns mostly the use of additives to existing
CE

PCMs and the combination of different types of PCMs. Cantor [14] is one of the first to study
the PCM properties during melting and solidification phases. The author conducted
experimental investigations on salt hydrate phase change material in order to define its
AC

thermal behavior during the incongruent melting and solidification phases. This property
leads to differences between the enthalpy of fusion and enthalpy of solidification. Author
undertook tests by means of differential scanning calorimetry (DSC), in order to determine
these differences for several salt hydrates. Results showed that this disparency between
melting and freezing behavior decreases its potential of thermal energy storage medium.
Authors found that the enthalpy of solidification at or near the peritectic temperature is never
more than 60% of the enthalpy of fusion for the ‘Na2SO4–10H2O’ hydrates salt. Following
this, Wu et al. [15] investigated tests on Al2O3–H2O nanofluids, in order to evaluate its
potential as a new phase change material for cold thermal energy storage systems. Several
samples with different mass fractions of nanofluids were prepared for a cooling test, and the
 ACCEPTED MANUSCRIPT

results showed that the ideal fraction of 0.2 wt% Al2O3 reduces the total freezing time of the
new PCM by 20.5% and the supercooling degree by 70.9%, while the thermal conductivity is
increased by 10.5%. Sciacovelli et al. [16] conducted an analysis of the melting process in a
single vertical shell-and-tube thermal energy storage system, using a computational fluid-
dynamic (CFD) model based on the enthalpy method to deal with the phase change
phenomenon. The temperature evolution and melting interface was analyzed, and the fluid
flow was fully resolved in the liquid phase in order to elucidate the role of natural convection
during a liquid phase change. Thermal conductivity in PCM is enhanced by adding highly
conductive nano-particles, leading to a reduction in the melting time by 15% with a particle
volume fraction of 4%. Authors noted that natural convection and inlet heat transfer fluid

T
conditions greatly affect the thermal behavior of the thermal storage system, and represent an

IP
additional difficulty for the determination of parameters such as the storing/releasing rate and
the melting/solidification time that are needed to design latent heat thermal storage systems.

CR
This problem has been solved by Li et al. [17] who investigated the encapsulation of a PCM
in a porous structure in a honeycomb form, and packed it into a stainless steel can to suppress
the natural convection of the melted PCM. Experimental results showed that natural

ratio of 70% to 30% in weight.

US
convection is suppressed when the PCM and the SiC-honeycomb are together with a mixing

PCMs have received considerable attention for their heat transfer enhancement in air-
AN
conditioning and refrigeration equipment due to their advantage of high latent heat. An
interesting study on the integration of cold latent heat thermal energy storage to air-
conditioning application has been presented by Ismail et al. [18]. The authors conducted a
M

numerical and experimental study for the component of ice bank for identifying the effect of
geometrical and operational parameters on the performance of such equipment. Numerical
solution has been realized using the finite difference method and a fixed grid scheme, and the
ED

results have been compared with experiment, analytical solutions and an approximate model
based upon the integral heat balance method. Complete energy analysis showed that the cold
plate temperature and the gap between the cold plates are the parameters that have the most
PT

influence on the stored thermal energy and on the time for complete solidification, while the
initial temperature of the PCM only slightly affects the energy storage process. Following
CE

this, Ismail et al. [19] performed a numerical model for simulating the solidification of a
phase change material around a cold cylinder crossed by an HTF. The mechanism of heat
transfer in the liquid PCM during solidification process is considered to be pure conduction,
AC

and the discretization method to be used in the solution of the heat transfer problem is the
finite volume method. The moving grid method utilizing the interface immobilization
technique has been used to solve the phase change problem. The numerical results have been
compared with available numerical and experimental solutions from the literature. The
comparison takes into account some parameters such as interface position, ratio of solidified
thickness, time for complete solidification and the Nusselt number. The authors examined two
discretization schemes (power law and central difference scheme) for tracking the interface
position, and they recommend using the central difference scheme because of the low
velocities in the pseudo-convective terms. They also concluded that under these conditions,
the immobilization technique gives same results as the enthalpic formulation in solving the
 ACCEPTED MANUSCRIPT

phase change problem. Interesting studies have been reported on the integration of PCMs
inside refrigeration systems where they are put into contact with the evaporator. Mention can
be made of the work of Wang et al. [20–22] who performed an experimental study on a
conventional refrigerator with PCMs mounted throughout its different parts. Results indicated
the advantages of each PCM's location on the energy performance of the refrigeration system,
and obtained a COP improvement of 4% and a reduction of the refrigerant's sub-cooling of
8%. Marques et al. [23] conducted a numerical study on a vapor-compression refrigerator
with a thermal storage device, and concluded that the integration of a PCM in the refrigerator
allows to extend the compressor shutdown cycle of 4.8h, while Alzuwaid et al. [24]
introduced a water-based PCM in a refrigerated cabinet and noted a 5% improvement in the

T
coefficient of performance, with a better control and stability of the air temperature inside the

IP
refrigerated cabinet. On the other hand, Gin et al. [25] analyzed the performance of a freezer
during defrosting and door opening cycles after introducing panels filled with PCM and

CR
positioned against the inside walls of a freezer. Authors demonstrated that the PCM helped to
limit the temperature increase and reduce the energy consumption during these cycles,
without justifying the quantity of PCM to be introduced into the refrigerator. An interesting

US
study on improving thermal performance of a domestic refrigerator was presented by Azzouz
et al. [26, 27]. The author studied the effect of adding a PCM slab between the insulation wall
and the evaporator, and showed that the PCM enhances heat transfer at the evaporator's
AN
surface, raises evaporation temperature, and increases COP of the refrigeration system.

However, a recurring problem in refrigeration systems is frost formation at the surface of the
evaporator, which has the potential to seriously affect the performance of such systems. This
M

solid ice layer forms as humid air comes into contact with the cold surface of the evaporator.
Frost is caused primarily by moisture in the ambient air entering the refrigerator, or partially
by the moisture released from the food stored in the refrigerator [28]. In fact, the temperature
ED

of the surface of the evaporator is lower than the dew point of the air and the solidification
temperature of the condensate. As a result, the moisture in contact with the evaporator surface
condenses, solidifies, and forms the frost layer. In order to determine the water mass transport
PT

that comes from the humid exterior ambient air and the humidity from fresh food in a
domestic refrigerator, Stein et al. [29] investigated the evaporation rates from the water's open
source inside a refrigerator's cabinet and the infiltration of moisture through the cabinet door's
CE

gaskets. Temperature, humidity, and power consumption were measured in order to show the
effect of the water surface area and the cabinet door gaskets on the overall humidity loading
AC

on a refrigerator. Experimental results indicated that the water surface area inside the
refrigerated cabinet significantly contributes to the cabinet's humidity level and the
evaporator's frost accumulation, and it also generates an additional annual consumption of
energy of up to 15–20% relative to the ambient humidity. Indeed, a thick layer of frost
reduces the heat transfer on the surface of the flat evaporator and affects the cooling of the air
surrounding the evaporator as shown by Cox-Smith and Carrington [30]. The authors
examined the performance of a domestic heat pump with a focus on a frosting system analysis
on the operation of the heat pump. They noted that below an ambient temperature of 7°C, the
phase change loading of the frost in contact with the evaporator impacts its effectiveness and
reduces the seasonal COP by up to 7%. They also noted that the effect of frost and its
 ACCEPTED MANUSCRIPT

removal would be to decrease the overall electric input by approximately 5%. In the case of a
finned ventilated evaporator, the area where air passes between the evaporator's fins starts to
become smaller, causing a larger pressure drop, which produces an additional energy
consumption of the refrigerator's compressor. Sengupta et al. [31] investigated the influence
of some parameters regarding frost formation on cylinders where humid air is forced over the
cylindrical coil maintained at subfreezing temperatures using an external refrigerator. In their
work, the authors presented a detailed description of the frost formation process, and
developed empirical formulations for the time dependency of both the frost thickness and the
heat transfer coefficient for a cylinder in a humid air cross flow. Ismail et al. [32] developed a
numerical model for evaluating the properties of frost formation around a single vertical

