University of Birmingham

Thermo-mechanical analysis of microcapsules

containing phase change materials for cold storage
Yu, Qinghua; Tchuenbou-Magaia, Fideline; Al-Duri, Bushra; Zhang, Zhibing; Ding, Yulong; Li,
Yongliang
DOI:
10.1016/j.apenergy.2017.12.021
License:
Creative Commons: Attribution-NonCommercial-NoDerivs (CC BY-NC-ND)

Document Version
Peer reviewed version

Citation for published version (Harvard):

Yu, Q, Tchuenbou-Magaia, F, Al-Duri, B, Zhang, Z, Ding, Y & Li, Y 2018, 'Thermo-mechanical analysis of
microcapsules containing phase change materials for cold storage', Applied Energy, vol. 211, pp. 1190-1202.
https://doi.org/10.1016/j.apenergy.2017.12.021

Link to publication on Research at Birmingham portal

Publisher Rights Statement:

Accepted for publication in Applied Energy, publication forthcoming

General rights
Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the
copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes
permitted by law.

•Users may freely distribute the URL that is used to identify this publication.
•Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private
study or non-commercial research.
•User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?)
•Users may not further distribute the material nor use it for the purposes of commercial gain.

Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document.

When citing, please reference the published version.

Take down policy
While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been
uploaded in error or has been deemed to be commercially or otherwise sensitive.

If you believe that this is the case for this document, please contact [email protected] providing details and we will remove access to
the work immediately and investigate.

Download date: 18. Oct. 2020

 Thermo-mechanical analysis of microcapsules containing phase

change materials for cold storage

Qinghua Yu, Fideline Tchuenbou-Magaia, Bushra Al-Duri, Zhibing Zhang, Yulong Ding,

Yongliang Li *

Birmingham Centre for Energy Storage, School of Chemical Engineering, University of

Birmingham, Birmingham B15 2TT, United Kingdom

*
Corresponding author. Tel.: +44 (0) 121 414 5135, Email: [email protected] (Y. Li)
Abstract

Microencapsulated phase change material slurries (MEPCMSs) offer a potentially

efficient and flexible solution for cryogenic-temperature cold storage. In this paper, the phase

change material (PCM) microcapsules prepared to form MEPCMSs for cryogenic-

temperature cold storage consist of Dowtherm J (DJ) as core material and melamine

formaldehyde (MF) as primary shell material. DJ is an aromatic mixture with diethylbenzene

as the main component. Composite shell materials are adopted to avoid cracking by adding

aluminium oxide (Al2O3) nanoparticles or copper (Cu) coating into/on MF shell. In order to

explore the heat transfer behaviour and mechanical stability of the microcapsules during the

solidification process of PCM, a thermo-mechanical model is established by taking into

account of energy conservation, pressure-dependent solid-liquid equilibria, Lamé’s equations

and buckling theory. Based on the proposed model, the effects of shell thickness, shell

compositions and microcapsule size are therefore studied on the variations of pressure

difference, freezing point, and latent heat. The cause of shell deformation is clearly explained

and the shell buckling modes are predicted using the model, which agree well with the

experimental observations. The critical core/shell size ratios of avoiding buckling are

proposed for the microcapsules with different compositions. Simultaneously incorporation of

Al2O3 nanoparticles and Cu coating into/on MF shell can markedly enhance the resistant to

buckling. In addition, special attention is paid to cold energy storage capacity of MEPCMSs,

which has considerable superiority compared to packed pebble beds.

Keywords: Phase change materials; Microencapsulation; Solidification; Shell buckling; Cold

storage.
Nomenclature
Roman letters 𝜀𝜀 strain
𝑎𝑎 shell thickness (m) 𝜅𝜅 correction factor
𝑐𝑐𝑝𝑝 specific heat (J·kg-1·K-1) 𝜆𝜆 thermal conductivity (W·m-1·K-1)
𝐸𝐸 Young’s modulus (Pa) 𝜈𝜈 Poisson’s ratio
𝐸𝐸𝐸𝐸 stored energy (J) 𝜌𝜌 density (kg·m-3)
𝑓𝑓 volumetric fraction 𝜎𝜎 stress (Pa)
𝐹𝐹 Legendre function
𝑔𝑔 chemical potential (kJ·kg-1) Subscripts
ℎ enthalpy (kJ·kg-1) 0 reference or initial
𝑘𝑘𝑓𝑓 foundation modulus (N/m3) 𝑎𝑎 atmospheric
𝐿𝐿 latent heat (kJ·kg-1) 𝑏𝑏 buckling
𝑛𝑛 buckling mode number 𝑐𝑐 shell
𝑃𝑃 pressure (Pa) 𝑐𝑐𝑐𝑐 critical
𝑟𝑟 radius (m) 𝑒𝑒 external surface of shell
𝑠𝑠 entropy (J·kg-1·K-1) 𝑒𝑒𝑒𝑒 equivalent
𝑡𝑡 time (s) 𝑓𝑓 freezing or freezing front
𝑇𝑇 temperature (K) 𝑖𝑖 shell/PCM interface or PCM
𝑢𝑢 displacement (m) 𝑙𝑙 liquid
𝑉𝑉 volume (m3) 𝑚𝑚 microcapsule
𝑟𝑟, 𝜃𝜃, 𝜑𝜑 spherical coordinates system
Greek letters 𝑠𝑠 solid
𝛼𝛼 thermal expansion coefficient (K-1)
𝛽𝛽 isothermal compressibility (Pa-1) Superscripts
𝛾𝛾 surface tension (N·m-1) ∗ holistic
𝛿𝛿, 𝜇𝜇 Lamé’s constant
1. Introduction

Liquid air energy storage (LAES) and pumped thermal electricity storage (PTES) are

two emerging grid scale thermal storage technologies, which are good solutions for the

intermittency and instability of electricity from renewable energy sources [1-3]. Cryogenic-

temperature cold storage is key to improving the overall performance of LAES and PTES

systems [4-6]. At present, the two systems generally utilize packed beds for cryogenic-

temperature cold storage. However, packed beds have much room for improvement in energy

storage capacity, efficiency and flexibility [7-10]. Microencapsulated phase change material

slurries (MEPCMSs) have great potential for dynamic and static cryogenic-temperature cold

storage applications as they combine the advantages of phase change materials (PCMs) and

liquid sensible energy storage materials, and are both transport media (heat transfer fluids)

and thermal storage media. MEPCMSs consist of a carrier liquid and PCM microcapsules

with a diameter of <100 μm, in general, small enough to be suspended in a carrier liquid.

Such partially melting and solidifying slurries can offer very high energy storage densities

and heat transfer rates in charging/discharging processes [11]. The good flowability of the

MEPCMSs allows them to be transported through pumping, and thus their flow rate can be

easily adjusted to realize the desired stored amount of cold energy and objective temperature.

Furthermore, their apparent specific heats at set temperatures can be designed by addition of

microcapsules with different melting point core PCM, in order to meet the significant specific

heat changes of transcritical/supercritical fluids [12]. Therefore, the MEPCMSs can offer a

much more flexible strategy for cold storage, which is extremely difficult to achieve using the

conventional packed bed.

The utilisation of MEPCMSs will also have a significant impact on the cryogenic

industry such as natural gas liquefaction and cold recovery in re-gasification, and air

separation/liquefaction [13, 14]. However, most of the research has been conducted only on

1
moderate or high temperature MEPCMSs with melting points above -20 ˚C [15-18], whereas

little research can be found on cryogenic MEPCMSs. Technically, the cryogenic MEPCMSs

are more challenging compared to MEPCMSs applied at moderate temperatures due to

deformation or fragility of the shell of microcapsules and poor heat transfer under cryogenic

conditions. The success of MPCMSs in cryogenic temperature cold storage is dependent on

the stability of microcapsules under repeated pumping, cyclic heating and cooling as well as

long-term storage. As a result, it is important to understand the thermo-mechanical behaviour

of MEPCMs in particular during the PCM solidification process.

Several studies have been devoted to the thermo-mechanical behaviour of encapsulated

PCM. A composite of mixed graphite and nitrate salts is considered as a solid sphere of PCM

encapsulated in a thick shell of graphite by Lopez et al. [19] and the shell was modelled as a

closed elastic spherical shell with a mobile internal wall and a non-moving external wall.

Based on this model, the effects of the shell Young’s modulus on the internal pressure,

melting point and latent heat, were examined. Pitié et al. [20] extended the model to a shell of

silicon carbide (SiC) with a free mobile external wall by incorporating the Lamé equations.

The variation of internal pressure due to the volume change during the melting process was

analytically calculated based on the extended model with a given volume fraction of melted

salts, leading to variations of melting point, enthalpy and stored energy. This indicates that

the coated PCM should have a low volumetric expansion causing a lower pressure increase so

that the coating SiC shell can avoid cracking. Based on the model, the temperature and

pressure evolutions during the melting and solidification processes of copper-encapsulated

nitrate spheres were simulated at a constant surrounding temperature by Parrado et al. [21]. In

the simulations the heat transfer equation was decoupled with the mechanical stress equation.

Zhao et al. [22] compared the time of the melting/solidification process between metal and

non-metal encapsulated PCM particles using numerical simulations of heat transfer regardless

2
of pressure variation. The above investigations are only based on high-temperature thermal

energy storage and thermo-mechanical analysis of millimeter-scale encapsulated PCM

particles. Mechanical response and properties of microcapsules near room temperature were

also evaluated via experiments by Giro-Paloma et al. [23] and Su et al. [24], without

considering the heat transfer behaviour. However, the thermo-mechanical behaviours of PCM

microcapsules have rarely been studied for the purpose of cold storage. In particular, the

effects of shell thickness and compositions on the thermo-mechanical behaviours have not

been clearly addressed by previous studies.

It should be noted that the PCM solidification processes in cold storage are different

from those in heat storage in terms of internal pressure and deformation mechanism of shells

[20, 25, 26]. Because of the volume shrinkage of PCM during solidification in cold storage

application, the internal pressure of microcapsule decreases while the external pressure is

constant [27, 28]. When the external pressure is higher than the internal pressure, the

spherical microcapsule shell is only subjected to uniform external pressure. The morphology

or deformation of such a pressurised spherical shell is then crucial to its properties, such as

optical, electromagnetic and heat transfer. The analytical studies of structural behaviour or

buckling of complete spherical/spheroidal shells under external pressure have been widely

conducted for various objects, including pressure vessels, spherical honeycombs [29], natural

fruits and vegetables [30], spherical viruses [31] and biological cells [32]. Timoshenko et al.

