materials
Article
Cryogenic Reliability Evaluation of Glass Fabric–Reinforced
Composites Using Novel Slip-Prevention Method
Yong-Cheol Jeong 1 , Dong-Ha Lee 1 , Seul-Kee Kim 2 , Jeong-Hyeon Kim 2, * and Jae-Myung Lee 1,2, *
Citation: Jeong, Y.-C.; Lee, D.-H.;
Kim, S.-K.; Kim, J.-H.; Lee, J.-M.
Cryogenic Reliability Evaluation of
Glass Fabric–Reinforced Composites
Using Novel Slip-Prevention Method.
Materials 2023, 16, 121. https://
doi.org/10.3390/ma16010121
Academic Editor: Milena Pavlíková
Received: 11 November 2022
Revised: 18 December 2022
Accepted: 19 December 2022
Published: 22 December 2022
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Materials 2023, 16, 121. https://doi.org/10.3390/ma16010121
1 Department of Naval Architecture and Ocean Engineering, Pusan National University,
Busan 46241, Republic of Korea
2 Hydrogen Ship Technology Center, Pusan National University, Busan 46241, Republic of Korea
* Correspondence: jeonghkim@pusan.ac.kr (J.-H.K.); jaemlee@pusan.ac.kr (J.-M.L.)
Abstract: Glass fabric–reinforced composites are the main insulating material components of the
secondary barrier of cargo containment systems (CCSs), because they prevent liquefied natural gas
(LNG) leakage during transport. Nevertheless, it is difficult to evaluate the material performance
of glass fabric–reinforced composites at cryogenic temperatures (−163 ◦ C) because it takes approxi-
mately 7 days to prepare the test specimens and because the slip-based test frequently fails. Although
glass fabric–reinforced composites for the secondary barrier of LNG CCSs show various structural
vulnerabilities, enhancing their material performance is significantly limited owing to the reasons
mentioned above. This study evaluated the structural vulnerabilities and failure characteristics of
glass fabric–reinforced composites by using the slip-prevention test method to determine the level
difference and adhesive vacancies. The failure surface and the thermal expansion of the composites
were also observed, to analyze their mechanical characteristics. By adopting our proposed test proce-
dure, the failure rate of the experiment decreased by approximately 80%, and the sample preparation
time for manufacturing was significantly shortened, to 1 day.
Keywords: cryogenic reliability; LNG cargo containment system; flexible secondary barrier
1. Introduction
Natural gas has received significant attention as an environmentally friendly fuel
because its combustion emits less carbon dioxide than conventional fossil fuels do [1].
Hence, its demand as a fuel has gradually increased in many industries, such as automotive
and shipbuilding. Liquefied natural gas (LNG), which is natural gas cooled to a liquid
state, at approximately −163 ◦ C, is the most effective and preferred method for shipping
and storing natural gas, because its volume is approximately reduced 600 times compared
with its gaseous state. Hence, structural reliability and thermal reliability at cryogenic
temperatures are the most important functional requirements for LNG storage systems, as
they require complex multilayered insulation systems to maintain temperatures such as
−163 ◦ C.
There are two main structural classifications for LNG cargo containment systems
(CCSs): independent and membrane [2,3]. Independent tanks are self-supporting and
separate from the hull, whereas membrane tanks are not autonomous and consist of a
thin layer (membrane) supported by insulation [4,5]. Membrane tanks are preferred in the
market for being more cost-effective and having significantly larger storage capacity than
their independent counterparts. Depending on the insulation system adopted for the CCS,
membrane tank designs can be classified into two types, Mark III and NO96; the former
type is widely employed for LNG CCSs because of its higher efficiency [6].
Figure 1 schematically illustrates a Mark III–type LNG insulation system. Its surface
is covered with a thin layer of cryogenic temperature-resistant 304 L austenitic stainless
steel, known as the primary barrier, that prevents LNG leakage. Further, the corrugated
https://www.mdpi.com/journal/materials
Materials 2022, 15, x FOR PEER REVIEW 2 of 16

Materials 2023, 16, 121 2 of 15

steel, known as the primary barrier, that prevents LNG leakage. Further, the corrugated
structure of the plate prevents thermal contraction and expansion [7]. The inside of the
structure of the plate prevents thermal contraction and expansion [7]. The inside of the
membrane panel contains two-layered reinforced polyurethane foam (R-PUF), embedded
membrane panel contains two-layered reinforced polyurethane foam (R-PUF), embedded
with a secondary barrier [8,9] to seal the tank against gas leakage for 15 days after the
with a secondary barrier [8,9] to seal the tank against gas leakage for 15 days after the
primary barrier has been fractured.
primary barrier has been fractured.

Figure 1. 1.
Figure Mark III–type
Mark LNG
III–type insulation
LNG system.
insulation system.