T
cylinder crossed by a laminar wet air. The frost properties taken into account in the model are

IP
the thickness, the density and the surface temperature distributions. The first step of the model
defines the momentum equation in terms of the vorticity and stream function, for determining

CR
the flow field solution. The second step defines the temperature and humidity fields from the
resolution of the temperature and the phase continuity equations for different Reynolds
number. The third step consists on modeling and solving the frost formation problem, by

US
using the flow, temperature and humidity fields already calculated. Numerical results have
been compared to experimental data form literature, and a good agreement has been found for
experimental time range of 600 sec. The authors noticed that frost thickness increases with the
AN
increase of time and humidity. Also, density, thickness and thermal conductivity are the
parameters that affect the most the temperature surfaces of the frost. Later, Ismail et al. [33]
developed a one-dimension transient model for evaluating frost formation on a flat cold
surface crossed by a wet air flow. The model has also been extended for the case of a wet
M

airflow that crosses two parallel cold plates forming a frost layer. The model uses the local
volume averaging technique to predict parameters such as frost temperature, density and frost
ED

thickness. The frost formation is assumed to have three periods; one-dimensional crystal
growth, three-dimensional crystal growth and the quasi-static growth. Authors compared the
predicted frost thickness as a function of time with the experimental data in literature, and
PT

noticed a difference of less than 10%. The difference is more important during the first stage
of frost formation, which is due to the initial conditions imposed in the numerical model and
to the precision of experimental measurements during this stage.
CE

Despite several studies on the use of PCMs in refrigeration systems, the problem of frost
formation and its effect on such systems in presence of PCMs has not been examined before.
AC

To the authors' knowledge, this aspect has only been studied in the case of refrigerators
equipped with evaporators alone, or in case of cold surfaces crossed by a wet air. The present
study aims to define the effect of a PCM and frost on the energy performance and the heat
transfer inside a conventional refrigerator equipped with a flat evaporator. Characteristics of
frost are compared to those of water as phase change material (PCM), along with the
mechanisms with which they affect energy performance and heat transfer within the
refrigerator. The first part of the paper deals with the thermal analysis conducted on a
refrigerator coupled to a PCM. The analysis, based on experimental measurements and a
mathematical model, aims to define the effect of PCM on the compressor’s downtime, energy
performance and heat transfer within the refrigerator. The second part concerns a comparative
 ACCEPTED MANUSCRIPT

study that seeks to define, under the chosen cooling conditions and the nature of the PCM and
frost, the effect of the thickness of water-PCM and frost on the evaporator’s thermal
performance and the heat transfer within the refrigerator. The comparison between the two
materials is justified because both are on the surface of the evaporator.

2. Thermal analysis of a refrigerator coupled to a PCM

A thermal analysis is conducted on a refrigerator coupled to a PCM, in order to determine its

thermal behavior. The analysis comprises two parts:

T
- Experimental measurements carried out on the refrigerator for the determination of

IP
temperature profiles (refrigerating cycle, air, and PCM), electrical consumption, and COP.

CR
- An analytical calculation for the determination of the different heat transfer coefficients
inside the refrigerator in the presence of PCM.

2.1. Experimental setup

US
The experimental study consists of a series of temperature and power consumption
measurements conducted on a conventional 400L refrigerator, which is equipped with a
AN
compressor using HFC-R134a as its refrigerant (Fig. 1). A 10 mm PCM layer is uniformly
mounted on the evaporator's surface, so that one side of the PCM is brought into direct
contact with the evaporator's surface, and the other side is in contact with the room's air.
M
ED
PT
CE
AC

Fig. 1. Thermocouples and PCM location in the refrigerator compartment

Temperatures are measured at three (3) equidistant levels in the refrigeration compartment,
and at one (1) level in the PCM. For each level of the refrigerator, four (4) thermocouples are
 ACCEPTED MANUSCRIPT

mounted. The average temperature of each level is defined as the arithmetic mean of the four
(4) measured temperatures. Also, five (5) thermocouples are placed at each component of the
refrigerator's cycle in order to follow the refrigerant's temperature evolution. Temperature
acquisition is conducted using a Hydra data logger and K type thermocouples. Characteristics
of the refrigerator and the PCM are summarized in the following table:

Table 1
Refrigerator and PCM characteristics
Refrigerator

T
Useful volume 400L

IP
Dimensions 1.45 *0.52 *0.52 m

Insulation Polyurethane 0.025 m

CR
Refrigerant HFC-R 134a

PCM

Phase change temperature

Dimensions
US -11°C

48*40*10 cm
AN
Thickness 10 mm

Total mass 0.58Kg

M

2.2. Calculation model

ED

The presence of a PCM layer at the evaporator's surface permits an extension of the
compressor's off time. Indeed, the cold stored into the PCM during the compressor's on cycle
PT

is released to the evaporator's surface when the compressor is stopped, and keeps the
thermostat bulb at a low temperature, allowing the compressor to be idle for a longer time.
CE

However, thicker layers of the PCM could allow for additional thermal resistance that reduces
the heat transferred from the chamber air to the evaporator's refrigerant. Thus, the
AC

introduction of a thicker layer of PCM would markedly prolong the refrigerator's downtime
without being able to maintain an acceptable temperature of the air inside the refrigerator. A
calculation model has been set in order to quantify the effect of the PCM's thickness on the
refrigerator's thermal behavior. The model is based on two (2) parameters, defined as:

- The overall heat transfer that takes place between the refrigerant flowing through an
evaporator and the air within a refrigerator during the compressor's on/off cycles, in the cases
where the evaporator's surface is in contact with a PCM ( H PCM ), and for when an evaporator
is without a PCM ( H ).
 ACCEPTED MANUSCRIPT

- The overall heat transfer that takes place between the refrigerant flowing through the
evaporator and the PCM during solidification and melting phases ( H PCM _ evap ).

In order to avoid a fine modeling of the problem, we describe the heat transfer by simple
correlations that give the global exchange coefficient in the different mediums of the
computational layout (air, PCM, evaporator, wall, and refrigerant). The heat transfer
coefficients are calculated for a PCM thickness ( X f ) ranging from 5 to 40 mm as illustrated
in Fig. 2. A detailed description of the heat transfer phenomenon is not the aim of this work,
and some assumptions are presumed in order to simplify the heat transfer problem. The

T
assumptions are as follows:

IP
- The heat transfer is assumed to be one-dimensional.

CR
- The temperature of the air inside the refrigerator is assumed to be at a constant of 5°C.