[33] was first to introduce the formulation and solving approach for pressurised buckling of

an empty and complete spherical shell based on the axisymmetric assumption and Rayleigh–

Ritz approach. Sato et al. [34] conducted comparative studies between the exact and

simplified approaches to validate the approximation based on the axisymmetric assumption

and Rayleigh–Ritz approach. These works show that the approximate formulations enable

sufficiently accurate values of the critical buckling pressure and the corresponding buckling

3
mode number to be obtained. It can also be inferred from the work of Sato et al. [34] that

when 𝑟𝑟𝑖𝑖 𝑘𝑘𝑓𝑓 ⁄𝐸𝐸𝑐𝑐 <10-3 (𝑟𝑟𝑖𝑖 is the core radius; 𝑘𝑘𝑓𝑓 is the core foundation modulus; and 𝐸𝐸𝑐𝑐 is the

shell Young’s modulus), the buckling behaviours of the shell filled with elastic materials is

the same as that for the empty shell.

This paper presents a first attempt to understand the thermo-mechanical behaviour of

spherical microcapsules containing PCM for cold storage application. The microcapsule

fabrication process as well as shell modification is described and the morphologies of

microcapsules are observed for mechanical analysis. A thermo-mechanical model is

established for a single microcapsule during the PCM solidification process, taking into

account energy conservation, pressure variation caused by volume shrinkage, pressure-

dependent solid-liquid equilibria, shell elastic deformation and buckling behaviour. As

𝑟𝑟𝑖𝑖 𝑘𝑘𝑓𝑓 ⁄𝐸𝐸𝑐𝑐 <10-5 in the present study which will be stated in Section 4, it is reasonable to

assume that the buckling theory and the corresponding solving approach for empty complete

spherical shells proposed by Timoshenko et al. [33] are applicable for the buckling analysis

of the shell of PCM microcapsules [34]. On the basis of the model, the influences of shell

thickness, shell composition and microcapsule size on the solidification process are studied,

including the variations of pressure difference, freezing point, latent heat, solidification

period and stored energy. The model is used to predict the critical bulking pressure and

buckling mode of microcapsules for specific shell thickness and composition. The predicted

buckling mode is then compared with experimental observations to validate the proposed

model. The energy storage capacities are also compared between MEPCMSs and typical

packed beds in LAES and PTES systems. This study can provide significant references for

the design of PCM microcapsules without buckling and with better cold storage performance

for MEPCMSs.

4
2. Microcapsule fabrications and testing

The microcapsule studied in this paper consists of melamine formaldehyde (MF) as

shell material and Dowtherm J (DJ) as core PCM. The DJ is an aromatic mixture containing

diethylbenzene as the main component with a freezing point of -81˚C, which was supplied by

Dow Chemical Company, US. MF precondensate was purchased from British Industrial

Plastics Ltd., UK. The MF shell microcapsules were fabricated via the in-situ polymerization

method [35]. The morphologies of fabricated microcapsules after undergoing thermal cycling

test were observed by a cryogenic scanning electron microscopy (Cryo-SEM, FEI Quanta

600 FEG SEM equipped with a Quorum PP2000T Cryo-stage). Fig. 1(a) displays the Cryo-

SEM image of the microcapsules with pure MF as shell material. It is observed that most of

the microcapsules are broken, which is likely due to high brittleness at cryogenic temperature

[36]. In order to avoid cracking, composite shell materials were adopted by adding

aluminium oxide (Al2O3) nanoparticles into MF or electroless copper (Cu) plating on the

surface of the MF shell to improve shell mechanical properties. All chemicals used in the

fabrication process of microcapsules were purchased from the Sigma-Aldrich, Inc., UK,

unless otherwise specified.

The procedure of adding Al2O3 nanoparticles is as follows: The surfaces of Al2O3

nanoparticles need to be modified by silane coupling agent KH-570 (Sinopharm Chemical

Reagent Co., Ltd., China) prior to its addition to the MF shell [37]. A certain amount of

Al2O3 nanoparticles were dispersed into C2H5OH by rapid agitation for 30 min and ultrasonic

vibration for 30 min. KH-570 dissolved in C2H5OH was subsequently added into the

suspension. After undergoing ultrasonic vibration for 30 min and continuous stirring for 2 h,

the reaction mixture was then filtered and rinsed with deionized water several times, and

finally dried in a vacuum dryer to obtain the modified Al2O3 nanoparticles. A certain amount

of modified Al2O3 nanoparticles were dispersed into core oil DJ with rapid agitation and

5
ultrasonic vibration. The other procedures about emulsification and polymerization were

similar to the fabrication of the pure MF shell microcapsules [35]. Eventually, most of the

Al2O3 nanoparticles were embedded into the MF shell [37, 38].

A typical procedure of electroless Cu plating is described as follows: Surface

sensitization was first implemented by dispersing the MF or MF/Al2O3 shell microcapsules

with clean surfaces in an aqueous solution of SnCl2 and HCl at 30℃ for 15 min. The

sensitised MF microcapsules were then cleaned with deionized water and dispersed in a

solution of PdCl2 at 30℃ for 15 min to accomplish surface activation. The activated MF

microcapsules were then cleaned with deionized water and dispersed in the electroless Cu

plating solution to form a Cu coating. The solution comprised CuSO4, NaOH, HCHO,

NaKC4H4O6 and Na2EDTA (ethylene diamine tetraacetic acid). The pH and temperature of

the plating solution were adjusted to about 12 and 30℃, respectively.

The microcapsule with Cu coating is referred to as MF-Cu; the one with both Al2O3

nanoparticles in shell and Cu coating is named MF-Cu-Al. The Cryo-SEM images of the MF-

Cu and MF-Cu-Al microcapsules are shown in Figs. 1(b) and 1(c), respectively. The images

show that buckling has occurred for some of the MF-Cu microcapsules while the MF-Cu-Al

microcapsules still keep their spherical shape. The occurrence of buckling is closely related to

the mechanical properties of shell materials and the ratio of shell thickness to core radius,

which will be discussed in Section 4. The microcapsules that did not crack during thermal

cycling were then cut into two by an ultramicrotome (Reichert-Jung Ultracut E) to observe

the morphology under the Cryo-SEM. Transmission electron microscope (TEM, JEOL

1200EX) was used to examine shell thickness of the microcapsules. It is also shown in Figs.

1(a-c) that the diameter of the microcapsules is around 10 µm. And the particle size

measurements indicate that the size distribution of the fabricated microcapsules ranges from

10 µm to 100 µm as shown in Fig. 1(d). The chemical studies were performed on a SEM

6
(Hitachi TM3030) equipped with an energy-dispersive spectrometer (EDS). The existence of

Cu on the surface of the so-called MF-Cu and MF-Cu-Al microcapsules is confirmed by the

Cu element peaks in the EDS spectra as shown in Figs. 1(e, f). Similarly, the existence of

Al2O3 on the surface of the so-called MF-Cu-Al microcapsules is certified by the Al element

peak in the EDS spectrum as shown in Fig. 1(f), which indicates that the Al2O3 nanoparticles

have been successfully integrated into the MF shell.

3. Mathematical Models

3.1. Geometry and initial hypotheses

The geometry of a spherical microcapsule is shown in Fig. 2, including a shell and

solid/liquid PCM. The external radius of the microcapsule is labelled 𝑟𝑟𝑒𝑒 . The position of the

shell/PCM interface and the solidification front are referred to as 𝑟𝑟𝑖𝑖 and 𝑟𝑟𝑓𝑓 , respectively.

These parameters are dependent on time 𝑡𝑡 during solidification process.

The main hypotheses applied in this model concerning the PCM and shell are as

follows: (a) the density 𝜌𝜌𝑙𝑙 , specific heat 𝑐𝑐𝑝𝑝𝑝𝑝 , and thermal conductivity 𝜆𝜆𝑙𝑙 are constant,

independent of pressure and temperature for the liquid phase of PCM; (b) the liquid pressure

within the shell is uniform; (c) as a result of the micro-size capsule, convection heat transfer

inside the shell is negligible; (d) viscous energy dissipation is also neglected; (e) the solid

phase of PCM possesses homogeneously constant values of density 𝜌𝜌𝑠𝑠 , specific heat 𝑐𝑐𝑝𝑝𝑝𝑝 and

thermal conductivity 𝜆𝜆𝑠𝑠 ; (f) the solid phase of PCM is deformable along with the shell

without effect on the shell deformation; (g) the shell is considered to be a homogeneous,

isotropic and exhibiting linear elastic behaviour indicated by Young’s modulus, with constant

values of density 𝜌𝜌𝑐𝑐 , specific heat 𝑐𝑐𝑝𝑝𝑝𝑝 and thermal conductivity 𝜆𝜆𝑐𝑐 ; (h) the external surface of

the shell is at known and uniform temperature and pressure; (i) the conditions of temperature

continuity and heat flux conservation are satisfied at the solidification front; (j) there are
7
equalities of temperature and pressure at the shell/PCM interface. The spherical symmetry

before buckling from above mentioned hypotheses allows reduction of the original three-

dimensional problem of transfer to a one-dimensional one [19, 20].

3.2. Expression of pressure variation due to volume shrinkage

During the solidification process, the volume shrinkage of the PCM caused by the

density difference between solid and liquid phases at a time 𝑡𝑡 is

𝜌𝜌𝑙𝑙 − 𝜌𝜌𝑠𝑠 ∗
∆𝑉𝑉 = 𝑉𝑉𝑙𝑙0 � � 𝑓𝑓 (𝑡𝑡) (1)
𝜌𝜌𝑠𝑠

where 𝑓𝑓 ∗ is the ratio of solidified volume at a time t to the initial volume 𝑉𝑉𝑙𝑙0 of liquid PCM,

which is referred to as solid fraction.

In view of the spherical symmetry of the studied microcapsule before buckling in a

spherical coordinate system (𝑟𝑟, 𝜃𝜃, 𝜑𝜑), the displacement, strain and stress fields of the shell

due to elastic deformation are only dependent on 𝑟𝑟 among the three coordinates. Furthermore,

the displacement 𝑢𝑢 only has radial component 𝑢𝑢𝑟𝑟 (i.e. 𝑢𝑢 = 𝑢𝑢𝑟𝑟 ); the strain only has normal

strain components 𝜀𝜀𝑟𝑟𝑟𝑟 , 𝜀𝜀𝜃𝜃𝜃𝜃 and 𝜀𝜀𝜑𝜑𝜑𝜑 with 𝜀𝜀𝜃𝜃𝜃𝜃 = 𝜀𝜀𝜑𝜑𝜑𝜑 ; and the stress also only has normal

stress components 𝜎𝜎𝑟𝑟𝑟𝑟 , 𝜎𝜎𝜃𝜃𝜃𝜃 and 𝜎𝜎𝜑𝜑𝜑𝜑 with 𝜎𝜎𝜃𝜃𝜃𝜃 = 𝜎𝜎𝜑𝜑𝜑𝜑 . Therefore, the equilibrium equation

without the body force can be simplified as

𝑑𝑑𝑑𝑑𝑟𝑟𝑟𝑟 2(𝜎𝜎𝑟𝑟𝑟𝑟 − 𝜎𝜎𝜃𝜃𝜃𝜃 )

+ = 0. (2)
𝑑𝑑𝑑𝑑 𝑟𝑟

The pressure at the shell/PCM interface is equal to the liquid pressure 𝑃𝑃, while the

pressure at the external surface of the shell is equal to ambient pressure 𝑃𝑃𝑎𝑎 (atmospheric).