Figure
Figure 2a2a illustrates
illustrates the
the schematic
schematic composition
composition of of
thethe secondary
secondary barrier
barrier in in Mark
Mark III–
III–
type LNG insulation systems. There are two main concerns: the adhesive
type LNG insulation systems. There are two main concerns: the adhesive vacancy and the vacancy and the
level difference. Adhesive vacancy can occur from the bonding process,
level difference. Adhesive vacancy can occur from the bonding process, whereas a level whereas a level
differencecan
difference canoccur
occurfrom
from differences
differencesin inmanufacturing
manufacturingtolerances during
tolerances duringthethe
manufacturing
manufac-
process
turing of theofLNG
process insulation
the LNG panel.
insulation The The
panel. secondary
secondarybarrier contains
barrier flexible
contains andand
flexible rigid
sections, respectively known as the flexible secondary barrier (FSB),
rigid sections, respectively known as the flexible secondary barrier (FSB), comprising comprising glass
fabric,
glass rubber,
fabric, andand
rubber, aluminum
aluminum foil,foil,
andandthe the
rigid secondary
rigid secondarybarrier (RSB),
barrier (RSB),comprising
comprisingglass
fabric and aluminum [10,11]. The flexibility of the secondary barrier
glass fabric and aluminum [10,11]. The flexibility of the secondary barrier is required be-is required because
of itsofsubjection
cause to thermal
its subjection loadsloads
to thermal under cryogenic
under temperatures
cryogenic and bending
temperatures and bending loadsloads
under
hogging and sagging. The bottom insulation panel is typically attached
under hogging and sagging. The bottom insulation panel is typically attached to the RSB to the RSB with
with a 30 mm gap to allow for thermal deformation. Subsequently, each RSB is attachedto
a 30 mm gap to allow for thermal deformation. Subsequently, each RSB is attached
the FSB with polyurethane (PU) adhesive, and the FSB is attached to the top bridge pad
to the FSB with polyurethane (PU) adhesive, and the FSB is attached to the top bridge pad
with epoxy adhesive. The R-PUF and FSB components exhibit a level difference owing to
with epoxy adhesive. The R-PUF and FSB components exhibit a level difference owing to
their disparate manufacturing tolerances during the manufacturing process of the LNG
their disparate manufacturing tolerances during the manufacturing process of the LNG
insulation panel [12,13], which is defined as uneven and curved parts of up to 3 mm between
insulation panel [12,13], which is defined as uneven and curved parts of up to 3 mm be-
the insulation panels, owing to the nonuniform flatness of the inner hull of the vessel.
tween the insulation panels, owing to the nonuniform flatness of the inner hull of the ves-
Figure 2b is the exaggeratedly drawn variation of the FSB caused by the level difference
sel. Figure 2b is the exaggeratedly drawn variation of the FSB caused by the level differ-
phenomenon. The occurrence of a level difference can likely weaken the performance of
ence phenomenon. The occurrence of a level difference can likely weaken the performance
the FSB because it can be bent, resulting in tensile load and stress concentration in the
of the FSB because it can be bent, resulting in tensile load and stress concentration in the
FSB. Hence, the mechanical properties of an FSB should be investigated on the basis of
FSB. Hence, the mechanical properties of an FSB should be investigated on the basis of
variations in the level difference.
variations in the
Figure level difference.
3 shows a schematic of the bonding-process-dependent adhesive vacancy of
the insulating system. Epoxy adhesive was applied to the top bridge pad in two or three
lines during the tank manufacturing process. The top bridge pad is bonded to the FSB by
using a pressure pad. Subsequently, the epoxy adhesive is spread onto the end of the FSB;
however, the epoxy adhesive does not reach the middle of the FSB, because of its swift
hardening or insufficient quantity. Hence, the mechanical properties of the FSB should be
investigated on the basis of variations in the adhesive filling ratio because various types of
vacancies might emerge between the FSB and the top bridge pad.
 Materials
Materials 2022, 15, x 2023, 16, 121REVIEW
FOR PEER 3 of 16 3 of 15

(a)

(a)
(b)

Figure 2. Schematic composition of (a) a secondary barrier in Mark III–type LNG insulation systems
and (b) variation of the mechanical characteristics caused by the level difference phenomenon.

Figure 3 shows a schematic of the bonding-process-dependent adhesive vacancy of

the insulating system. Epoxy adhesive was applied to the top bridge pad in two or three
lines during the tank manufacturing process. The top bridge pad is bonded to the FSB by
using a pressure pad. Subsequently, the epoxy adhesive is spread onto the end of the FSB;
however, the epoxy adhesive does not reach the middle of the FSB, because of its swift
hardening or insufficient quantity.(b) Hence, the mechanical properties of the FSB should be
investigated on
Figure 2. Schematic the basis
compositionof variations in the adhesive
of (a) a secondary filling
barrierIII–type
in Mark ratio because
III–type various types
LNG insulation
Figure 2. Schematic composition of (a) a secondary barrier in Mark LNG insulation systems systems
of vacancies
and might
(b) variation of theemerge between
mechanical the FSB and
characteristics theby
caused top
thebridge pad.
level difference phenomenon.
and (b) variation of the mechanical characteristics caused by the level difference phenomenon.

Figure 3 shows a schematic of the bonding-process-dependent adhesive vacancy of

the insulating system. Epoxy adhesive was applied to the top bridge pad in two or three
lines during the tank manufacturing process. The top bridge pad is bonded to the FSB by
using a pressure pad. Subsequently, the epoxy adhesive is spread onto the end of the FSB;
however, the epoxy adhesive does not reach the middle of the FSB, because of its swift
hardening or insufficient quantity. Hence, the mechanical properties of the FSB should be
investigated on the basis of variations in the adhesive filling ratio because various types
of vacancies might emerge between the FSB and the top bridge pad.