- The surface temperature of the evaporator is assumed to be constant during the on and off
cycles of the compressor. It is assumed to be -19°C during the operating cycle and -3°C

US
during the inactive cycle, respectively. These values are calculated based on the averages
measured from the evaporator's surface temperatures. The surface temperature varies between
-3°C and -19°C for the thermostat setting position (3) as shown in Fig. 4.
AN
- The phase change phenomenon is complete for the melting and solidification processes, and
it is assumed that sensible heat is negligible compared to the latent heat of
M

melting/solidification.

- The refrigerator operating mode is a succession of stopping and operating cycles, and is
ED

considered as permanent.

- The thermal properties of the PCM are assumed to be the same during the solidification and
PT

fusion processes. They are estimated on the basis of an average between the liquid and solid
phases.
CE

- The convection in liquid PCM during the solidification process is negligible. The heat
transfer mode in both the liquid and the solid PCM is assumed to be through conduction [34].

- The heat transfer in PCM during the melting process is dominated by natural convection.
AC

The conduction is assumed to be negligible [35].

 ACCEPTED MANUSCRIPT

T
Fig. 2. Calculated convective heat transfer coefficients within the refrigerator.

IP
CR
2.3. PCM's phase change temperature and PCM arrangement within the refrigerator

The definition of the phase change temperature of the PCM to be integrated into the

US
refrigerator depends on the PCM's position in the refrigerator. It may be in direct contact, or
distanced from the evaporator, depending on the configuration of the evaporator and the
presence of a fan inside the refrigerator.
AN
The flat plate evaporator type is usually mounted without a fan, necessitating the PCM to be
placed in direct contact with the evaporator's surface (see Fig. 3). Usually, the PCM is
M

distanced from a ventilated finned evaporator. Unlike in a ventilated finned evaporator, the
chosen phase change temperature of a PCM placed directly against an evaporator's surface
must be close to the evaporator's surface temperature so that the PCM can fully solidify
ED

during the compressor's on cycle.

PT
CE
AC

Fig. 3. Design of a flat plate evaporator coupled to a PCM.

In a flat plate evaporator without forced ventilation, the heat transfer of the air inside the
refrigerator is mainly governed by natural convection. A low heat transfer rate within the
 ACCEPTED MANUSCRIPT

refrigerator allows the PCM to be in direct contact with the evaporator's surface. Thus, the
phase change temperature should be lower than the air temperature of the chamber, and
approach the plate evaporator's temperature, so that the PCM can solidify during the
compressor's on cycle, and melt again during the compressor's off cycle. During the off cycle
of the compressor, the stored cold is released partly to the air surrounding the evaporator, and
partly to the thermostat bulb mounted on the evaporator's surface, leading to the extension of
the compressor's downtime.

The determination of the phase change temperature of the PCM for a plate evaporator
depends on the value of the evaporator's surface temperature, which allows the phase change

T
phenomenon to occur (see Fig. 4). In the case of a ventilated finned evaporator, the PCM is
distanced from the evaporator and is only in contact with the air in the chamber. Thus, the

IP
phase change temperature of the PCM depends on the profile of the air inside the refrigerator.

CR
For the temperature measurements, the Kapton® sheathed thermocouple model with naked
junction has been used. The thermocouple is of type K, and its dimensions are 1.5 m in length,
and 1.3 mm in diameter. The measuring range is -250 to +400°C. Time lag of the

US
thermocouple response is 0.5 sec. The accuracy of the measurement is 10-4°C. Increment of
time is 5 sec.
AN
-2

-4
M

-6 Evaporator
PCM
-8
Temperature (°C)

ED

-10

-12
PT

-14
CE

-16

-18
AC

-20
0 20 40 60 80 100 120
Time (min)

Fig. 4. Measured phase change and evaporation temperatures for a plate evaporator.
 ACCEPTED MANUSCRIPT

2.4. Effect of PCM in extending a compressor’s downtime

The control of the desired air temperature into a conventional refrigeration system is usually
based on the type of thermostat, with the choice of one or the other type depending mainly on
the size in volume of the refrigeration system.

For refrigerators with large storage volumes, such as cold rooms, thermal gains in the
refrigerator come mainly through the isolated walls. The opening/closing rate of the
refrigerator door is generally low compared to small volume domestic refrigerators, and the
stored products can last several months inside such refrigerators. For this type of application,

T
an ambient thermostat to control the room's temperature is recommended. Thus, a precise
control of the room temperature is enough to ensure long-term storage of perishable stored

IP
products. A fan is also necessary for this type of refrigerator in order to homogenize the air
temperature inside and to avoid air stratification.

CR
Unlike cold rooms, small and medium-sized domestic refrigerators have a high
opening/closing rate of the refrigerator door, and product storage is usually short-lived. The

US
requirements in terms of maintaining and controlling the temperature inside such refrigerators
are less than those of the long-term storage refrigerators. In this type of refrigeration system, it
is necessary to use a surface thermostat, in which the thermostat bulb is in direct contact with
AN
the evaporator's surface. For this type of thermostat, the control of the compressor's operation
is based on the surface evaporator's temperature, and not on the air temperature, such as for an
ambient thermostat. This type of thermal control prevents excessive compressor starts with
M

each manipulation of the refrigerator door, which generates an electric overconsumption by

the compressor.
ED

Compressor's on cycle: During the operation time of the compressor, the cold refrigerant
interacts thermally with the PCM through its surface, and leads to a solidification of the PCM
and the cold's storage. A direct contact of the PCM with the evaporator's surface increases the
PT

heat transfer at the evaporator's surface. Improved heat transfer increases the evaporator's
efficiency and permits easier evaporation of the refrigerant, and consequently generates an
increase in the COP of the refrigerating machine.
CE

Compressor's off cycle: While the compressor is off, melting of solidified PCM permits the
release of the stored 'cold' both to the air in the chamber, and through the evaporator's surface.
AC

The operation of the refrigerator depends exclusively on the surface thermostat, which has a
bulb that is in direct contact with the surface of the evaporator. The cold is released partly to
the air surrounding the evaporator, and partly to thermostat bulb mounted on the evaporator's
surface, resulting in an extension of the compressor's downtime, and a reduction in the
number of daily on cycles and daily electricity consumption of the refrigerator. Heat transfer
improvement by the PCM on the evaporator's surface is of capital importance in extending the
downtime of the refrigerator's compressor.
 ACCEPTED MANUSCRIPT

2.5. PCM effect on the heat transfer inside the refrigerator

During the melting process (compressor’s off cycle), the heat transfer in the PCM is
dominated by natural convection, and the conduction is assumed to be negligible [35]. The
estimation of convective heat transfer coefficient hPCM of the phase change material during
the melting process (compressor off-cycle) is calculated from approximation proposed by Lim
and Bejan [35], which is valid for Pr >1 in rectangular enclosures and based on the Rayleigh
number RaPCM . The evaporation temperature is assumed to be constant and equal to -3°C.
This value represents the average of the measured temperature of the evaporator surface
during the compressor off cycle.