Thus, the boundary conditions for the elastic deformation of the shell are

𝜎𝜎𝑟𝑟𝑟𝑟 (𝑟𝑟 = 𝑟𝑟𝑖𝑖0 ) = −𝑃𝑃, 𝜎𝜎𝑟𝑟𝑟𝑟 (𝑟𝑟 = 𝑟𝑟𝑒𝑒0 ) = −𝑃𝑃𝑎𝑎 , (3)

where 𝑟𝑟𝑖𝑖0 and 𝑟𝑟𝑒𝑒0 are the initial radii of the internal and external walls of the shell,

respectively.

8
 The shell undergoes temperature change ∆𝑇𝑇 during the PCM solidification process. The

thermal stress should be taken into account, which is proportional to the thermal expansion

coefficient of the shell material 𝛼𝛼𝑐𝑐 . By combining strain-displacement and stress-strain

relations with thermal stress [20], the stress-displacement relations are obtained as

𝑑𝑑𝑑𝑑 2𝑢𝑢 𝑑𝑑𝑑𝑑

𝜎𝜎𝑟𝑟𝑟𝑟 = 𝛿𝛿 � + � + 2𝜇𝜇 − (3𝛿𝛿 + 2𝜇𝜇)𝛼𝛼𝑐𝑐 ∆𝑇𝑇, (4)
𝑑𝑑𝑑𝑑 𝑟𝑟 𝑑𝑑𝑑𝑑

𝑑𝑑𝑑𝑑 2𝑢𝑢 2𝜇𝜇𝜇𝜇

𝜎𝜎𝜃𝜃𝜃𝜃 = 𝛿𝛿 � + �+ − (3𝛿𝛿 + 2𝜇𝜇)𝛼𝛼𝑐𝑐 ∆𝑇𝑇, (5)
𝑑𝑑𝑑𝑑 𝑟𝑟 𝑟𝑟

where 𝛿𝛿 and 𝜇𝜇 are Lamé’s constants related to the Young’s modulus 𝐸𝐸𝑐𝑐 and Poisson’s ratio 𝜈𝜈𝑐𝑐

as

𝐸𝐸𝑐𝑐 𝜈𝜈𝑐𝑐
𝛿𝛿 = ,
(1 + 𝜈𝜈𝑐𝑐 )(1 − 2𝜈𝜈𝑐𝑐 )
(6)
𝐸𝐸𝑐𝑐
𝜇𝜇 = .
2(1 + 𝜈𝜈𝑐𝑐 )

Substituting Eqns. (4) and (5) into Eqn. (2), yields the simplified Lamé’s equations as

𝑑𝑑 2 𝑢𝑢 2 𝑑𝑑𝑑𝑑 2𝑢𝑢
+ − = 0. (7)
𝑑𝑑𝑟𝑟 2 𝑟𝑟 𝑑𝑑𝑑𝑑 𝑟𝑟 2

Solving Eqn. (7) with the boundary conditions in Eqn. (3), yields the elastic description

of the system as
3 3 3 3
1 𝑟𝑟𝑖𝑖0 𝑟𝑟𝑒𝑒0 (𝑃𝑃 − 𝑃𝑃𝑎𝑎 ) (𝑟𝑟𝑖𝑖0 𝑃𝑃 − 𝑟𝑟𝑒𝑒0 𝑃𝑃𝑎𝑎 )𝑟𝑟
𝑢𝑢(𝑟𝑟) = 𝑟𝑟𝛼𝛼𝑐𝑐 ∆𝑇𝑇 + 3 � + �, (8)
𝑟𝑟𝑒𝑒0 − 𝑟𝑟𝑖𝑖03 4𝑟𝑟 2 𝜇𝜇 3𝛿𝛿 + 2𝜇𝜇

3 3
1 𝑟𝑟𝑖𝑖0 𝑟𝑟𝑒𝑒0 (𝑃𝑃 − 𝑃𝑃𝑎𝑎 ) 3 3
𝜎𝜎𝑟𝑟𝑟𝑟 (𝑟𝑟) = 3 3 �− + (𝑟𝑟𝑖𝑖0 𝑃𝑃 − 𝑟𝑟𝑖𝑖0 𝑃𝑃𝑎𝑎 )�, (9)
𝑟𝑟𝑒𝑒0 − 𝑟𝑟𝑖𝑖0 𝑟𝑟 3

3 3
1 𝑟𝑟𝑖𝑖0 𝑟𝑟𝑒𝑒0 (𝑃𝑃 − 𝑃𝑃𝑎𝑎 ) 3 3
𝜎𝜎𝜃𝜃𝜃𝜃 (𝑟𝑟) = 3 3 � + (𝑟𝑟𝑖𝑖0 𝑃𝑃 − 𝑟𝑟𝑒𝑒0 𝑃𝑃𝑎𝑎 )�.
𝑟𝑟𝑒𝑒0 − 𝑟𝑟𝑖𝑖0 2𝑟𝑟 3 (10)

For 𝑟𝑟 = 𝑟𝑟𝑖𝑖0 , the volume displacement is written as

4 3
∆𝑉𝑉 = 𝜋𝜋[(𝑢𝑢(𝑟𝑟𝑖𝑖0 ) + 𝑟𝑟𝑖𝑖0 )3 − 𝑟𝑟𝑖𝑖0 ], (11)
3

9
 From Eqns. (1) and (11), it can be derived that

3 𝜌𝜌𝑙𝑙 − 𝜌𝜌𝑠𝑠
𝑢𝑢(𝑟𝑟𝑖𝑖0 ) = 𝑟𝑟𝑖𝑖0 � � 𝑓𝑓 ∗ (𝑡𝑡) + 1 − 1�, (12)
𝜌𝜌𝑠𝑠

Before shell formation in the in-situ polymerization process, the liquid PCMs are

dispersed in water as spherical droplets. Due to the surface tension, the internal pressure of

the droplet is larger than the external pressure. The internal pressure of the droplet can be

calculated according to the Young–Laplace equation. It can be assumed that the initial

internal pressure inside microcapsules after shell formation is equal to the internal pressure of

the droplet. The initial internal pressure is denoted as 𝑃𝑃𝑖𝑖 . Then the initial displacement at

𝑟𝑟 = 𝑟𝑟𝑖𝑖0 before solidification of PCM can be calculated as follows:

3 3 3 3
1 𝑟𝑟𝑖𝑖0 𝑟𝑟𝑒𝑒0 (𝑃𝑃𝑖𝑖 − 𝑃𝑃𝑎𝑎 ) (𝑟𝑟𝑖𝑖0 𝑃𝑃𝑖𝑖 − 𝑟𝑟𝑒𝑒0 𝑃𝑃𝑎𝑎 )𝑟𝑟
𝑢𝑢0 = 3 3 � 2
+ �. (13)
𝑟𝑟𝑒𝑒0 − 𝑟𝑟𝑖𝑖0 4𝑟𝑟 𝜇𝜇 3𝛿𝛿 + 2𝜇𝜇

Therefore, Eqn. (12) can be changed to

3 𝜌𝜌𝑙𝑙 − 𝜌𝜌𝑠𝑠
𝑢𝑢(𝑟𝑟𝑖𝑖0 ) = 𝑟𝑟𝑖𝑖0 � � 𝑓𝑓 ∗ (𝑡𝑡) + 1 − 1� + 𝑢𝑢0 , (14)
𝜌𝜌𝑠𝑠

and by combining with Eqn. (8) gives

3 3 3 3 (1
2(𝑟𝑟𝑒𝑒0 − 𝑟𝑟𝑖𝑖0 )𝐸𝐸𝑐𝑐 � �(𝜌𝜌𝑙𝑙 − 𝜌𝜌𝑠𝑠 )𝑓𝑓 ∗ ⁄𝜌𝜌𝑠𝑠 + 1 − (1 + 𝛼𝛼𝑐𝑐 ∆𝑇𝑇) + 𝑢𝑢0 ⁄𝑟𝑟𝑖𝑖0 � + 3𝑟𝑟𝑒𝑒0 − 𝜈𝜈𝑐𝑐 )𝑃𝑃𝑎𝑎
𝑃𝑃 = 3 (2 3 . (15)
𝑟𝑟𝑖𝑖0 − 4𝜈𝜈𝑐𝑐 ) + 𝑟𝑟𝑒𝑒0 (1 + 𝜈𝜈𝑐𝑐 )

3.3. Expression of freezing point and latent heat dependent on pressure

The liquid-solid phase equilibrium can be considered to exist at the solidification front,

making the chemical potential of liquid phase equal to that of solid phase. The chemical

potential can be approached by a second order Taylor expansion on the basis of some

fundamental thermodynamic relations, which is expressed as [19]

1 1 𝑐𝑐𝑝𝑝𝑝𝑝0 2
𝑔𝑔𝑗𝑗 �𝑇𝑇𝑓𝑓 , 𝑃𝑃� = 𝑔𝑔𝑗𝑗0 − 𝑠𝑠𝑗𝑗0 �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 � + (𝑃𝑃 − 𝑃𝑃0 ) − �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 � (16)
𝜌𝜌𝑗𝑗0 2 𝑇𝑇0

10
 1 𝛽𝛽𝑗𝑗0 𝛼𝛼𝑗𝑗0
− (𝑃𝑃 − 𝑃𝑃0 )2 + �𝑇𝑇 − 𝑇𝑇𝑓𝑓0 �(𝑃𝑃 − 𝑃𝑃0 ),
2 𝜌𝜌𝑗𝑗0 𝜌𝜌𝑗𝑗0 𝑓𝑓

where the index 𝑗𝑗 = 𝑙𝑙 or 𝑠𝑠 denotes liquid or solid phase; 𝑇𝑇𝑓𝑓 represents the freezing

temperature at the pressure 𝑃𝑃; 𝑔𝑔𝑗𝑗0 = 𝑔𝑔𝑗𝑗 �𝑇𝑇𝑓𝑓0 , 𝑃𝑃0 � represents the chemical potential at 𝑇𝑇𝑓𝑓0 and

𝑃𝑃0 ; 𝑇𝑇𝑓𝑓0 is the freezing temperature at 𝑃𝑃0 representing reference pressure (atmospheric); 𝑠𝑠𝑖𝑖 is

the specific entropy; 𝛼𝛼𝑖𝑖 represents the thermal expansion coefficient; and 𝛽𝛽𝑖𝑖 represents the

isothermal compressibility; the subscript 0 refers to �𝑇𝑇𝑓𝑓0 , 𝑃𝑃0 � conditions.