Figure3.3.Schematic
Figure Schematicillustration
illustration
of of
thethe manufacturing
manufacturing (bonding)-process-dependent
(bonding)-process-dependent adhesive
adhesive va-
vacancy
cancy
in MarkinIII–type
Mark III–type LNG insulation
LNG insulation systems.systems.

Over the past decades, several studies have reported on the mechanical properties
of glass fiber–reinforced composites, which are similar to the secondary barrier in LNG
CCSs [14–17]. Oh et al. investigated the fatigue performance of an LNG insulation system
containing a newly developed metallic secondary barrier, which showed improved tight-
ness and strength over that of conventional secondary barriers [18]. Additionally, many
researchers have reported the mechanical properties of adhesively bonded joints as sec-
ondary barriers [19,20]. For example, Lee et al. studied the debonding failure characteristics
Figure 3. of
Schematic illustration
epoxy- and of the manufacturing
polyurethane-based (bonding)-process-dependent
adhesives adhesive
at cryogenic temperatures. va-
Furthermore, they
cancy in Mark III–type LNG insulation systems.
 Over the past decades, several studies have reported on the mechanical properties of
glass fiber–reinforced composites, which are similar to the secondary barrier in LNG CCSs
[14–17]. Oh et al. investigated the fatigue performance of an LNG insulation system con-
taining a newly developed metallic secondary barrier, which showed improved tightness
and strength over that of conventional secondary barriers [18]. Additionally, many re-
Materials 2023, 16, 121 4 of 15
searchers have reported the mechanical properties of adhesively bonded joints as second-
ary barriers [19,20]. For example, Lee et al. studied the debonding failure characteristics
of epoxy- and polyurethane-based adhesives at cryogenic temperatures. Furthermore,
qualitatively
they investigated
qualitatively the debonding
investigated the debondingfailure damage
failure damagepattern fromfrom
pattern ambient to cryogenic
ambient to cry-
temperatures. Additionally, the material behavior of the multilaminate
ogenic temperatures. Additionally, the material behavior of the multilaminate bonding bonding system
was precisely analyzed on the basis of the stress–strain relationship
system was precisely analyzed on the basis of the stress–strain relationship [21]. [21].
Althoughseveral
Although severalresearchers
researchershave havestudied
studiedthethemechanical
mechanicalbehavior
behaviorofofthe the secondary
secondary
barrier, only a few studies have considered the effects of the level difference
barrier, only a few studies have considered the effects of the level difference on adhesively on adhesively
bondedjoints.
bonded joints.Yoon
Yoonetetal.al.reported
reported ononthe
theinfluence
influenceofofthethecuring
curingpressure
pressureandande-glass
e-glass
fabricdensity
fabric densityononthe
thebonding
bondingperformance
performanceofofadhesive
adhesivejoints
jointswith
witha a3 3mm
mmlevel
leveldifference
difference
at cryogenic temperatures [22,23]. Nevertheless, the simultaneous investigation of theof
at cryogenic temperatures [22,23]. Nevertheless, the simultaneous investigation me-the
mechanical performance of FSBs with a measurable level difference and adhesive vacancies
chanical performance of FSBs with a measurable level difference and adhesive vacancies
is completely absent. This study evaluates the structural vulnerabilities and failure charac-
is completely absent. This study evaluates the structural vulnerabilities and failure char-
teristics of a glass fabric–reinforced composite, where the influences of the level difference
acteristics of a glass fabric–reinforced composite, where the influences of the level differ-
and the adhesive vacancies were considered by employing a novel slip-prevention test
ence and the adhesive vacancies were considered by employing a novel slip-prevention
method. Parallelly, the failure surface and thermal expansion degree of the composites
test method. Parallelly, the failure surface and thermal expansion degree of the compo-
were meticulously observed to evaluate their mechanical performance.
sites were meticulously observed to evaluate their mechanical performance.
2. Experimental Section
2.Proposed
Experimental Section
Mechanical Test Method
2.1. Proposed Mechanical TestofMethod
The mechanical test the secondary barrier for the MARK III type was conducted
The mechanical
by using GTT M 3101 test(Gaztransport
of the secondary barrier forChevreuse,
& Technigaz, the MARK France).
III type was conducted
Figures 4 and 5
byshow
usingillustrations
GTT M 3101 of the preparations of the testing samples and experimentaland
(Gaztransport & Technigaz, Chevreuse, France). Figures 4 5
setup,
show illustrations
respectively. In theofconventional
the preparations of the testing
test method, samples
the grip sectionand experimental
of the textile/fibersetup, re-
is bonded
spectively. In thewith
to the plywood conventional test method,
a PU adhesive. the fractures
However, grip section of the textile/fiber
occasionally occur in theis bonded
adhesive
tojoints
the plywood with
at the grip a PU adhesive.
section prior to theHowever,
fracture fractures occasionally
of the specimen, occur in the
and slippage can adhesive
also occur
between
joints at thethe
gripjigsection
and the specimen
prior at the grip
to the fracture of thesection.
specimen,Particularly,
and slippage slippage
can alsooccurs
occurat
cryogenic
between thetemperatures, makingataccurate
jig and the specimen the grip measurements impossible.
section. Particularly, slippage Furthermore,
occurs at cry-the
curing time for the PU adhesive is 7 days, as stated in the conventional
ogenic temperatures, making accurate measurements impossible. Furthermore, the curing test standard.
time for the PU adhesive is 7 days, as stated in the conventional test standard.