T
IP
K PCM NuPCM
hPCM  (1)
Xf

CR
NuPCM  0.35RaPCM
0.25
(2)

g  (Tevap  TPCM ) X 3f
RaPCM 
aPCM  PCM
US (3)
AN
The estimation of the PCM convective heat transfer coefficient hPCM during the solidification
process (compressor’s on cycle) is based on the conductivity and the thickness of the PCM,
M

and only conduction is taken into account [34]. This is expressed by Equation (4). The
evaporation temperature is assumed to be constant and equal to -19°C. This value represents
the average of the measured temperature of the evaporator surface during the compressor on
ED

cycle. The thermal properties of the PCM are assumed to be constant, and are calculated on
the basis of an average between the liquid and solid phases.
PT

K PCM
hPCM  (4)
Xf
CE

Figure 5 illustrates the evolution of convective heat transfer coefficients within PCM hPCM
vis-a-vis the variation of the PCM thickness X f in the range of 5-40 mm, for both melting and
AC

solidification processes. Under same operating conditions for both solidification and melting
processes, the variation of hPCM is only depending on the variation of X f . hPCM is defined for
the cases of melting by equations 1-3, and for solidification processes by equation 4.

It is noted that heat transfer coefficient hPCM is more significant during the melting process,
and its values are slightly more sensitive to the PCM's thickness when compared to the
solidification process. This is due to the fact that during solidification, heat transfer is
governed by conduction, while heat transfer is globally governed by natural convection during
the melting process. The decreasing curve of hPCM for the melting process is more pronounced
 ACCEPTED MANUSCRIPT

compared to the case of solidification. An increase in PCM thickness intensifies convection

heat transfer into the PCM, however, this increase is offset by the increase of PCM thermal
resistance with the increase in its thickness.

500

450
solidification process
400 melting process

350

T
hpcm (W/m2.K)

300

IP
250

CR
200

150

100

50 US
AN
0
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
PCM thickness Xf (m)

Fig. 5. Heat transfer coefficients through PCM (hPCM). Solidification/melting phases.

M

In order to highlight the effect of PCM thickness on heat transfer inside the refrigerator, we
ED

define :

- H PCM as the overall heat transfer coefficient which takes place between the refrigerant
PT

flowing through the evaporator and the air inside the refrigerator during the compressor's
on/off cycles, for an evaporator coupled to a PCM (Equation (6)). H PCM is determined for a
CE

physical layout that is defined by; aluminum wall of the evaporator, a refrigerant fluid layer
which passes between the two evaporator walls, a PCM layer of a variable thickness X f , and
an air layer surrounding the PCM.
AC

- H as the overall heat transfer coefficient which takes place between the refrigerant flowing
through the evaporator and the air inside the refrigerator during the compressor's on and off
cycles, for an evaporator without a PCM (Equation (5)). H is defined for a physical layout,
which consists of; aluminum wall of the evaporator, a refrigerant fluid layer and an air layer
surrounding evaporator surface.

H is considered constant as it only depends on the operating conditions of the refrigerator,

and not on the variation of the PCM thickness. However, unlike H , H PCM is not constant and
it depends on the variation of the PCM thickness, as it is defined as a function of hPCM .
 ACCEPTED MANUSCRIPT

1 1 e 1
   (5)
H hair Kevap hrefrigerant

1 1 1 e 1
    (6)
H PCM hPCM hair K evap hrefrigerant

The calculation of the convective heat transfer coefficient for an evaporation of a refrigerant
flowing through an evaporator is based on the widely used formula of Dittus-Boettler [36].

T
0.8 0.4
 DGL   CpL  L 

IP
hrefrigerant D
 0.023     (7)
KL  L   K L 

CR
.
m
With: GL  L (8)
s

US
The convective heat exchange coefficient of the inside air is calculated from the
approximations of McAdams [37]. For natural convection, it is assumed that the difference of
AN
temperature between the evaporator's surface and the air surrounding the evaporator is the
driving force for the heat transfer inside the refrigerator. The heat transfer coefficient for a
natural convection is based on the Rayleigh number Raair :
M

g  (Tair  T ) L3
Raair  (9)
aairair
ED

While:
PT

T  Tevap when H is calculated.

T  TPCM when H PCM is calculated.

CE

According to the value of the Rayleigh number calculated, the flow is turbulent ( Nuair > 109).
AC

The Nusselt number Nuair is expressed as follows:

1
Nuair  0.13Raair3 (10)
 ACCEPTED MANUSCRIPT

The convective heat transfer coefficient of the air is written as follows:

K air Nuair
hair  (11)
L

Coefficients H and H PCM are presented on Fig. 6 and Fig. 7 for a compressor's on cycle and
off cycle respectively. Comparison between the two coefficients H and H PCM shows that
more the PCM thickness grows and more H PCM decreases vis-a-vis H . The difference between

T
the two coefficients is defined by a parameter in percentage. This latter is expressed in

IP
percentage, and has been defined as the ‘PCM heat transfer loss coefficient’ DPCM ,
represented by Equation (12):

CR
 H  H PCM 
DPCM    .100 (12)
 H 

US
AN
6 2.85

overall heat transfer coefficient Hpcm (W/m2.K)

overall heat transfer coefficient H (W/m2.K)

5.5 2.8
M
ED

5 2.75
PT

4.5 2.7
CE

4 2.65
AC

3.5 2.6
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
PCM thickness Xf (m)

Fig. 6. Overall heat transfer coefficients (H) and (HPCM). Compressor's on cycle.
 ACCEPTED MANUSCRIPT

4 2.412

overall heat transfer coefficient Hpcm (W/m2.K)

overall heat transfer coefficient H (W/m2.K)

3.5 2.41

3 2.408

2.5 2.406

T
IP
2 2.404

CR
1.5 2.402
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04

US
PCM thickness Xf (m)

Fig. 7. Overall heat transfer coefficients (H) and (HPCM). Compressor's off cycle.
AN
It is noted that PCM substantially reduces heat transfer within the refrigerator. The heat
transfer coefficient H is reduced by 18.28 % during the compressor off cycle (melting
process), while reduction reaches 45.18% during the compressor on cycle (solidification
M

process). These values are obtained by calculating the average of DPCM in the range of PCM
thickness X f (5-40 mm). Figure 8 illustrates the evolution of the coefficient DPCM for both
ED

melting and solidification processes.

PT
CE
AC
 ACCEPTED MANUSCRIPT

50

PCM heat transfer loss coefficient Dpcm (%)

45

40
Average value =45.28% Solidification process
Melting process
35

30

T
IP
25
Average value= 18.28%

CR
20

15
0 0.01 0.02 0.03 0.04 0.05 0.06

US
PCM thickness Xf (m)

Fig. 8. PCM heat transfer loss coefficient DPCM. Solidification/ melting processes.
AN

The air temperature profiles in the refrigerator compartment during the on/off cycles and for
M

an evaporator with and without a PCM are shown in Fig. 9. The temperature is measured at
three (3) equidistant levels in the refrigeration compartment. For each level of the refrigerator,
four (4) thermocouples are mounted. The average temperature of each level is defined as the
ED

average of the four (4) measured temperatures. The reduction in heat transfer within the
refrigerator generates an increase in the refrigerator air temperature of up to 1.1°C when
compared to the refrigerator without PCM. This is obtained by calculating the difference
PT

between the average measured air temperature for a refrigerator with PCM and without PCM.
CE
AC
 ACCEPTED MANUSCRIPT

3.5 case without PCM

case with PCM

3
Temperature (°C)

2.5

T
IP
1.5

CR
1

0.5
0 20 40 60 80 100 120

US
Time (min)

Fig. 9. Air temperature profiles. Cases with and without PCM. Compressor's on/off cycle.
AN

2.6. PCM effect on refrigerator's operating condition

M

As mentioned in section (2.4), the use of a PCM as a thermal storage device extends the
ED

refrigerator's downtime, and thus reduces the number of compressor’s on cycles for a period
of time (see Fig. 10). A high number of compressor’s on cycles generate an energy
overconsumption of the refrigerator, which is mainly due to 2 parameters:
PT

-When the compressor stops, a pressure balance in the refrigerant circuit of the refrigerator
occurs between the condenser and the evaporator. Thus, the high pressure refrigerant in the
CE

condenser migrates to the low pressure refrigerant in the evaporator. This transfer of mass
from high to low pressure results in a transfer of energy from the condenser to the evaporator.
In order to offset this additional energy, the compressor is required to work longer during
AC

each on cycle [25].