Applying the liquid-solid equilibrium condition (𝑔𝑔𝑙𝑙 = 𝑔𝑔𝑠𝑠 ), the following equation for

freezing temperature as a function of pressure is obtained:

1 1 1 𝑐𝑐𝑝𝑝𝑝𝑝0 − 𝑐𝑐𝑝𝑝𝑝𝑝0 2
0 = −(𝑠𝑠𝑙𝑙0 − 𝑠𝑠𝑠𝑠0 )�𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 � + � − � (𝑃𝑃 − 𝑃𝑃0 ) − � � �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 �
𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0 2 𝑇𝑇0
(17)
1 𝛽𝛽𝑙𝑙0 𝛽𝛽𝑠𝑠0 𝛼𝛼𝑙𝑙0 𝛼𝛼𝑠𝑠0
− � − � (𝑃𝑃 − 𝑃𝑃0 )2 + � − � �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 �(𝑃𝑃 − 𝑃𝑃0 ).
2 𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0 𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0

Then via factorizing by �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 � for Eqn. (17), the solution is derived as

−𝑐𝑐 + √𝑐𝑐 2 − 4𝑏𝑏𝑏𝑏

𝑇𝑇𝑓𝑓 (𝑃𝑃) = 𝑇𝑇𝑓𝑓0 + , (18)
2𝑏𝑏

with

1 𝑐𝑐𝑝𝑝𝑝𝑝0 − 𝑐𝑐𝑝𝑝𝑝𝑝0
𝑏𝑏 = � �,
2 𝑇𝑇𝑓𝑓0

𝛼𝛼𝑙𝑙0 𝛼𝛼𝑠𝑠0
𝑐𝑐 = (𝑠𝑠𝑙𝑙0 − 𝑠𝑠𝑠𝑠0 ) − � − � (𝑃𝑃 − 𝑃𝑃0 ), (19)
𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0

1 1 1 𝛽𝛽𝑙𝑙0 𝛽𝛽𝑠𝑠0
𝑑𝑑 = − � − � (𝑃𝑃 − 𝑃𝑃0 ) + � − � (𝑃𝑃 − 𝑃𝑃0 )2 .
𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0 2 𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0

In a similar way, the variation of latent heat with pressure can be predicted. The

enthalpy difference between the liquid and solid phases (i.e. latent heat) at thermodynamic

equilibrium (𝑔𝑔𝑙𝑙 = 𝑔𝑔𝑠𝑠 ) can be expressed as [19]

𝐿𝐿𝑓𝑓 �𝑇𝑇𝑓𝑓 , 𝑃𝑃� = ∆𝑠𝑠𝑓𝑓 �𝑇𝑇𝑓𝑓 , 𝑃𝑃�𝑇𝑇𝑓𝑓 . (20)

11
 For estimation of entropy variation ∆𝑠𝑠𝑓𝑓 �𝑇𝑇𝑓𝑓 , 𝑃𝑃�, Eqns. (16) allows writing

𝜕𝜕𝑔𝑔𝑗𝑗 𝑐𝑐𝑝𝑝𝑝𝑝0 𝛼𝛼𝑗𝑗0

𝑠𝑠𝑗𝑗 ≡ − � = 𝑠𝑠𝑗𝑗0 + �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 � − (𝑃𝑃 − 𝑃𝑃0 ), (21)
𝜕𝜕𝜕𝜕 𝑃𝑃 𝑇𝑇0 𝜌𝜌𝑖𝑖0

and thus

𝑐𝑐𝑝𝑝𝑝𝑝0 − 𝑐𝑐𝑝𝑝𝑝𝑝0 𝛼𝛼𝑙𝑙0 𝛼𝛼𝑠𝑠0

∆𝑠𝑠𝑓𝑓 = (𝑠𝑠𝑙𝑙0 − 𝑠𝑠𝑠𝑠0 ) + � � �𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 � − � − � (𝑃𝑃 − 𝑃𝑃0 ). (22)
𝑇𝑇𝑓𝑓0 𝜌𝜌𝑙𝑙0 𝜌𝜌𝑠𝑠0

Values of parameters 𝑠𝑠𝑙𝑙0 , 𝑠𝑠𝑠𝑠0 , 𝜌𝜌𝑙𝑙0 , 𝜌𝜌𝑠𝑠0 , 𝑐𝑐𝑝𝑝𝑝𝑝0 , 𝑐𝑐𝑝𝑝𝑝𝑝0, 𝛼𝛼𝑙𝑙0 , 𝛼𝛼𝑠𝑠0 , 𝛽𝛽𝑙𝑙0 , 𝛽𝛽𝑠𝑠0 are usually available

in the corresponding thermodynamic data bases. Eqns. (18-20) and (22) account for the

variation of freezing temperature and latent heat with pressure.

3.4. Heat transfer modelling for spherical microcapsules

The enthalpy method based on a fixed grid [39] was used to model the PCM

solidification process while the temperature was directly solved. According to the hypotheses

(a) to (g), energy conservation equation can be written as

𝜕𝜕 ��𝜌𝜌𝑐𝑐𝑝𝑝 �𝑒𝑒𝑒𝑒 𝑇𝑇𝑖𝑖 � 1 𝜕𝜕 𝜕𝜕𝑇𝑇𝑖𝑖 𝜕𝜕�𝜌𝜌𝑒𝑒𝑒𝑒 Δℎ𝑓𝑓 �

2
= �𝜆𝜆 𝑒𝑒𝑒𝑒 𝑟𝑟 � − for 0 ≤ 𝑟𝑟 ≤ 𝑟𝑟𝑖𝑖 ,
𝜕𝜕𝜕𝜕 𝑟𝑟 2 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕
(23)
𝜕𝜕�𝜌𝜌𝑐𝑐 𝑐𝑐𝑝𝑝𝑝𝑝 𝑇𝑇𝑐𝑐 � 1 𝜕𝜕 𝜕𝜕𝑇𝑇𝑐𝑐
= 2 �𝜆𝜆𝑐𝑐 𝑟𝑟 2 � for 𝑟𝑟𝑖𝑖 < 𝑟𝑟 ≤ 𝑟𝑟𝑒𝑒 ,
𝜕𝜕𝜕𝜕 𝑟𝑟 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕

where �𝜌𝜌𝑐𝑐𝑝𝑝 �𝑒𝑒𝑒𝑒 is the equivalent heat capacity; 𝑇𝑇𝑖𝑖 denotes the temperature distributions in the

PCM layer; 𝜆𝜆𝑒𝑒𝑒𝑒 is the equivalent thermal conductivity; 𝜌𝜌𝑒𝑒𝑒𝑒 is the equivalent density; Δℎ𝑓𝑓

represents the solidification enthalpy which can be defined as a product of latent heat 𝐿𝐿𝑓𝑓 and

local liquid fraction 𝑓𝑓𝑙𝑙 , i.e. Δℎ𝑓𝑓 = 𝑓𝑓𝑙𝑙 𝐿𝐿𝑓𝑓 ; and 𝑇𝑇𝑐𝑐 is the temperature distributions in the shell

layer. �𝜌𝜌𝑐𝑐𝑝𝑝 �𝑒𝑒𝑒𝑒 , 𝜌𝜌𝑒𝑒𝑒𝑒 and 𝜆𝜆𝑒𝑒𝑒𝑒 are given by

�𝜌𝜌𝑐𝑐𝑝𝑝 �𝑒𝑒𝑒𝑒 = 𝜌𝜌𝑙𝑙 𝑐𝑐𝑝𝑝𝑝𝑝 𝑓𝑓𝑙𝑙 + 𝜌𝜌𝑠𝑠 𝑐𝑐𝑝𝑝𝑝𝑝 (1 − 𝑓𝑓𝑙𝑙 ),

(24)
𝜌𝜌𝑒𝑒𝑒𝑒 = 𝜌𝜌𝑙𝑙 𝑓𝑓𝑙𝑙 + 𝜌𝜌𝑠𝑠 (1 − 𝑓𝑓𝑙𝑙 ),

12
 𝜆𝜆𝑒𝑒𝑒𝑒 = 𝜆𝜆𝑙𝑙 𝑓𝑓𝑙𝑙 + 𝜆𝜆𝑠𝑠 (1 − 𝑓𝑓𝑙𝑙 ).

For the pure PCM with a fixed freezing point, the relationship between local liquid

fraction and temperature can be described as

1, 𝑇𝑇𝑝𝑝 ≥ 𝑇𝑇𝑓𝑓
𝑓𝑓𝑙𝑙 (𝑟𝑟, 𝑡𝑡) = � . (25)
0, 𝑇𝑇𝑝𝑝 < 𝑇𝑇𝑓𝑓

Boundary conditions of the problem are

𝜕𝜕𝑇𝑇𝑖𝑖
−𝜆𝜆𝑒𝑒𝑒𝑒 = 0 at 𝑟𝑟 = 0,
𝜕𝜕𝜕𝜕

𝜕𝜕𝑇𝑇𝑖𝑖 𝜕𝜕𝑇𝑇𝑐𝑐 (26)

𝜆𝜆𝑒𝑒𝑒𝑒 = 𝜆𝜆𝑐𝑐 , and 𝑇𝑇𝑖𝑖 = 𝑇𝑇𝑐𝑐 at 𝑟𝑟 = 𝑟𝑟𝑖𝑖 ,
𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕

𝑇𝑇𝑐𝑐 = 𝑇𝑇𝑒𝑒 (𝑡𝑡) at 𝑟𝑟 = 𝑟𝑟𝑒𝑒 ,

where 𝑇𝑇𝑒𝑒 (𝑡𝑡) is the temperature at the external surface of shell. Initially, 𝑇𝑇𝑖𝑖 (𝑟𝑟, 0) = 𝑇𝑇𝑐𝑐 (𝑟𝑟, 0) =

𝑇𝑇0 , which is uniform. For the integration of the phase change into pressure variation in Eqn.