Materials 2022, 15, x FOR PEER REVIEW 5 of 16

(a)

(b)

Schematic illustration
Figure 4. illustration
Figure 4. Schematic of the preparations
of the preparations of the
of the testing testing(a)
samples: samples: (a) conventional
conventional method method
and (b)method.
and (b) proposed proposed method.
 (b)
Materials 2023, 16, 121 5 of 15
Figure 4. Schematic illustration of the preparations of the testing samples: (a) conventional method
and (b) proposed method.

(a) (b)

FigureFigure 5. Schematic
5. Schematic illustration
illustration of the of the preparations
preparations of theofexperimental
the experimental
setup:setup: (a) conventional
(a) conventional
method and (b) proposed
method and (b) proposed method. method.

To overcome
To overcome these problems,
these problems, we proposed
we proposed a novel mechanical
a novel mechanical test method.testThe
method.
spec- The
imen is initially bonded to a glass fiber matrix by using a cyanoacrylate adhesive and then and
specimen is initially bonded to a glass fiber matrix by using a cyanoacrylate adhesive
then clamped with a machine bolt. Stiff clamping prevents the rotation of the specimen
clamped with a machine bolt. Stiff clamping prevents the rotation of the specimen parts
parts after matrix cracking, ensuring a nearly uniform stress distribution over the cross
after matrix cracking, ensuring a nearly uniform stress distribution over the cross section.
section. Furthermore, the curing time for the cyanoacrylate adhesive is only 1 day.
Furthermore, the curing time for the cyanoacrylate adhesive is only 1 day.
Stefanov et al. reported a dynamic mechanical analysis of carbon black-filled cyanoacry-
Stefanov et al. reported a dynamic mechanical analysis of carbon black-filled cy-
late adhesive bulk films. Cyanoacrylate adhesives require shorter curing times for reaching
anoacrylate adhesive bulk films. Cyanoacrylate adhesives require shorter curing times for
a solid state than do other adhesives, while possessing good thermal properties at cryo-
reaching a solid state than do other adhesives, while possessing good thermal properties
genic temperatures, because gelation during phase separation maintains the morphology
at cryogenic temperatures, because gelation during phase separation maintains the mor-
in place [24]. Therefore, in our work, compared with the conventional test method, the
phology in place [24]. Therefore, in our work, compared with the conventional test
premature fracturing at the adhesive joints and slippage of the specimen at the grip section
method, the premature fracturing at the adhesive joints and slippage of the specimen at
were attenuated because of the cyanoacrylate adhesive, owing to its excellent thermal
the grip section were attenuated because of the cyanoacrylate adhesive, owing to its ex-
properties and enhanced grip section. Furthermore, the conventional test method required
cellentanthermal
adhesiveproperties andofenhanced
curing time grip section.
7 days, whereas Furthermore,
the proposed the conventional
test method required only test1 day.
method required an adhesive curing time of 7 days, whereas the proposed test
Therefore, fractures did not occur at the adhesive joints under cryogenic conditions. More- method
required
over,only 1 day.
a large Therefore, fractures
deformation did not
did not occur occur
under at thetensile
a high adhesive
load,joints under in
resulting cryo-
accurate
genic fracture
conditions. Moreover, a large deformation did not occur under a
unit values, better strength, and a high success rate, confirmed by the high tensile load,
decrease
resulting in test
of the accurate
failurefracture
rate tounit
20%.values,
Thus, better strength, and
the experimental a high success
reproducibility rate,
and con-
reliability of
firmed by the decrease of the test failure rate to 20%. Thus, the experimental
our novel tensile test method and its ability to control the fracture location parameters reproduci-
bility were
and reliability
verified. of our novel tensile test method and its ability to control the fracture
location parameters
Experimentalweresamples
verified.were prepared to investigate the effects of the level difference
and the adhesive filling ratio of the secondary barrier. The FSB (Huntchinson, Paris,
France) and R-PUF panel (KANGRIM INSULATION, Changwon, Korea) were mechanically
bonded using an epoxy adhesive with a density of 120 kg/m3 to produce a conventional
insulation structure. All specimens for the mechanical tests were fabricated according to
the strip test standard ISO 1421 [25]. The gauge length was set to 30 mm, which is the
distance between the bottom insulation panels. The total length, width, and thickness
of the specimens were 100, 50, and 0.7 mm, respectively (Figure 6). The total bonding
lengths between the FSB and R-PUF panels from the edges of the gauge length were 15 mm
(adhesive filling ratio of 50%), 21 mm (adhesive filling ratio of 70%), and 30 mm (adhesive
filling ratio of 100%). To induce an adhesive filling vacancy, the middle of the FSB was
taped to prevent adherence to the R-PUF, with tape lengths of 15, 9, and 0 mm. The bonding
 ing to the strip test standard ISO 1421 [25]. The gauge length was set to 30 mm, which is
the distance between the bottom insulation panels. The total length, width, and thickness
of the specimens were 100, 50, and 0.7 mm, respectively (Figure 6). The total bonding
lengths between the FSB and R-PUF panels from the edges of the gauge length were 15
Materials 2023, 16, 121 mm (adhesive filling ratio of 50%), 21 mm (adhesive filling ratio of 70%), and 30 mm 6(ad- of 15
hesive filling ratio of 100%). To induce an adhesive filling vacancy, the middle of the FSB
was taped to prevent adherence to the R-PUF, with tape lengths of 15, 9, and 0 mm. The
bonding areaFSB
area of the of determined
the FSB determined the adhesive
the adhesive filling
filling ratio ratio
of the of the specimen.
specimen. Further-
Furthermore, the
more, the adhesive vacancy was positioned in the middle of the FSB to replicate
adhesive vacancy was positioned in the middle of the FSB to replicate actual performance actual
performance
when the topwhen
bridgethe topisbridge
pad pad
attached is attached
with withlines
two or three two ofor epoxy
three lines
glue.of epoxy
After glue.
bonding,
After bonding,was
the specimen the compressed
specimen was compressed
with with 10
10 kg of weight forkga of weight
week for the
to cure a week to cure theof
adhesiveness
adhesiveness
the epoxy. of the epoxy.