-The high current at each start of the compressor's cycle contributes to an increase in power
consumption, and could be significant over a long period of the refrigerator’s operation.
 ACCEPTED MANUSCRIPT

-2
case without PCM
-4 case with PCM

-6

-8
Temperature (°C)

-10

-12

-14

T
-16

IP
-18

CR
-20 Compressor Off
Compressor off
-22
0 20 40 60 80 100 120

US
time (min)

Fig. 10. Evaporation temperatures during the compressor’s on/off cycles. With PCM and without PCM.
AN
Table 2 shows the measured values of the daily operating time, the numbers of compressor’s
on cycles, and the power consumption under test conditions.
M

Table 2
Daily operating time, number of on cycles and power consumption of
the refrigerator (experimental).
ED

Reduction in the Reduction in Reduction in the daily

number of daily on- power operating time
cycles consumption
PT

14.5 % 10 % 17.6 %
CE

Power consumption of the refrigerator is measured by an energy meter (Phywe model) for a
AC

period of 24 hours. The value of the reduction in power consumption in Table 2 is obtained
between a refrigerator with and without PCM.

Reduction in the number of daily cycles between a refrigerator with and without PCM is
obtained by defining N as the number of cycles over a period of 24 hours, expressed by the
following equation. The value of the reduction in number of daily cycles in Table 2 is
obtained by a comparison between a refrigerator with and without PCM.
 ACCEPTED MANUSCRIPT

(24.3600)
N (13)
(to  ts )

The parameters to and ts (in sec) are defined from the experimental measurements operated
on the refrigerator. They represent respectively the operating and stopping time of one (1)
refrigerating cycle. These values are quasi-constant as the refrigerator operating mode
consists on a series of successive stopping and operating cycles under constant operating
conditions.

The value of the reduction in daily operating time tmo in Table 2 is obtained between a

T
refrigerator with and without PCM. The daily operating time is defined by the following

IP
equation:

CR
tmo  to .N (14)

2.7. PCM effect on the coefficient of performance (COP)

US
The improvement of the evaporator's performance is associated with an increase in the
AN
evaporation temperature of the refrigerant. This allows the refrigerant flowing into the
evaporator to evaporate more easily, reducing the compressor's work and improving COP.
M

The improvement of the heat transfer at the evaporator generates an increase in the vapors of
the refrigerant inside evaporator. This allows a slight increase in the saturation pressure of the
refrigerant and an increase in its evaporation temperature [26, 27]. This increase in
ED

evaporation temperature is shown in Fig. 10.

In order to define the effect of PCM on the performance of the evaporator, the overall heat
PT

transfer coefficient from evaporator H PCM _ evap is determined (Equation (15)). This coefficient
quantifies the heat flux which takes place between the refrigerant evaporator and the PCM,
when the evaporator is in contact with PCM. This coefficient gives an overview on the
CE

performance of the evaporator after adding the PCM, and is compared to the constant overall
heat transfer coefficient from evaporator without PCM, H .
AC

This coefficient is determined for a physical layout that is defined by; aluminum wall of the
evaporator, a refrigerant fluid layer and a PCM layer of a variable thickness X f .

1 1 e 1
   (15)
H PCM _ evap hPCM Kevap hrefrigerant
 ACCEPTED MANUSCRIPT

Fig. 11 and Fig. 12 illustrate the overall heat transfer coefficient H PCM _ evap compared to the
case without PCM. They show evolution of the coefficient according to thickness, for melting
and solidification processes, respectively.

Heat transfer coefficients are calculated during the compressor's on and off cycles. It is noted
that an improvement in the heat transfer at the evaporator is from 260W/m2.K to 33.9W/m2.K,
calculated in the thickness range of 5–40 mm during the compressor's on cycle. This
improvement is due to the dominance term of a forced convective heat transfer coefficient of
the refrigerant through evaporator.

T
An improvement in the order of ~ 16W/m2K during the compressor's off cycle is also noticed,

IP
but with a less magnitude, which is due to the fact that the refrigerant is in a stagnation phase
when the compressor stops. Heat transfer doesn’t present any noticeable variation despite the

CR
thickness variation, which is most likely due to the dominance of natural convection into the
PCM.

US 300
AN

Hpcm-evap (W/m2.K)
5 200
M
H (W/m2.K)

ED

4 100
PT
CE

3 0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
PCM thickness Xf (m)
AC

Fig. 11. Heat transfer evolution at the evaporator. Solidification case.

 ACCEPTED MANUSCRIPT

4 16.5

Hpcm-evap (W/m2.K)
H (W/m2.K)

2 16

T
IP
CR
0 15.5
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04

US
PCM thickness Xf (m)

Fig. 12. Heat transfer evolution at the evaporator. Melting case.

AN
The measured COP values of the refrigeration cycle associated with their refrigerant's
evaporation temperatures are presented in Table 3. The experiment showed an average
increase of the evaporation temperature by 2°C and an average COP increase of 5.05% after
M

the addition of a 10 mm thickness PCM at the evaporator's surface.

Table 3
ED

COP values and the evaporator's temperature

COP with
PCM Tevap (°C) COP Tevap (°C) COP ΔTevap (°C)
without improvement
PT

PCM

2.77 -19.36 2.63 -21.25 5.05 % 1.89

CE

3. PCM vs Frost
AC

Frost formation is a phenomenon that often occurs in refrigerators equipped with evaporators
without a defrost system. Beyond a certain thickness, frost formed on an evaporator's surface
may decrease the heat transfer that takes place between the refrigerant flowing through the
evaporator and the air surrounding the evaporator within the refrigerator. A decrease in the
overall heat transfer at the evaporator lowers the evaporator's performance and generates an
overconsumption of energy by the compressor.

In this section, the overall heat transfer coefficients through the evaporator in the presence of
PCM and frost are calculated, in order to highlight the influence of the thickness of each, the
 ACCEPTED MANUSCRIPT

frost and the pure water PCM, on the heat transfer inside the refrigerator. Frost is considered
as water under its solid phase, while PCM is considered as pure water under its solid/liquid
transition cycle. The studied thickness range of the frost and the PCM are from 5 to 40 mm.

3.1. Calculation assumptions

In order to study the PCM/frost effect on heat transfer inside a refrigerator, some assumptions
and considerations are taken into account:

- Frost is a gradual deposit of ice at the evaporator's surface, formed very slowly over time
from air moisture, which undergoes condensation and solidification. It is considered as water

T
under its solid phase, which does not melt during the compressor's off cycle. Therefore, it is

IP
assumed that the heat transfer through the frost is by conduction.