(15), an expression calculating 𝑓𝑓 ∗ (𝑡𝑡) is required, which can be written as:

3 𝑟𝑟𝑖𝑖 2
𝑓𝑓 ∗ (𝑡𝑡) = 1 − � 𝑟𝑟 𝑓𝑓𝑙𝑙 (𝑟𝑟, 𝑡𝑡)𝑑𝑑𝑑𝑑. (27)
𝑟𝑟𝑖𝑖03 0

The total energy stored within the microcapsule during solidification mainly consists of

latent energy and sensible energy, which can be expressed as

𝑓𝑓 ∗ 𝑇𝑇(𝑓𝑓 ∗ )
∗)
𝐸𝐸𝐸𝐸(𝑓𝑓 = � 𝜌𝜌𝑙𝑙 𝑉𝑉𝑙𝑙0 𝐿𝐿𝑓𝑓 𝑑𝑑𝑑𝑑 + � 𝑉𝑉𝑙𝑙0 �𝜌𝜌𝑙𝑙 𝑐𝑐𝑝𝑝𝑝𝑝 𝑓𝑓𝑙𝑙 + 𝜌𝜌𝑠𝑠 𝑐𝑐𝑝𝑝𝑝𝑝 (1 − 𝑓𝑓𝑙𝑙 )�𝑑𝑑𝑑𝑑. (28)
0 𝑇𝑇(𝑓𝑓 ∗ =0)

3.5. Buckling of uniform pressurised spherical shells

If a spherical shell is subjected to uniform external pressure, it may retain its spherical

form and undergo only a uniform compression with radial displacement. The magnitude of

the uniform compressive stress in this case can be calculated by Eqns. (9) and (10). If the

pressure increases beyond a certain limit, the spherical equilibrium form of the compressed

shell may become unstable and buckling occurs [33]. As described before, the axisymmetric

assumption has little effect on the calculated values of the critical buckling pressure and the
13
corresponding buckling mode number [34]. Thus, it is assumed that the buckling deformation

is axisymmetric with respect to the vertical axis in order to simplify the calculation of the

critical pressure. Considering the discrepancy between theory and experiment existing during

buckling of spherical shell under uniform external pressure [33, 40], the actual critical

buckling pressure can be obtained by modifying the theoretical expression, which is as

follows [33]:

2𝐸𝐸𝑐𝑐 𝑎𝑎2
𝑃𝑃𝑐𝑐𝑐𝑐 = 𝜅𝜅 (29)
𝑟𝑟𝑖𝑖2 �3(1 − 𝜈𝜈𝑐𝑐2 )

where 𝑎𝑎 is the shell thickness and 𝜅𝜅 is the correction factor which equals about 0.7 [41].

Because of axisymmetric buckling, the small displacements of the shell during bulking

from the compressed spherical form only have the components 𝑢𝑢𝑏𝑏𝜃𝜃 and 𝑢𝑢𝑏𝑏𝑟𝑟 , respectively, in

meridian and radial directions 𝜃𝜃 and 𝑟𝑟. The two components are calculated by [33]:
∞
𝑑𝑑𝑑𝑑𝑛𝑛 (cos 𝜃𝜃)
𝑢𝑢𝑏𝑏𝜃𝜃 = � 𝐴𝐴𝑛𝑛
𝑑𝑑𝑑𝑑
𝑚𝑚=1
(30)
∞

𝑢𝑢𝑏𝑏𝑟𝑟 = � 𝐵𝐵𝑛𝑛 𝐹𝐹𝑛𝑛 (cos 𝜃𝜃)

𝑛𝑛=1

where 𝐹𝐹𝑛𝑛 (cos 𝜃𝜃) is Legendre functions; 𝐴𝐴𝑛𝑛 , 𝐵𝐵𝑛𝑛 can be obtained by solving

𝐴𝐴𝑛𝑛 [𝑚𝑚𝑛𝑛 + (1 + 𝜈𝜈𝑐𝑐 ) + φ]+𝐵𝐵𝑛𝑛 [𝜔𝜔𝜔𝜔𝑛𝑛 + (1 + 𝜈𝜈𝑐𝑐 ) + φ] = 0,

𝐴𝐴𝑛𝑛 [𝜔𝜔𝑚𝑚𝑛𝑛2 + (1 + 𝜈𝜈𝑐𝑐 )(𝑚𝑚𝑛𝑛 + 2) + φ(𝑚𝑚𝑛𝑛 + 2)] (31)

+ 𝐵𝐵𝑛𝑛 [𝜔𝜔𝑚𝑚𝑛𝑛2 + (3 + 𝜈𝜈𝑐𝑐 )𝜔𝜔𝜔𝜔𝑛𝑛 + 2(1 + 𝜈𝜈𝑐𝑐 ) − φ(𝑚𝑚𝑛𝑛 − 2)] = 0,

where 𝜔𝜔 = 𝑎𝑎2 ⁄(12𝑟𝑟𝑖𝑖2 ); 𝑚𝑚𝑛𝑛 = 𝑛𝑛(𝑛𝑛 + 1) − 2; 𝑛𝑛 is an integer representing the buckling mode

number. The relation between φ and 𝑚𝑚𝑛𝑛 is as follows [33]:

(1 − 𝜈𝜈𝑐𝑐2 ) + 𝜔𝜔[𝑚𝑚𝑛𝑛2 + 2𝑚𝑚𝑛𝑛 + (1 + 𝜈𝜈𝑐𝑐 )2 ]

φ= (32)
𝑚𝑚𝑛𝑛 + (1 + 3𝜈𝜈𝑐𝑐 )

According to equation (32), 𝑛𝑛 is selected to obtain the smallest value of φ where

buckling may occur.

14
4. Results and discussions

Heat transfer and mechanical behaviour of the microcapsule during PCM solidification

are simulated using validated models in Section 3. The temperature at the external surface of

the microcapsule is decreasing as the microcapsules flow with the slurry in a heat exchanger

for the charging process of cold energy. It is thus assumed that the temperature at the external

surface of the microcapsule decreases from -80˚C to -85˚C at a cooling rate of 5˚C/min over

the freezing point of -81˚C in the simulations. The external pressure of microcapsule is

constant and equal to atmospheric pressure (≈0.1 MPa). As mentioned in Section 3.2, the

initial internal pressure calculated according to the Young–Laplace equation equals about

0.11 MPa for the microcapsule with 𝑟𝑟𝑖𝑖 = 5 µm.

The thermodynamic properties of PCM (DJ) at atmospheric pressure 𝑃𝑃0 are listed in

Table 1 [42]. The theoretical properties of pure MF are supplied in Table 2 for reference [43].

It should be noted that the actual properties of the fabricated shell depend on the actual

polymerization effect. Thus they are variable and different from the theoretical values,

especially for the Young’s modulus. The research of Giro-Paloma et al. [23] manifests that

the effective Young's modulus 𝐸𝐸𝑚𝑚 of primordial microcapsules with pure MF as shell is

subject to approximately normal random distribution with an average value of 30 MPa. The

value of 𝐸𝐸𝑚𝑚 has a linear relationship with the actual Young's modulus of the shell 𝐸𝐸𝑐𝑐 , which

is 𝐸𝐸𝑚𝑚 = 0.16𝑎𝑎𝐸𝐸𝑐𝑐 /(𝑟𝑟𝑖𝑖 + 𝑎𝑎) [44]. The calculated average value of 𝐸𝐸𝑐𝑐 for MF shell is therefore

about 0.6 GPa according to the values of 𝑟𝑟𝑖𝑖 and 𝑎𝑎 [23]. As described in Section 2, the shell is

made of composite materials of MF and Cu or Al2O3. The compositions of MF-Cu shell are

specified as 95% MF and 5% Cu by volume while those of MF-Cu-Al shell are specified as

90% MF, 5% Al2O3 and 5% Cu by volume. It is estimated by weighted average that 𝐸𝐸𝑐𝑐 is

around 1.0 GPa for MF-Cu shell and around 3.0 GPa for MF-Cu-Al shell. The Poisson’s
15
ratios for the three kinds of compositions are all between 0.2 and 0.4 and the value in this

range has little effect on the results. The Poisson’s ratio of the composite materials is thus

assumed to be 0.3. The estimated Young’s modulus can be used as a characteristic value to

represent the property changes caused by compositions for comparative analysis.

4.1. Validation of the model

In order to validate the proposed model, the results calculated based on the model

established in this paper were compared with those in literature [19, 39] for the same

problems and properties. Figs. 3(a, b) compare the temperature profile at t = 500 s and

solidification rate in terms of liquid fraction for a square cavity containing PCM between the

present study and the literature [39]. Results from both this study and literature are similar,

suggesting that the heat transfer model for phase change based on the enthalpy method in

Section 3.4 is reliable. Figs. 3(c, d) compare the variations of internal pressure, melting

temperature and latent heat during melting coupled with heat transfer for salt particles coated

in a graphite matrix between the current study and literature [19]. The results obtained in this

study show a satisfactory agreement with the literature, indicating that the pressure variation

model in Section 3.2 and pressure-dependent dynamic equilibrium model in Section 3.3

together with the heat transfer model in Section 3.4 are sufficiently accurate. Comparison of

the predicted results with experimental results in Section 4.5 also validates the buckling

models in Section 3.5.

4.2. Effects of shell thickness

The effects of shell thickness are analyzed for MF-Cu microcapsules with 𝑟𝑟𝑖𝑖 = 5 µm in

this section. Fig. 4(a) illustrates the evolutions of differences between external and internal

pressures under different shell thicknesses during the solidification process (𝑓𝑓 ∗ : 0 → 1). The

internal pressure is calculated from Eqn. (15). As a consequence of PCM volume shrinkage,

16
the internal pressure will be progressively decreased until zero and thus the pressure

difference will be progressively increased until 0.1 MPa. The increasing rate of pressure

difference decreases with the increase of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 . The critical pressures calculated by Eqn. (29)

are traced to compare the progress of the pressure differences to the buckling limits of the

shells. Buckling occurs only when the pressure difference increases to the critical pressure.

The position as buckling occurs is marked according to the critical pressure represented by

horizontal line in the figure. Shell bulking will not occur for 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 20 or 50, while shell

bulking will occur for 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 100 or 120. This accounts for the phenomenon that buckling

occurs for some of the MF-Cu microcapsules in Fig 1(b). Compared with the case where

𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 100, the solid fraction as bulking occurs for 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 120 is smaller because the critical

pressure is lower, although the pressure difference increases more slowly. The above results

imply that reducing shell thickness leads to shell buckling or makes buckling occur at the

earlier solidification stage. The buckling limit indicates that it becomes impossible to

consider the heat transfer behaviour of PCM based on the mentioned approach when the

pressure difference is beyond the critical pressure. In order to compare the different shell

thicknesses and compositions, the following Figs. 4(b-d) and 5 still consider up to 100% of

solidified PCM without buckling.