(a) (b)
Figure
Figure6.6.Shapes
Shapesand
anddimensions
dimensionsofofthe
themechanical
mechanicaltest
testspecimens:
specimens:(a)
(a)without
withoutan
anadhesive
adhesivefilling
filling
vacancy
vacancyandand(b)
(b)with
withan
anadhesive
adhesivefilling
fillingvacancy.
vacancy.

Plywood
Plywoodlayers layersofofvarious
various thicknesses
thicknesses were inserted
were in each
inserted specimen
in each specimen to induce var-
to induce
ious levellevel
various differences. Specimens
differences. Specimens with a level
with a leveldifference
difference of of
3 and
3 and6 6mm mmwere werefabricated
fabricated
using
usingplywood
plywood with with 6T and 12T,
6T and 12T, respectively,
respectively,because
becauseofofthe the5.65.6
mmmm level
level difference
difference of
of the
the
LNG LNG CCS CCS insulation
insulation panel
panel [26].
[26]. TheThe plywood
plywood hadhad dimensions
dimensions of×
of 30 3050 × 50
mm mm to adjust
to adjust the
the jig joint
jig joint size.size.
TheThe jig joint
jig joint of each
of each specimen
specimen adheredadhered
to thetostrips
the strips to prevent
to prevent slippage slippage
during
during the tensile
the tensile tests. tests.
There
Thereare areseveral
severaldifferent
differentmechanical
mechanicaltests testsininthe
theISO
ISOstandards
standardsfor forindividual
individualstrips;
strips;
however,
however,no noappropriate
appropriatetest teststandard
standardexists
existsfor forstrips
stripswith
withaalevelleveldifference.
difference.Hence,Hence,we we
designed
designed aa special
special tensile
tensiletest
testwith
withsymmetrical
symmetrical specimens
specimens to verify
to verify the tensile
the tensile perfor-
performance
mance of a secondary
of a secondary barrierbarrier
with awith
levela level difference.
difference. Furthermore,
Furthermore, mechanical
mechanical teststests
werewere also
also performed on specimens with a 0 mm level difference, following the ISO 1421 strip
performed on specimens with a 0 mm level difference, following the ISO 1421 strip test
test standard, while the designed symmetrical tensile test was performed on specimensa
standard, while the designed symmetrical tensile test was performed on specimens with
with
leveladifference.
level difference. The tensile
The tensile tests were
tests were conducted
conducted at a crosshead
at a crosshead speed speed of 5 mm/min
of 5 mm/min and
◦
and
at −at120
−120C,°C, which
which is the
is the temperature
temperature that
that thethe secondary
secondary barrierofofthe
barrier theMARK
MARKIIIIIItank tankis
istypically
typicallyexposed
exposedtotoininthethefield.
field.Table
Table1 1lists
liststhe
theparameters
parametersused usedforforthethe tensile
tensile tests.
tests.

Table 1. Mechanical test scenarios for the FSB.

Material Temperature Level Difference Filling Ratio

50%
0 mm 70%
100%
50%
FSB −120 ◦ C 3 mm 70%
100%
50%
6 mm 70%
100%

Figure 7 shows the schematic diagram of specimens with different level differences/
adhesive filling ratios. Although the FSB is an anisotropic material, only the tensile load
direction generated by hogging and sagging was considered in this study, owing to the
 100%
50%
6 mm 70%
100%
Materials 2023, 16, 121 7 of 15
Figure 7 shows the schematic diagram of specimens with different level differ-
ences/adhesive filling ratios. Although the FSB is an anisotropic material, only the tensile
load direction generated by hogging
shrinkage and insulation
of the two sagging was considered
panels in this study,
in the opposite owing
direction. Theto test was performed
the shrinkage of theontwo insulation panels in the opposite direction. The test was per-
five specimens to ensure the validity of the results. The average of three main values
formed on five specimens to ensure
was accepted, the validity
apart from theofmaximum
the results.
andTheminimum
average of three main
values. Moreover, the measured
values was accepted,ultimate
apart from the maximum and minimum values. Moreover,
load was divided by two in the symmetrical mechanical the meas-
tests, to determine the
ured ultimate load was divided
ultimate loadbyoftwo
the in the symmetrical mechanical tests, to determine
FSB.
the ultimate load of the FSB.