CR
- The PCM is assumed to solidify and melt during the compressor's on and off cycles,
respectively. PCM is assumed to be a liquid pure water that solidifies during the compressor's
on cycle, and an ice solidly packed that melts during the compressor's off cycle. Natural

US
convection is taken into account in the heat transfer into PCM.

- Frost generally does not melt because its temperature substantially follows the evaporator's
temperature because it is in direct contact with the evaporator's surface. (Measured
AN
evaporator's surface temperature oscillates between a maximum of -3 and a minimum of -
19°C as mentioned in section 2.2). Therefore, frost is considered to be kept in its solid phase,
because its temperature is always lower than that of its phase change temperature (around
M

0°C). However, the phase change temperature of PCM is -11°C, and is chosen to be close to
the evaporator's surface temperature, in order to allow its liquid / solid phase change transition
to occur (see Fig. 4).
ED

- The process of the formation of frost on subfreezing surfaces is complicated. Leoni et al.
[38] conducted an experimental study on frost formation and development on cold flat plate
PT

surfaces, and showed that the plate temperature and the relative air humidity are the most
influential parameters regarding frost properties (thickness, density and thermal conductivity).
For purposes of comparison with the PCM, it is assumed that thermal conductivity of frost
CE

depends only on the density, and is considered constant in the entire frost layer independent of
the thickness variations within [39].
AC

- The properties of PCM-water are considered constant for the same operating temperature,
and its characteristics are taken from the database of the software COOLPACK [40].

3.2. Definition of thermal resistance of the frost and the PCM:

The calculation of the thermal conductivity of the frost on a flat surface is based on the
approximation of Kim et al. [39]:
 ACCEPTED MANUSCRIPT

K frost  0,132  3,13.104  frost  1,6.107  2frost (16)

Where  frost is the frost's density represented by the following correlation [38]:
  Tev  
1
Tair  3 
 frost   0, 627. 1  
. 1    0,565  ice (17)
  273.15   273.15  
 

Fig. 13 and Fig. 14 illustrate the evolution of the thermal resistance to heat transfer of the pure
water RPCM in comparison with frost R frost , for the solidification melting and processes,
respectively. These coefficients (expressed in m2.K/W) are defined by the inverse of the heat

T
transfer coefficient. They are only depending on the variation of thickness.

IP
RPCM is calculated from the characteristics of pure water (COOLPACK database) and its

CR
evolution depends on hPCM . The PCM is assumed to solidify and melt during the compressor's
on and off cycles. For the calculation of hPCM , it is assumed that only convection is taken into

solidification process [34].

US
account during the melting process [35], and only conduction is taken into account during

Moreover, R frost is calculated from the characteristics of frost according to Kim et al. [39].
AN
Unlike PCM, frost is considered to be always in its solid phase as it has been explained in
section 3.1, and only conduction is considered for the calculation of R frost .
M

R frost is considered as the thermal resistance caused by the frost, and RPCM is the thermal
resistance caused by the PCM, and are expressed by the following :
ED

1
RPCM  (18)
hPCM
PT

Xf
R frost  (19)
K frost
CE
AC
 ACCEPTED MANUSCRIPT

0.08

0.07
Rfrost
Thermal resistance R (m2.K/W)

0.06 Rpcm

0.05

0.04

T
0.03

IP
0.02

CR
0.01

0
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04

US
Material thickness Xf (m)

Fig. 13. Thermal resistance Rfrost vs RPCM. Compressor's on cycle.

AN

0.08
M

0.07
Rfrost
Rpcm
Thermal resistance R (m2.K/W)

ED

0.06

0.05
PT

0.04
CE

0.03

0.02
AC

0.01

0
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Material thickness Xf (m)

Fig. 14. Thermal resistance Rfrost vs RPCM. Compressor's off cycle.

 ACCEPTED MANUSCRIPT

It can be seen from Fig. 13 and Fig. 14 that the thermal resistance of the frost layer is always
greater than the thermal resistance of the PCM. The difference between these two profiles is
small in the interval of thinner widths, and increases with the increase in thickness. However,
PCM presents a lower thermal resistance and seems to be less sensitive to the increase in its
thickness in the range of the widths considered. This can be explained by the fact that the
PCM is more thermally conductive than frost. Also, the heat transfer through the PCM during
the off cycle of the compressor is governed by natural convection. We infer that this mode of
transfer compensates for the increase in thermal resistance correlating to the increase in the
thickness of the PCM. It can also be seen from Fig. 13 and Fig. 14 that the evolution of frost
thermal resistance is the same under constant operating conditions of the refrigerator. This is

T
due to the absence of phase change during both compressor’s on and off cycles.

IP
CR
3.3. Effect of the frost and the PCM on the evaporator's performance

In order to evaluate the effect of PCM and frost on the performance of the evaporator, the

US
overall heat transfer coefficient from evaporator is determined. This latter permits to quantify
the heat flux passing through the evaporator in contact with PCM, in comparison with an
evaporator in contact with frost.
AN
The overall heat transfer coefficient which takes place between the refrigerant evaporator and
the PCM is defined earlier by H PCM _ evap , while the coefficient between the refrigerant
M

evaporator and frost is defined by H frost _ evap . The latter is determined for a physical layout that
is defined by; aluminum wall of the evaporator, a refrigerant fluid layer and a frost layer, and
is expressed by Equation (20).
ED

1 Xf e 1
   (20)
PT

H frost _ evap K frost K evap hrefrigerant

CE

Fig. 15 and Fig. 16 illustrate the evolution of H PCM _ evap , in comparison with H frost _ evap
according to variation of the thickness, for the melting and solidification processes,
AC

respectively. It is noted from Fig. 15 that there is a marked improvement of heat transfer at the
evaporator during the compressor's on cycle for both the PCM and the frost, with a slight
advantage for the PCM. Nevertheless, it can be noted that this improvement decreases rapidly
as the thickness increases due to the conduction mode that governs heat transfer inside both
the PCM and the frost during this cycle. However, it is noticed that during the compressor's
off cycle (see Fig. 16), the heat transfer is improved for both the PCM and the frost but at a
lesser magnitude than for the on cycle. This is due to the fact that heat transfer is more
affected by the forced thermal convection of the refrigerant in motion inside the evaporator as
it operates, than when the refrigerant is in its stagnation phase during the compressor's off
cycle. It is also noted that the value of H PCM _ evap is almost insensitive to the increases in layer
 ACCEPTED MANUSCRIPT

thickness in the range of the widths considered. This is due to the heat transfer mode which is
assumed to be natural convection in PCMs.

300
Overall heat transfer coefficient (W/m2.K)

250

Hpcm-evap
200 Hfrost-evap

T
150

IP
CR
100

50

0
0.005 0.01 0.015 0.02
US 0.025 0.03 0.035 0.04
AN
Material thickness Xf (m)

Fig. 15. Heat transfer evolution at the evaporator. Compressor's on cycle.

M

17

16
ED
Overall heat transfer coefficient (W/m2.K)

15

14
PT

Hpcm-evap
13 Hfrost-evap

12
CE

11

10
AC

7
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Material thickness Xf (m)

Fig. 16. Heat transfer evolution at the evaporator. Compressor's off cycle.
 ACCEPTED MANUSCRIPT

3.4. Effect of the frost and the PCM on heat transfer

In order to quantify the effect of PCM and frost on heat transfer, we introduce the coefficients
DPCM and Dfrost , which we define as heat transfer loss coefficients of the PCM and the frost
layer, respectively. The coefficients DPCM and Dfrost have been calculated in the range of
PCM thickness X f (5-40 mm).