The freezing point variations ( 𝑇𝑇𝑓𝑓 − 𝑇𝑇𝑓𝑓0 ) calculated from Eqn. (18) at different

solidification stages are shown in Fig. 4(b). It can be found that the freezing point slightly

decreases until a constant value is reached according to zero internal pressure as the

solidification process carries on. The coated PCM will no longer solidify at constant

temperature before the internal pressure decreases to zero. The decreasing rate of freezing

point dependent on the internal pressure decreases with the increase of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 . The lowest

freezing point before shell buckling for 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 100 is lower than that for 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 120. Fig. 4(c)

shows the latent heat variations ( 𝐿𝐿𝑓𝑓 − 𝐿𝐿𝑓𝑓0 ) calculated from Eqn. (20) at different
17
solidification stages. As both the internal pressure and freezing point decrease, the latent heat

slightly increases. The increase of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 slows the increase of latent heat in the solidification

process.

Fig. 4(d) illustrates the solidification time of PCM obtained by solving Eqn. (23) under

different shell thicknesses. The thickness has little influence on the total solidification period

of PCM at a shell thermal conductivity of 0.5 W·m-1·K-1. The effects of shell thicknesses

with different thermal conductivities on the solidification time of PCM are shown in Fig. 5.

For 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 100, the thermal conductivity of shell has nearly no influence on the PCM

solidification period. For 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 20, the solidification period of PCM increases with the

decrease in the thermal conductivity of the shell. The difference in the solidification period

between 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 20 and 100 also increases with the decrease of the thermal conductivity of

the shell. This implies that the shell thickness has considerable effect on the solidification

time and its increase extends the solidification period of PCM when the shell has low thermal

conductivity. This suggests that the shell thickness affects the heat transfer behaviour of

MEPCMSs via influencing the solidification time inside microcapsules.

4.3. Effects of shell compositions

Comparative analysis is carried out among MF, MF-Cu and MF-Cu-Al microcapsules

with 𝑟𝑟𝑖𝑖 =5 µm and 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 =100. The differences of shell compositions are reflected in the

differences in the Young’s moduli as described above. Fig. 6(a) displays the evolution of

differences between external and internal pressures with solidification of PCM under

different shell compositions. Among the three kinds of microcapsules, the MF-Cu-Al

microcapsule with highest Young’s modulus has fastest increasing rate of pressure difference.

The pressure differences for the MF and MF-Cu microcapsules reach the critical buckling

pressures at 𝑓𝑓 ∗ = 0.56 and at 𝑓𝑓 ∗ = 0.52, respectively. As a result, shell buckling occurs at the

corresponding positions for the two microcapsules. Compared with the MF microcapsule, the
18
MF-Cu microcapsule buckles at an earlier solidification stage because of a higher rate of

increase in the pressure difference, although it has a higher critical pressure. Because the MF-

Cu-Al microcapsule possesses higher critical buckling pressure, the shell buckling does not

take place during solidification. This implies that the increase of Young’s modulus augments

the critical pressure and therefore avoids shell buckling. The reason why there are different

morphologies between the MF-Cu and MF-Cu-Al microcapsules in Figs. 1(b, c) can easily be

obtained from the above analysis. The calculation analysis and experimental test indicate that

using composite shell to elevate the Young’s modulus is a feasible method to avoid buckling.

As shown in Fig. 6(b), the MF-Cu-Al microcapsule has the lowest freezing point at the

same solidification stage among the three types of microcapsules before their freezing points

become constant, which is determined by its internal pressure. The lowered freezing point

will slow down the solidification process. The freezing point persistently decreases during the

whole solidification process for the MF microcapsule, rather than remaining constant at the

later solidification stage similar to the situation of other microcapsules. Fig. 6(c) indicates

that the MF-Cu-Al microcapsule exhibits the highest average value of latent heat among the

three types of microcapsules. This results in an increase of cold energy stored by releasing

latent heat. Fig. 6(d) shows that the shell compositions have a slight effect on the

solidification time. The MF-Cu-Al microcapsule exhibits the slowest solidification process,

which coincides with the situation of freezing point as shown in Fig. 6(b).

4.4. Effects of microcapsule size

The solidification processes are comparatively analysed for the MF-Cu-Al

microcapsules with different sizes at 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 =100. It is obvious that the solidification period of

PCM increases as the microcapsule size increases. The exact solidification periods for

microcapsules with different sizes are shown in Fig. 7(a). The solidification periods for

microcapsules with 𝑟𝑟𝑖𝑖 = 5, 25 and 50 are 0.5 s, 2.1 s and 4.5 s, respectively. The solidification
19
period will directly influence the charging efficiency of cold energy. In addition, the starting

time for solidification is independent of the microcapsule size. The stored energy calculated

by Eqn. (28) is also examined for microcapsules with different sizes at the same 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 as

shown in Figs. 7(b, c). The stored latent energy increases linearly regardless of microcapsule

size, while the stored sensible energy is nearly constant because of the narrow temperature

changes during the solidification process.

4.5. Critical pressure and buckling mode

According to the buckling theory, the condition for shell buckling is that the difference

between external and internal pressures should be greater than the critical buckling pressure.

The external pressure is atmospheric pressure and the minimum internal pressure is zero, so

that the maximum pressure difference is atmospheric pressure for the microcapsules. For

prescribed shell materials or compositions, the critical buckling pressure is only dependent on

𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 in accordance with Eq. (29). Fig. 8(a) depicts the variation of critical buckling pressure

with respect to 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 for the MF, MF-Cu and MF-Cu-Al shells. The figure shows that the

critical pressure decreases with an increase in 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 and the MF-Cu-Al shell has highest

critical pressure among the three shells at the same 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 . When 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 ≥ 72, 92 and 160 for the

MF, MF-Cu and MF-Cu-Al shells respectively, the critical pressure decreases below

atmospheric pressure (≈0.1 MPa), indicating that the pressure differences will exceed the

critical pressure at a certain solidification stage and buckling will thus occur. Thus, the

conditions for avoiding buckling during the solidification process for MF, MF-Cu and MF-

Cu-Al shells are 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 < 72, 92 and 160, respectively. This implies that the MF-Cu

microcapsule is easier to buckle than the MF-Cu-Al one, which explains the morphology

difference between the two kinds of microcapsules as shown in Figs 1(b, c).

As shown in Fig. 8(b), the buckling mode number obtained by solving Eqn. (32)

increases with 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 and coincides with the calculation of Sato et al. [34]. This suggests that
20
the microcapsules present different buckling modes at different 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 , because they are

corresponding to the mode numbers. Figs. 9(a, b) demonstrate the buckling modes of MF-Cu

microcapsules calculated from Eqn. (30) at 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 140 and 14, of which the mode numbers

are 23 and 6, respectively. The buckling waves on one side of symmetry axis are labeled with

numbers. The buckling deformation is asymmetric about the equator for the odd mode

number as shown in Fig. 9(a), whilst it is symmetric about the equator for the even mode

number as shown in Fig. 9(b).

Fig. 9(c) displays the cross-sectional Cryo-SEM image of a buckled MF-Cu

microcapsule obtained experimentally. The buckling on one side of the axis of symmetry

vanished when the microcapsules were cut. The calculated buckling deformation and mode

number at 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 140 as shown in Fig. 9(a) agree well with the image observed

experimentally as shown in Fig. 9(c). The condition for buckling is satisfied for the MF-Cu

microcapsule of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 140, which means that the prediction about buckling is reasonable.

From the Cryo-SEM image in Fig. 9(c) the shell thickness cannot be obtained, but it can be

derived from the value of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 in Fig. 9(a). The resulting shell thickness is about 36 nm.

The cross-sectional TEM image of a buckled MF-Cu microcapsule obtained through

experiments is shown in Fig. 9(d). It can be observed from this figure that the microcapsule

diameter is around 11 μm and the shell thickness is 386 nm. The value of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 for the

microcapsule in Fig. 9(d) is about 14, which is the same as that in Fig. 9 (b). By Comparing

Figs. 9(b) and 9(d), it is easily found that the predicted buckling deformation and mode

number are highly consistent with the experimental observations at the same 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 . This

further confirms the validity of the proposed buckling model in Section 3.5. It should be

pointed out that the buckling condition for a microcapsule of 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 = 14 under atmospheric

pressure is 𝐸𝐸 ≤ 23 MPa, which is smaller than the adopted value of 𝐸𝐸 for the MF-Cu

microcapsule above. It can be inferred that the small value of 𝐸𝐸 is a consequence of a poor

quality of polymerization and electroless plating, which is likely to randomly occur during

21
the fabrication process of microcapsules.

4.6. Energy storage capacity of MEPCMSs

A cold storage unit is the essential component for the PTES system [7, 45] and can also

be applied to improve the round trip efficiency of the LAES system [8, 46]. Both McTigue et

al. [7] and Sciacovelli et al. [8] used a packed bed as the cold storage unit for the PTES and

LAES systems, respectively. In the packed bed, the storage medium is spherical pebbles

made of Fe3O4 with an average void fraction of 0.35 for the PTES system, while it is

spherical quartzite rocks with an average void fraction of 0.38 for the LAES system. In this

study, it is attempted to use a tank containing MEPCMS as the cold storage unit instead of the

packed bed in the cryogenic temperature region. The selected MEPCMS consists of a carrier

liquid and the DJ microcapsules with an assumed volumetric concentration of 20%. Because

the working temperature regions in the cold storage units of the two systems are different, the

R22 and propane are adopted as the carrier liquid for PTES and LAES systems, respectively.

The thermophysical properties of R22 and propane come from the commercial software

REFPROP 8.0 developed by National Institute of Standards and Technology (NIST). The

equivalent thermophysical properties of MEPCMSs were calculated by the method shown in

[47].

Table 3 compares the energy storage capacity between the typical packed beds and the

selected MEPCMSs at the same pressure and temperature conditions for the PTES and LAES

systems, respectively. The calculation of energy storage density only considered the static

energy balance of heat transfer. For the PTES system, the mass-based energy storage density

of the MEPCMS is about 2.4 times that of the packed bed and the volume-based energy

storage density of the MEPCMS is 5.2 MJ/m3 greater than that of the packed bed. For the

LAES system, the MEPCMS has about 3.8 times the mass-based energy storage density and

around 1.8 times the volume-based energy storage density of the packed bed. Because of a

temperature gradient existing along the packed bed in actual applications, not all of the cold
22
storage medium can be fully utilized [48, 49]. Therefore, the difference in the energy storage

density between the MEPCMS and packed bed will be further enlarged in actual applications.