(a)

, x FOR PEER REVIEW 8 of 16

(b)

(c)
Figure 7. Schematic diagram of specimens
Figure with
7. Schematic different
diagram of level differences/adhesive
specimens with different filling ratios: (a)
level differences/adhesive filling ratios:
0 mm/50%, (b) 3 mm/70%, (a)and (c) 6 mm/100%.
0 mm/50%, (b) 3 mm/70%, and (c) 6 mm/100%.

The mechanical tests were conducted by using a universal testing machine (UTM;
KSU-5M, Kyung Sung, Ansan, Korea) connected to a high-pressure liquid nitrogen con-
tainer to investigate the mechanical properties and fracture points of the FSB in a MARK
III tank. The UTM was used with a cryogenic insulation chamber, as shown in Figure 8.
 (c)
Materials 2023, 16, 121 Figure 7. Schematic diagram of specimens with different level differences/adhesive filling ratios:
8 of (a)
15
0 mm/50%, (b) 3 mm/70%, and (c) 6 mm/100%.

The mechanical tests were conducted by using a universal testing machine (UTM;
The mechanical tests were conducted by using a universal testing machine (UTM; KSU-
KSU-5M, Kyung Sung, Ansan, Korea) connected to a high-pressure liquid nitrogen con-
5M, Kyung Sung, Ansan, Korea) connected to a high-pressure liquid nitrogen container to
tainer to investigate the mechanical properties and fracture points of the FSB in a MARK
investigate the mechanical properties and fracture points of the FSB in a MARK III tank.
III tank. The UTM was used with a cryogenic insulation chamber, as shown in Figure 8.
The UTM was used with a cryogenic insulation chamber, as shown in Figure 8. Liquid
Liquid nitrogen (−196 °C) was distributed throughout the insulation chamber to maintain
nitrogen (−196 ◦ C) was distributed throughout the insulation chamber to maintain the
the desired temperature. Furthermore, the cryogenic mechanical test specimens were
desired temperature. Furthermore, the cryogenic mechanical test specimens were cooled
cooled until their temperature matched that of the surrounding environment before con-
until their temperature matched that of the surrounding environment before conducting
ducting the mechanical tests.
the mechanical tests.

Figure 8.
Figure 8. Photograph
Photograph of the
the UTM
UTM apparatus
apparatus used
used for
for the
the mechanical
mechanical tests.
tests.

The
The mechanical
mechanical load
load and displacement
displacement were
were measured
measured byby using
using a constantly
constantly moving
moving
crosshead at − ◦ C, the operating temperature
crossheadof ofthe
theUTM.
UTM.Although
Although the LNG
the LNGwas stored
was stored at163
−163 °C, the operating tempera-
◦ C, owing to the insulation materials; hence, all cryogenic tests were
of theofFSB
ture FSB−was
thewas 120 −120 °C, owing to the insulation materials; hence, all cryogenic tests
conducted at −120 ◦ C [27]. The target temperature of −of ◦ C (the temperature faced by
120−120
were conducted at −120 °C [27]. The target temperature °C (the temperature faced
secondary barriers during operation) was maintained using a thermometer and a control
system in the cryogenic chamber. Furthermore, the liquid nitrogen was circulated by a fan
to realize a uniform distribution of heat.
The thermal stress generated during cooling was eliminated by controlling the jig
position. The mechanical behavior was observed through the window of the cryogenic
chamber, owing to the variable mechanical behavior of the glass fibers in the FSB.

3. Results and Discussion

3.1. Maximum Tensile Load and Reliability
Table 2 presents the maximum fracture loads and displacements of the tensile test
specimens related to their level difference and adhesive filling ratio at −120 ◦ C. Overall,
the fracture load and displacement increased and decreased, respectively, with increasing
the adhesive filling ratio. However, the fracture load and displacement both decreased
with an increase in the level difference.
Materials 2023, 16, 121 9 of 15

Table 2. Mechanical properties of the FSB under each condition.

Material Temp (◦ C) Level Difference (mm) Filling Ratio (%) Fracture Load (kN) Displacement
(mm)
50 10.36 4.25
0 70 10.64 4.03
100 11.84 3.79
FSB with 50 9.51 4.11
epoxyadhesive −120
3 70 9.92 3.97
100 10.63 3.63
50 9.48 3.61
6 70 9.67 3.29
100 10.32 3.27