The coefficient DPCM compares the overall heat transfer coefficient which takes place between
the refrigerant evaporator and the air inside refrigerator in presence of PCM H PCM , with the

T
same coefficient for the case of an evaporator alone H . This coefficient is expressed in

IP
percentage, and has been defined as the ‘PCM heat transfer loss coefficient’, represented by
Equation (12).

CR
Dfrost is a comparison between the overall heat transfer coefficient which take place between
the refrigerant evaporator and the air inside refrigerator in presence of frost H frost (equation

US
22), with the case of an evaporator alone. This coefficient has been defined as the ‘frost heat
transfer loss coefficient’, represented by Equation (21):
AN
 H  H frost 
D frost    .100 (21)
 H 
M

1 1 Xf e 1
With:     (22)
H frost hair K frost K evap hrefrigerant
ED

The evolution of DPCM and Dfrost in the range of PCM thickness X f (5-40 mm) is showed by
Fig. 17 and Fig. 18, for the cases of melting and solidification processes, respectively. Two
PT

values of thickness (10 and 40 mm) for both PCM and frost have been given as an example, to
show how the thickness of each material affects the heat transfer inside the refrigerator.
CE

It is noted from Fig. 17 that for an evaporator coupled to a PCM, during the compressor’s on
cycle, a 10 mm thickness of PCM presents a loss coefficient DPCM of 42.7% in the overall
heat transfer compared to an evaporator without a PCM. This reduction is 46.1% for a
AC

thickness of 40 mm. However, the variation of the frost loss coefficient Dfrost shows more
sensitivity to the thickness than the PCM loss coefficient DPCM . Frost presents a loss
coefficient of 44.46 % for a 10 mm thickness, and reaches 51.8% for a thickness of 40 mm.
PCM and frost impact the heat transfer inside the refrigerator with the same magnitude.

However, when the compressor stops (see Fig. 18), it is noted that the PCM has a lesser
impact on the heat transfer compared to the frost. For comparison, a 10 mm thickness PCM
shows a loss of 18% in heat transfer compared to an evaporator without a PCM and an 18.3%
loss for a thickness of 40 mm. On the other hand, frost presents a heat transfer loss coefficient
 ACCEPTED MANUSCRIPT

Dfrost of 21.22% for a 10 mm thickness, and reaches 30.3% for a thickness of 40 mm. Frost
may then affect more significantly the heat transfer that occurs between the refrigerant
flowing through the evaporator and the air within the refrigerator, causing an increase in the
air temperature inside the refrigerator, which presents a higher risk of deterioration of stored
products.

52

51
Heat transfer loss coefficient D (%)

50 Dpcm 51.8%

T
Dfrost
49

IP
48

CR
46.1%
47
44.5%
46

45

44
42.7% US
AN
43

42
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
M

Material thickness Xf (m)

Fig. 17. Thermal loss coefficients, DPCM and Dfrost. Compressor's on cycle.
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

32
30.3%
Heat transfer thermal loss coefficient D (%)
30
Dpcm
28 Dfrost

26

24 21.22%

T
22

IP
18.3%

CR
20 18.05%

18
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04

US
Material thickness Xf (m)

Fig. 18. Thermal loss coefficients, DPCM and Dfrost. Compressor's off cycle.
AN
M

4. Conclusion

The study based on experimental measurements and a mathematical model deals with the
ED

effects of PCM (phase change material) on the operating conditions and the heat transfer
within a domestic refrigerator equipped with a flat plate evaporator. A comparative study
dealing with the effects of the PCM and the frost on the heat transfer inside the refrigerator is
PT

also discussed. The results from analytical and experimental findings can be summarized as
follows:
CE

- The PCM's phase change temperature must be close to the evaporator's temperature in order
for the phase change phenomenon to occur.
AC

- The addition of PCM layer generates an improvement in the heat transfer at the evaporator
resulting in COP and a reduction in the daily consumption. However, it would reduce the heat
transfer that occurs between the refrigerant and the air inside refrigerator, generating a slight
increase in the air temperature.

- The frost's effect on heat transfer is more evident during the compressor's off cycle when
compared to PCM. However, PCM and frost impact the heat transfer with the same
magnitude during the compressor's on cycle.
 ACCEPTED MANUSCRIPT

NOMENCLATURE

𝑎𝑎𝑖𝑟 =Thermal diffusivity of air (m2/s)

Cp =Specific heat [J/(kg K)]

D =Evaporator thickness (m)

e =Thickness of evaporator wall (m)

T
IP
g =Gravity acceleration (m/s2)

h =Convective heat transfer coefficient [W/(m2 K)]

CR
H =Overall heat transfer coefficient [W/(m2 K)]

L
=Thermal conductivity

=Characteristic length US [W/(m K)]

(m)
AN
𝑚̇𝐿 =Refrigerant mass flow (Kg/s)

s =Surface crossed by refrigerant [Kg/(s m2)]

M

Tevap =Evaporator temperature (°C)

ΔTevap =Evaporator temperature variation (°C)

ED

Tair =Air temperature inside refrigerator (°C)

TPCM =PCM phase change temperature (°C)

PT

Xf =PCM/frost thickness (m)

CE

Greek letters
AC

𝜈 =Cinematic viscosity of air (m2/s)

µ =Dynamical viscosity of refrigerant (Pa s)

𝜌 =Ice density (Kg/m3)

𝛽 =Thermal expansion coefficient (1/K)

 ACCEPTED MANUSCRIPT

Dimensionless

Ra =Rayleigh number Ra = gβ(T − T∞ )L3 /a ν

Nu =Nusselt number Nu = hL/K

T
Subscripts

IP
CR
L =Liquid refrigerant

PCM =Phase Change Material

air =Air inside the refrigeratorUS

AN
frost =Frost

ice =Ice
M

ACKNOWLEDGEMENTS
ED

The authors are thankful to the Unité de Développement des Equipements Solaires (UDES
epst CDER) for the research support.
PT

REFERENCES
CE

[1] L.A. Chidambaram, A.S. Ramana, G. Kamaraj, R. Velraj, Review of solar cooling
methods and thermal storage options, Renewable and Sustainable Energy Reviews 15 (2011)
AC

3220-3228.
[2] M. Laidi, S. Hanini, B. Abbad, M. Berdja, M. Ouali, The study and performance of a
modified ENIEM conventional refrigerator to serve as a photovoltaic powered one under
Algerian climate conditions, Journal of Renewable and Sustainable Energy 4 (2012) 053112.
[3] M.M. Farid, A.M. Khudhair, S.A. Razack, S. Al-Hallaj, A review on phase change energy
storage: materials and applications, Energy Conversion and Management 45 (2004) 1597–
1615.
[4] B. Zalba, J. Marin, L.F. Cabeza, H. Mehling, Review on thermal energy storage with
phase change materials: heat transfer analysis and applications, Applied Thermal Engineering
23 (2003) 251–283.
 ACCEPTED MANUSCRIPT

[5] E. Oro, A De Garcia, A. Castell, M.M. Farid, L.F. Cabeza, Review on phase change
materials (MCPs) for cold thermal energy storage applications, Applied Energy 99 (2012)
513–533.
[6] M. Bashar, K. Siddiqui, Investigation of heat transfer during melting of a PCM by a U-
shaped heat source, International journal of energy research 41 (2017) 2091–2107.
[7] M.H. Abokersh, M. Osman, O. El-Baz, M. El–Morsi, O. Sharaf, Review of the phase
change material (PCM) usage for solar domestic water heating systems
(SDWHS), International journal of energy research 42 (2017) 329–357.
[8] U. Stritih, V. Butala, Energy saving in building with PCM cold storage, International
journal of energy research 31 (2007) 1532–1544.