In view of its higher energy storage capacity, using the MEPCMS as a cold storage medium

can result in more compact PTES and LAES systems.

5. Conclusions

A numerical model was established to describe the thermo-mechanical behavior of

spherical microcapsules containing PCM for cryogenic-temperature cold storage. The model

combines energy conservation equations, pressure-dependent solid-liquid equilibria, Lamé’s

equations and buckling theory. During the charging process of cold energy, the PCM

solidification results in volume shrinkage and the pressure inside the microcapsule thus

decreases. The main consequences of this depressurisation are a progressive augmentation of

the difference between external and internal pressures, a progressive diminution of freezing

point of the PCM and a progressive increase of its latent heat. When the pressure difference

increases to the critical buckling pressure, shell buckling will occur.

The influences of shell thickness and compositions on the thermo-mechanical

behaviour of a microcapsule during the PCM solidification process were studied on the basis

of the developed model. The decrease of shell thickness slows down the decrease of internal

pressure, and thus diminishes the changing rates of freezing point and latent heat. The shell

thickness has little effect on the solidification time of PCM when the shell material has a high

thermal conductivity. When the critical pressure reduces below 0.1 MPa with the decrease in

shell thickness at the same core radius, shell buckling will occur during the solidification

process of PCM; Further decreasing shell thickness leads to the occurrence of shell buckling

at an earlier solidification stage. The Young’s modulus of the shell is increased by embedding

Al2O3 nanoparticles into or electroless Cu plating on the surface of MF the shell. The increase

23
in the Young’s modulus of the shell speeds up the variations of internal pressure, freezing

point and latent heat and thus leads to a slower solidification process, while enhancing the

resistance to buckling. The conditions for avoiding buckling during the solidification process

for MF, MF-Cu and MF-Cu-Al shells are that the ratio of core radius to shell thickness is less

than 72, 92 and 160, respectively. The buckling mode predicted by the thermo-mechanical

model is highly consistent with the experimental observations and the mode number increases

with the ratio of core radius to shell thickness. The model can be applied to predict the

conditions of avoiding shell buckling as well as the shell thickness or Young’s modulus based

on observed buckling mode.

The shell buckling and solidification time of PCM microcapsules are crucial to the heat

transfer behavior of MEPCMSs and charging efficiency of cold energy. The comparative

analysis indicates that MEPCMSs have higher cold energy storage capacity than packed

pebble beds in PTES and LAES systems. The present study can provide significant guidance

for precisely tailoring the key parameters of PCM microcapsules to enable successful and

high-efficiency applications of MEPCMSs for cold storage.

Acknowledgement

The authors gratefully acknowledge the financial support of the Engineering and

Physical Sciences Research Council (EPSRC) of the United Kingdom under grants

EP/N000714/1 and EP/N021142/1.

24
References

[1] Guizzi GL, Manno M, Tolomei LM, Vitali RM. Thermodynamic analysis of a liquid air energy

storage system. Energy 2015;93:1639-47.

[2] Peng H, Shan X, Yang Y, Ling X. A study on performance of a liquid air energy storage system

with packed bed units. Appl Energy 2018;211:126-35.

[3] Thess A. Thermodynamic Efficiency of Pumped Heat Electricity Storage. Phys Rev Lett

2013;111:110602.

[4] Morgan R, Nelmes S, Gibson E, Brett G. Liquid air energy storage – Analysis and first results

from a pilot scale demonstration plant. Appl Energy 2015;137:845-53.

[5] She X, Peng X, Nie B, Leng G, Zhang X, Weng L, et al. Enhancement of round trip efficiency of

liquid air energy storage through effective utilization of heat of compression. Appl Energy

2017;206:1632-42.

[6] White A, Parks G, Markides CN. Thermodynamic analysis of pumped thermal electricity storage.

Appl Therm Eng 2013;53:291-8.

[7] McTigue JD, White AJ, Markides CN. Parametric studies and optimisation of pumped thermal

electricity storage. Appl Energy 2015;137:800-11.

[8] Sciacovelli A, Vecchi A, Ding YL. Liquid air energy storage (LAES) with packed bed cold

thermal storage – From component to system level performance through dynamic modelling. Appl

Energy 2017;190:84-98.

[9] Peng H, Li R, Ling X, Dong H. Modeling on heat storage performance of compressed air in a

packed bed system. Appl Energy 2015;160:1-9.

[10] Peng H, Yang Y, Li R, Ling X. Thermodynamic analysis of an improved adiabatic compressed

air energy storage system. Appl Energy 2016;183:1361-73.

[11] Ma X, Omer SA, Riffat SB, Zhang W. Investigation of energy transportation capability of a

phase change slurry through a cold storage-cooling coil system. Int J Energy Res 2009;33:999-1004.

[12] Griffiths PW, Eames PC. Performance of chilled ceiling panels using phase change material

slurries as the heat transport medium. Appl Therm Eng 2007;27:1756-60.

25
[13] Chiu C-H. Commercial and technical considerations in the development of offshore liquefaction

plant. 23rd World Gas Conference. Amsterdam, Netherlands; 2006. p. 2485-96.

[14] Li Y, Wang X, Ding Y. An optimal design methodology for large-scale gas liquefaction. Appl

Energy 2012;99:484-90.

[15] Delgado M, Lázaro A, Mazo J, Zalba B. Review on phase change material emulsions and

microencapsulated phase change material slurries: Materials, heat transfer studies and applications.

Renew Sust Energ Rev 2012;16:253-73.

[16] Qiu Z, Ma X, Zhao X, Li P, Ali S. Experimental investigation of the energy performance of a

novel Micro-encapsulated Phase Change Material (MPCM) slurry based PV/T system. Appl Energy

2016;165:260-71.

[17] Wang T, Wang S, Luo R, Zhu C, Akiyama T, Zhang Z. Microencapsulation of phase change

materials with binary cores and calcium carbonate shell for thermal energy storage. Appl Energy

2016;171:113-9.

[18] Alva G, Huang X, Liu L, Fang G. Synthesis and characterization of microencapsulated myristic

acid–palmitic acid eutectic mixture as phase change material for thermal energy storage. Appl Energy

2017;203:677-85.

[19] Lopez J, Caceres G, Palomo Del Barrio E, Jomaa W. Confined melting in deformable porous

media: A first attempt to explain the graphite/salt composites behaviour. Int J Heat Mass Transf

2010;53:1195-207.

[20] Pitié F, Zhao CY, Caceres G. Thermo-mechanical analysis of ceramic encapsulated phase-

change-material (PCM) particles. Energy Environ Sci 2011;4:2117-24.

[21] Parrado C, Cáceres G, Bize F, Bubnovich V, Baeyens J, Degrève J, et al. Thermo-mechanical

analysis of copper-encapsulated NaNO3–KNO3. Chem Eng Res Des 2015;93:224-31.

[22] Zhao W, Neti S, Oztekin A. Heat transfer analysis of encapsulated phase change materials. Appl

Therm Eng 2013;50:143-51.

[23] Giro-Paloma J, Barreneche C, Martínez M, Šumiga B, Fernández AI, Cabeza LF. Mechanical

response evaluation of microcapsules from different slurries. Renew Energy 2016;85:732-9.

26
[24] Su J-F, Wang X-Y, Dong H. Micromechanical properties of melamine–formaldehyde

microcapsules by nanoindentation: Effect of size and shell thickness. Mater Lett 2012;89:1-4.

[25] Alam TE, Dhau JS, Goswami DY, Stefanakos E. Macroencapsulation and characterization of

phase change materials for latent heat thermal energy storage systems. Appl Energy 2015;154:92-101.

[26] Lashgari S, Arabi H, Mahdavian AR, Ambrogi V. Thermal and morphological studies on novel

PCM microcapsules containing n-hexadecane as the core in a flexible shell. Appl Energy

2017;190:612-22.

[27] Liu M-J, Fan L-W, Zhu Z-Q, Feng B, Zhang H-C, Zeng Y. A volume-shrinkage-based method

for quantifying the inward solidification heat transfer of a phase change material filled in spherical

capsules. Appl Therm Eng 2016;108:1200-5.

[28] Assis E, Ziskind G, Letan R. Numerical and experimental study of solidification in a spherical

shell. J Heat Transf 2009;131:024502.

[29] Tarnai T. Buckling patterns of shells and spherical honeycomb structures. Comput Math Appl

1989;17:639-52.

[30] Yin J, Cao Z, Li C, Sheinman I, Chen X. Stress-driven buckling patterns in spheroidal core/shell

structures. Proceedings of the National Academy of Sciences 2008;105:19132-5.

[31] Ru CQ. Buckling of empty spherical viruses under external pressure. J Appl Phys

2009;105:124701.

[32] Vliegenthart GA, Gompper G. Compression, crumpling and collapse of spherical shells and

capsules. New J Phys 2011;13:045020.

[33] Timoshenko SP, Gere JM. Theory of elastic stability: McGraw-Hill; 1961.

[34] Sato M, Wadee MA, Iiboshi K, Sekizawa T, Shima H. Buckling patterns of complete spherical

shells filled with an elastic medium under external pressure. Int J Mech Sci 2012;59:22-30.

[35] ACM Comput. Surv.Long Y, York D, Zhang Z, Preece JA. Microcapsules with low content of

formaldehyde: preparation and characterization. J Mater Chem 2009;19:6882-7.

27
[36] Hechtel K. Design considerations for the use of plastic materials in cryogenic environments.

Curbell Plastics, Inc.; 2014. https://www.curbellplastics.com/Research-Solutions/Technical-

Resources/Technical-Resources/Plastic-Materials-in-Cryogenic-Environments.

[37] Jiang X, Luo R, Peng F, Fang Y, Akiyama T, Wang S. Synthesis, characterization and thermal

properties of paraffin microcapsules modified with nano-Al2O3. Appl Energy 2015;137:731-7.

[38] Niu X-W, Sun Y-M, Ding S-N, Chen C-C, Song B, Xu H-B, et al. Synthesis of enhanced urea–

formaldehyde resin microcapsules doped with nanotitania. J Appl Polym Sci 2012;124:248-56.

[39] Voller VR, Cross M, Markatos NC. An enthalpy method for convection/diffusion phase change.

Int J Numer Methods Eng 1987;24:271-84.

[40] Kármán TV, Tsien HS. The buckling of spherical shells by external pressure. J Aeronautical Sci

1939;7:43-50.