The maximum tensile load of the tensile test specimen with a level difference of 0 mm
and an adhesive filling ratio of 100% was 2.7% and 14.3% higher than those of the specimens
with adhesive filling ratios of 70% and 50%, respectively. However, the maximum tensile
loads of the specimens with a level difference of 3 mm and an adhesive filling ratio of 100%
were 7.1% and 11.7% higher than those of the specimens with adhesive filling ratios of
70% and 50%, respectively. Finally, a level difference of 6 mm and an adhesive filling ratio
of 100% led to maximum tensile loads that were 6.7% and 8.8% higher than those of the
specimens with adhesive filling ratios of 70% and 50%, respectively.
Furthermore, the maximum tensile load of the tensile test specimen with an adhesive
filling ratio of 50% and a level difference of 0 mm was 8.8% and 9.3% higher than those of the
specimens with a level difference of 3 and 6 mm, respectively. Accordingly, the maximum
tensile load for the specimen with an adhesive filling ratio of 70% and a level difference
of 0 mm was 2.6% and 10.0% higher than those of the specimens with a level difference
of 3 and 6 mm, respectively. However, the maximum tensile load for the specimen with
an adhesive filling ratio of 100% and a level difference of 0 mm was 11.5% and 14.8%
higher than those of the specimens with a level difference of 3 and 6 mm, respectively. The
percentage increase in the maximum tensile load of the specimens with respect to their level
difference and adhesive filling ratio strongly indicated that both these factors influence the
maximum tensile load of the FSB.
It can thus be concluded that an FSB with a level difference of 6 mm and an adhesive
filling ratio of 50% is the most prone to failure, with a high probability of fracture occurring
in the middle. On the other hand, the most stable conditions are when the FSB has zero
level difference and a 100% adhesive filling ratio; in this case, any fracture is likely to
manifest at the edge of the FSB.

3.2. Effect of the Level Difference

FSB fractures in the field originate from several causes, including impacts, thermal
loads, and structural issues; however, tensile loads are the most common cause, as pre-
viously mentioned. Therefore, in this study, tensile tests were performed on the FSB
specimens to investigate the effect of the level differences and adhesive vacancies. Figure 9
shows the load-displacement curves for the specimens according to the level difference at
−120 ◦ C. These curves show that the displacement and load decreased as the level differ-
ence increased. The specimen with a level difference of 3 mm exhibited a more compact
degradation than that with a level difference of 0 mm (Table 2). However, the specimen with
a level difference of 6 mm exhibited a severe degradation: its fracture load was 11.85 kN
when the level difference was 0 mm with an adhesive filling ratio of 100%; furthermore,
the fracture load was 10.32 kN, with a level difference of 6 mm and an adhesive filling
ratio of 100%. The tensile properties of the specimens with a level difference of 6 mm and
an adhesive filling ratio of 100% were 12.9% lower than those with a level difference of
0 mm and the same adhesive filling ratio. This indicated that a level difference above 6 mm
 shows the load-displacement curves for the specimens according to the level difference at
-120 °C. These curves show that the displacement and load decreased as the level differ-
ence increased. The specimen with a level difference of 3 mm exhibited a more compact
degradation than that with a level difference of 0 mm (Table. 2). However, the specimen
with a level difference of 6 mm exhibited a severe degradation: its fracture load was 11.85
kN when the level difference was 0 mm with an adhesive filling ratio of 100%; further-
Materials 2023, 16, 121 10 of 15
more, the fracture load was 10.32 kN, with a level difference of 6 mm and an adhesive
filling ratio of 100%. The tensile properties of the specimens with a level difference of 6
mm and an adhesive filling ratio of 100% were 12.9% lower than those with a level differ-
ence of 0 mm and thecritically promotes
same adhesive anratio.
filling FSB’s susceptibility
This indicated thatto afracturing because of structural issues. This
level difference
above 6 mm critically behavior
promotes stems
an FSB’s susceptibility
mainly from thetostress
fracturing because of structural
concentration, which also affects the fracture location.
issues. This behavior Figure
stems mainly
10 showsfrom that
the stress concentration,
the stress which also
concentration affects thewith an increase in the level
deteriorated
fracture location. Figure 10 shows that the stress concentration deteriorated
difference of the specimens. Overall, a level difference of with an6in-
mm and above was inferred to
crease in the level difference
significantly increase the probability of an FSB to sustain and
of the specimens. Overall, a level difference of 6 mm failure during operation, owing
above was inferred to significantly increase the probability of an FSB to sustain failure
to bending- and thermal-derived tensile stress.
during operation, owing to bending- and thermal-derived tensile stress.

22, 15, x FOR PEER REVIEW 11 of 16

(a)

(b)

(c)
Figure 9. Tensile load and displacement
Figure 9. Tensilecurves of specimens
load and with various
displacement level
curves of differences
specimens : (a)various
with 0 level differences: (a) 0 mm,
mm, (b) 3 mm, and (c) 6 mm.
(b) 3 mm, and (c) 6 mm.
 (c)
Materials 2023, 16, 121 11 of 15
Figure 9. Tensile load and displacement curves of specimens with various level differences : (a) 0
mm, (b) 3 mm, and (c) 6 mm.

(a) (b)
Materials 2022, 15, x FOR PEER REVIEW 12 of 16
Figure10.
Figure 10.Variation
Variationininthe
the(a)(a) maximum
maximum displacement
displacement andand (b) maximum
(b) maximum loadload of tensile
of tensile specimens
specimens
withvarying
with varyingadhesive
adhesive filling
filling rations
rations in relation
in relation to the
to the level
level difference.
difference.