T
[9] C. Tzivanidis, K.A. Antonopoulos, E.D. Kravvaritis, Parametric analysis of space cooling

IP
systems based on night ceiling cooling with PCM–embedded piping, International journal of
energy research 36 (2012) 18–35.

CR
[10] T. Kousksou, A. Arid, A. Jamil, Y. Zeraouli, Thermal behavior of building material
containing microencapsulated PCM, Thermochimica Acta 550 (2012) 42–47.
[11] M. Esen, T. Ayhan, Development of a model compatible with solar assisted cylindrical

US
energy storage tank and variation of stored energy with time for different phase change
materials, Energy Conversion and Management 37(12) (1996) 1775–1785.
[12] M. Esen, A. Durmus¸ and A. Durmus, Geometric design of solar–aided latent heat store
AN
depending on various parameters and phase change materials, Solar Energy 62(1) 1998 19–
28.
[13] M. Esen, Thermal performance of a solar-aided latent heat store used for space
heating by heat pump, Solar Energy 69(1) 2000 15–25.
M

[14] S. Cantor, DSC study of melting and solidification of salt hydrates, Themochimica Acta
33 (1979) 69–86.
ED

[15] S. Wu, D. Zhu, X. Li, H. Li, J. Lei, Thermal energy storage behavior of Al2O3–H2O
nanofluids, Thermochimica Acta 483 (2009) 73–77.
[16] A. Sciacovelli, F. Colella, V. Verda, Melting of PCM in a thermal energy storage unit:
PT

Numerical investigation and effect of nanoparticle enhancement. International journal of

energy research 37 (2013) 1610–1623.
[17] Y. Li, B. Guo, G. Huang, P. Shu, H. Kiriki, S. Kubo, K. Ohno, T. Kawai, Eutectic
CE

compound (KNO3/NaNO3 :PCM) quasi-encapsulated into SiC-honeycomb for suppressing

natural convection of melted PCM, International journal of energy research 39 (2015) 789–
804.
AC

[18] K.A.R. Ismail, O.C. Quispe, J.R. Henriquez, A numerical and experimental study on a
parallel plate ice bank, Applied Thermal Engineering, Applied Thermal Engineering 19
(1999) 163–193.
[19] K.A.R. Ismail, A. Batista de Jesus, Modeling and solution of the solidification problem
of PCM around a cold cylinder, numerical heat transfer 36(a) (1999) 95–114.
[20] F. Wang, G. Maidment, J. Missenden, R. Tozer, The novel use of phase change materials
in refrigeration plant. Part 1: Experimental investigation, Applied Thermal Engineering 27
(2007) 2893–2901.
 ACCEPTED MANUSCRIPT

[21] F. Wang, G. Maidment, J. Missenden, R. Tozer, The novel use of phase change materials
in refrigeration plant. Part 2: Dynamic simulation model for the combined system, Applied
Thermal Engineering 27 (2007) 2902–2910.
[22] F. Wang, G. Maidment, J. Missenden, R. Tozer, The novel use of phase change materials
in refrigeration plant. Part 3: MCP for control and energy savings, Applied Thermal
Engineering 27 (2007) 2911–2918.
[23] A. Marques, G. Davies, G. Maidment, J. Evans, I. Wood, Novel design and performance
enhancement of domestic refrigerators with thermal storage, Applied Thermal Engineering 63
(2014) 511-519.
[24] F. Alzuwaid, Y. Ge, S. Tassou, A. Raeisi, L. Gowreesunker, The novel use of phase

T
change materials in a refrigerated display cabinet: An experimental investigation, Applied

IP
Thermal Engineering 75 (2015) 770-778.
[25] B. Gin, M.M. Farid, P. Bansal, Effect of door opening and defrost cycle on a freezer with

CR
phase change panels, Energy Conversion and Management 51 (2010) 2698-2706.
[26] K. Azzouz, D. Leducq, D. Gobin, Performance enhancement of a household refrigerator
by addition of latent heat storage, International Journal of Refrigeration 31 (2008) 892–901.

US
[27] K. Azzouz, D. Leducq, D. Gobin, Enhancing the performance of household refrigerators
with latent heat storage: an experimental investigation, International journal of Refrigeration
32 (2009) 1634-1644.
AN
[28] Z. Li, D. Zhao, G. Ding, T. Ren, S. Miao, X. Han, T. Noda, Improving defrosting
performance by controlling frost distribution to match defrosting heat distribution in frost-free
household refrigerators, International Journal of Refrigeration 77 (2017) 136–148
[29] M.A. Stein, C. Inan, C. Bullard, T. Newell, Closed door moisture transport in
M

refrigerator/freezers, International journal of energy research 26 (2002) 793–805.

[30] I.R. Cox-Smith, C.G. Carrington, Assessment of a domestic heat pump 2: Component
ED

and frosting analysis, International journal of energy research 14 (1990) 27–37.

[31] S. Sengupta, S.A. Sherif, K.V. Wong, Empirical heat transfer and frost thickness
correlations during frost deposition on a cylinder in cross-flow in the transient regime,
PT

International journal of energy research 22 (1998) 615–624.

[32] K.A.R. Ismail, C. Salinas, M. M. Gonçalves, Frost growth around a cylinder in a wet air
stream, International Journal of Refrigeration 20(2) (1997) 106–119.
CE

[33] K.A.R. Ismail, C. Salinas, Modeling of frost formation over parallel cold plates,
International Journal of Refrigeration 22 (1999) 425–441.
[34] P. Lamberg, R. Lehtiniemi, A.M. Henell, Numerical and experimental investigation of
AC

melting and freezing processes in phase change material storage, International Journal of
Thermal Sciences 43 (2004) 277–287.
[35] J.S. Lim, A. Bejan, The Prandtl number effect on melting dominated by natural
convection, ASME Journal of heat transfer 114(3) (1992) 784–787.
[36] Y. Çengel, A. Ghajar, Heat and mass transfer: Fundamentals and Applications, McGraw-
Hill, 4th Edition, (2011) ISBN-10: 0077366646
[37] S. Ben Amara, écoulements et transferts thermique par convection naturelle dans les
milieux macroporeux alimentaires, application aux réfrigérateurs ménagers. PhD thesis,
l’Institut National Agronomique Paris-Grignon, France (2005).
 ACCEPTED MANUSCRIPT

[38] A. Leoni, M. Mondot, F. Durier, R. Revellin, P. Haberschill, Frost formation and

development on flat plate: Experimental investigation and comparison to predictive methods,
Experimental Thermal and Fluid Science 88 (2017) 220–233.
[39] C. Kim, J. Lee, K.S. Lee, Numerical modeling of frost growth and densification on a cold
plate using frost formation resistance, International Journal of Heat and Mass Transfer 115
(2017) 1055–1063.
[40] COOLPACK Refrigeration Utilities database.

T
IP
CR
US
AN
M
ED
PT
CE
AC