[41] Pan B, Cui W. An overview of buckling and ultimate strength of spherical pressure hull under

external pressure. Mar Struct 2010;23:227-40.

[42] Dowtherm J heat transfer fluid: Product technical data. The Dow Chemical Company; 1997.

http://www.dow.com/heattrans/.

[43] Harper CA. Modern Plastics Handbook. New York: McGraw-Hill; 1999.

[44] Liu M. Understanding the mechanical strength of microcapsules and their adhesion on fabric

surfaces [Doctoral Dissertation]. United Kingdom: University of Birmingham; 2010.

[45] Frate GF, Antonelli M, Desideri U. A novel Pumped Thermal Electricity Storage (PTES) system

with thermal integration. Appl Therm Eng 2017;121:1051-8.

[46] Li Y, Cao H, Wang S, Jin Y, Li D, Wang X, et al. Load shifting of nuclear power plants using

cryogenic energy storage technology. Appl Energy 2014;113:1710-6.

[47] Zeng R, Wang X, Chen B, Zhang Y, Niu J, Wang X, et al. Heat transfer characteristics of

microencapsulated phase change material slurry in laminar flow under constant heat flux. Appl

Energy 2009;86:2661-70.

[48] Araki H, Nakabaru M, Chino K. Simulation of heat transfer in the cool storage unit of a liquid–

air energy storage system. Heat Transf—Asian Res 2002;31:284-96.

28
[49] Chai L, Liu J, Wang L, Yue L, Yang L, Sheng Y, et al. Cryogenic energy storage characteristics

of a packed bed at different pressures. Appl Therm Eng 2014;63:439-46.

29
 (a) (b)

12

Volume fraction (%)

10
8
6
4
2
0
10 μm
1 10 100 1000
(c) Microcapsule size (𝜇𝜇m) (d)

(e)

(f)

Fig. 1 Microcapsules: (a,b,c) Cryo-SEM images of MF, MF-Cu and MF-Cu-Al microcapsules; (d) Size

distribution; (e,f) EDS spectra of MF-Cu and MF-Cu-Al microcapsules.

30
 Liquid PCM

Solid PCM

Shell

Fig. 2 Geometry of the spherical microcapsule containing PCM.

31
 Dimensionless Temperature

0.5 1

0.3 0.9

Liquid Fraction
0.1 0.8

-0.1 0.7
Reference [39] Reference [39]
-0.3 Current study 0.6 Current study
-0.5 0.5
0 0.2 0.4 0.6 0.8 1 0 100 200 300 400 500
Dimensionless Distance Time (s)

(a) (b)

600 600 110

Melting Temperature (K)

Reference [19]
500 Current study

Latent Heat (J/g)

Pressure (MPa)

100
400 550
300 90
200 500
80
100 Reference [19]
Current study
0 450 70
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Liquid Fraction Liquid Fraction

(c) (d)

Fig. 3 Comparison with references: (a) Temperature profile at t = 500 s; (b) Solidification rate; (c)

Pressure variation with solidification; (d) Variations of melting temperature and latent heat with

solidification.

32
 0.12 0.2

Freezing Point Variation (K)

Pressure Difference (MPa)

ri/a=20
0.10
ri/a=50
0.08 0.0
ri/a=100
Buckling occurs
0.06 ri/a=120
-0.2 Buckling occurs
0.04 ri/a=20
0.02 ri/a=50
-0.4
ri/a=100
0.00
ri/a=120
-0.02 -0.6
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Solid Fraction Solid fraction

(a) (b)

16 12.5
Latent Heat Variation (J/kg)

ri/a=20
12 12.4 ri/a=50
Buckling occurs 12.3 ri/a=100
Time (s)

8 ri/a=120
12.2
4 ri/a=20
ri/a=50 12.1
0 ri/a=100 12.0
ri/a=120
-4 11.9
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Solid Fraction Solid Fraction

(c) (d)

Fig. 4 Effects of shell thickness during solidification: (a) Evolution of pressure differences; (b) Evolution

of freezing point variation; (c) Evolution of latent heat variation; (d) Time with respect to solid fraction.

Critical position of buckling is labelled.

33
 12.8
ri/a=100, λc=0.50 Wm K
-1 -1

ri/a=100, λc=0.15 Wm K
-1 -1

ri/a=100, λc=0.05 Wm K
-1 -1

ri/a=20, λc=0.50 Wm K
-1 -1
12.6
ri/a=20, λc=0.15 Wm K
-1 -1

ri/a=20, λc=0.05 Wm K
-1 -1

12.4
Time (s)

12.2

12.0

11.8
0 0.2 0.4 0.6 0.8 1
Solid Fraction

Fig. 5 Effect of shell thickness with different thermal conductivities on solidification time.

34
 0.12 0.2

Freezing Point Variation (K)

Pressure Difference (MPa)

MF
0.10
MF-Cu
0.08 0.0
MF-Cu-Al
0.06 Buckling occurs
Buckling occurs -0.2
0.04
MF
0.02 -0.4
MF-Cu
0.00 MF-Cu-Al
-0.02 -0.6
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Solid Fraction Solid Fraction

(a) (b)

16 12.5
Latent Heat Variation (J/kg)

12 12.4

12.3
Time (s)

8
Buckling occurs 12.2
4
12.1 MF
MF
0 MF-Cu 12.0 MF-Cu
MF-Cu-Al MF-Cu-Al
-4 11.9
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Solid Fraction Solid Fraction

(c) (d)

Fig. 6 Effects of shell compositions during solidification: (a) Evolution of difference pressures; (b)

Evolution of freezing point variation; (c) Evolution of latent heat variation; (d) Time with respect to solid

fraction. Critical position of buckling is labelled.

35
 17

16 ri=5 µm
ri=25 µm
15 ri=50 µm
Time (s)

14

13

12

11
0 0.2 0.4 0.6 0.8 1
Solid Fraction

(a)

10 14
12
8
Energy (10 J)

Energy (10 J)

10
-7

-5

6 8
Latent Energy
Sensible Energy 6 Latent Energy
4 Sensible Energy
Total Energy
4 Total Energy
2
2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Solid Fraction Solid Fraction

(b) (c)

Fig. 7 Effects of microcapsule size during solidification: (a) Time with respect to solid fraction; (b,c)

Evolution of the stored energy at 𝑟𝑟𝑖𝑖 = 5 μm and 25 μm.

36
 0.8 Buckling Mode Number 30
Critical Pressure (MPa)

0.7
MF 25
0.6
MF-Cu 20
0.5 MF-Cu-Al
0.4 15
0.3
10
0.2 Current study
5 Reference [27]
0.1
0 0
60 80 100 120 140 160 10 40 70 100 130 160
ri/a ri/a

(a) (b)

Fig. 8 (a) Critical buckling pressure with respect to 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎 for different shell compositions; (b) Buckling

mode numbers with respect to 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎.

37
 ri/a=140 ri/a=14
11 12 3
10
9
8
7
2
6
5
4
3
2 1 µm 1 µm
6 4 2 0 2 4 6 6 4 2 0 2 4 6

(a) (b)

8 9 10 11
7 12
6
5
3
4

3
2
2
1
1

(c) (d)

Fig. 9 Calculated buckling modes at 𝑟𝑟𝑖𝑖 ⁄𝑎𝑎= 140 (a) and 14 (b); Cross-sectional images of buckling MF-Cu

microcapsules under Cryo-SEM (c) and TEM (d). The waves are partially labeled with numbers and the

dot-dash line represents symmetry axis. The buckling on one side of symmetry axis vanished when the

microcapsules were cut.

38
 Table 1 Properties of DJ used in simulations [42].
Properties Symbol Value Unit
DJ in Density 𝜌𝜌𝑙𝑙0 931.3 kg·m-3
liquid Specific heat 𝑐𝑐𝑝𝑝𝑝𝑝 1584 J·kg-1·K-1
state Compressibility 𝛽𝛽𝑙𝑙 1.72×10-10 Pa-1
Thermal expansion 𝛼𝛼𝑙𝑙 8.33×10-4 K-1
Thermal 𝜆𝜆𝑙𝑙 0.148 W·m-1·K-1
conductivity
Surface tension 𝛾𝛾𝑙𝑙 0.028 N·m-1
DJ in Density 𝜌𝜌𝑠𝑠0 950.0 kg·m-3
solid Specific heat 𝑐𝑐𝑝𝑝𝑝𝑝 1500 J·kg-1·K-1
state Compressibility 𝑐𝑐𝑝𝑝𝑝𝑝 0 Pa-1
Thermal expansion 𝛼𝛼𝑠𝑠 0 K-1
Thermal 𝜆𝜆𝑠𝑠 0.152 W·m-1·K-1
conductivity
L↔S Freezing temperature 𝑇𝑇𝑓𝑓0 -81 ℃
at 𝑃𝑃0
Latent heat at 𝐿𝐿𝑓𝑓0 80 kJ·kg-1
�𝑇𝑇𝑓𝑓0, 𝑃𝑃0 �

39
 Table 2 Theoretical properties of MF [43].
Properties Symbol Value Unit
Density 𝜌𝜌𝑐𝑐 1500 kg·m-3
Specific heat 𝑐𝑐𝑝𝑝𝑝𝑝 1200 J·kg-1·K-1
Thermal expansion 𝛼𝛼𝑐𝑐 6.0×10-5 K-1
Thermal 𝜆𝜆𝑐𝑐 0.5 W·m-1·K-1
conductivity
Young’s modulus 𝐸𝐸𝑐𝑐 7.0×109 Pa
Poisson’s ratio 𝜈𝜈𝑐𝑐 0.29 --

40
Table 3 Comparison of energy storage capacity between a typical packed bed and the MEPCMS for PTES

and LAES. The storage materials are Fe3O4 (density 5175 kg/m3) and quartzite (density 2560 kg/m3) in the

packed beds for PTES and LAES systems, respectively. The carrier liquids are R22 and propane in the

MEPCMSs for PTES and LAES systems, respectively.

Unit Packed bed MEPCMS Packed bed MEPCMS

(PTES) (PTES) (LAES) (LAES)
Pressure bar 1.05 1.05 1.49 1.49
Temperature range K 123~223 123~223 92.7~192.7 92.7~192.7
Void fraction -- 0.35 -- 0.38 --
Average specific heat J·kg-1·k-1 520 1125 541 1861
Equivalent latent heat kJ·kg-1 -- 11.1 -- 21.8
Mass-based energy kJ·kg-1 52.0 123.6 54.1 207.9
storage density
Volume-based energy MJ·m-3 174.9 180.1 85.9 154.1
storage density

41