3.3.Effect
3.3. Effectofofthe
theAdhesive
Adhesive Filling
Filling Ratio
Ratio
Figure 11 shows the typical load-displacement variations of the tensile test specimens
Figure 11 shows the typical load-displacement variations of the tensile test specimens
according to the adhesive filling ratio at −120 ◦ C. A decrease in the adhesive filling ratio
according to the adhesive filling ratio at −120 °C. A decrease in the adhesive filling ratio
increased the tensile fracture load and decreased the tensile fracture displacement. The
increased the tensile fracture load and decreased the tensile fracture displacement. The
specimen with an adhesive filling ratio of 100% exhibited significantly larger mechanical
specimen with an
reinforcement thanadhesive
that with filling ratio offilling
an adhesive 100% ratio
exhibited
of 50%. significantly
Accordingly,larger mechanical
the specimen
reinforcement than that with an adhesive filling ratio of 50%.
with an adhesive filling ratio of 70% had more compact reinforcement than that withAccordingly, the specimen
an
with an adhesive
adhesive filling
filling ratio ratio Furthermore,
of 50%. of 70% had more compactwith
the specimen reinforcement
an adhesivethan that
filling with
ratio of an
adhesive
50% andfilling
a levelratio of 50%.
difference of 0Furthermore,
mm exhibitedthe specimen
a tensile withload
fracture an adhesive filling
of 10.36 kN, ratio of
whereas
50%
thatand a level
of the difference
specimen with anof adhesive
0 mm exhibited a tensile
filling ratio of 100% fracture
and a load
level of 10.36 kN,
difference of whereas
0 mm
that of the specimen
exhibited one of 11.85 with
kN.anTheadhesive
tensile filling
fractureratio
load ofof
100% and a levelwith
the specimens difference of 0 mm
an adhesive
filling ratio
exhibited one ofof
100%
11.85and a level
kN. The difference of 0 mm
tensile fracture wasof14.3%
load higher thanwith
the specimens those anwith an
adhesive
adhesive
filling filling
ratio ratio and
of 100% of 50% and the
a level same level
difference of difference.
0 mm wasThis 14.3%suggested that the
higher than adhesive
those with an
filling ratio
adhesive critically
filling ratioaffects
of 50%theandoccurrence
the same oflevel
fractures in FSBs. This
difference. Therefore, adhesives
suggested that should
the adhe-
be applied to 100% of the FSB surface to minimize the fracturing probability.
sive filling ratio critically affects the occurrence of fractures in FSBs. Therefore, adhesives

(a) (b)
Figure
Figure11.
11.Variation inthe
Variation in themaximum
maximum (a)(a) displacement
displacement andand (b) tensile
(b) tensile load
load of the of the tensile
tensile test speci-
test specimens
mens according to the adhesive filling
according to the adhesive filling ratio. ratio.

Figure 12a shows the scanning electron microscopy (SEM) images of the fracture sur-
face of the FSB, verifying its glass fiber composition; furthermore, it shows a bundle of
glass fibers at the fracture surface of an FSB specimen. Figure 12b shows the impregnated
epoxy adhesive penetrating the glass fibers of the FSB. It was assumed that the impreg-
Materials 2023, 16, 121 12 of 15

Figure 12a shows the scanning electron microscopy (SEM) images of the fracture
surface of the FSB, verifying its glass fiber composition; furthermore, it shows a bundle of
glass fibers at the fracture surface of an FSB specimen. Figure 12b shows the impregnated
epoxy adhesive penetrating the glass fibers of the FSB. It was assumed that the impregnated
epoxy adhesive acted as an epoxy coating, absorbing the tensile load rather than the FSB
while hardening the glass fibers on the surface. Additionally, the stress in the FSB was
dispersed and transmitted to the epoxy matrix through the epoxy resin chain. Therefore,
the epoxy adhesive effectively improved the tensile strength of the FSB by reducing the
stress concentration at the defects while increasing the maximum tensile fracture load [28].
Also, the adhesive-penetrated FSBs possessed higher surface energy and work values than
their untreated equivalents [29], suggesting that applying the epoxy adhesive on the FSB
Materials 2022, 15, x FOR PEER REVIEW
surface strongly promoted crack healing. The penetrated epoxy can be used as a 13 of 16
surface
treatment, reinforcing the tensile strength of the FSB by absorbing the tensile load [30].

(a) (b)

Figure 12.
Figure SEMimages
12. SEM imagesof
ofthe
thefracture
fracturesurface
surfaceof
of tensile
tensiletest
test specimens:
specimens: (a)
(a)surface
surfaceof
ofglass
glassfibers
fibers in
in
nonadhesive-treated FSB;
nonadhesive-treated FSB; (b)
(b) epoxy
epoxy adhesive-penetrated
adhesive-penetrated glass
glass fibers
fibers of FSB.

3.4. Fracture Location within Specimens

3.4. Fracture Location within Specimens
Fractures generally occurred near the weaker points of our FSB specimens and were
Fractures generally occurred near the weaker points of our FSB specimens and were
influenced by two main parameters: the level difference and the adhesive filling ratio.
influenced by two main parameters: the level difference and the adhesive filling ratio.
Figure 13 shows the fracture characteristics according to the level difference and the
Figure 13 shows the fracture characteristics according to the level difference and the ad-
adhesive filling ratio of the specimens. Two main fracture types were observed. The first
hesive filling ratio of the specimens. Two main fracture types were observed. The first was
was located at the end of the gauge length of the specimen, with a level difference of
located at the end of the gauge length of the specimen, with a level difference of 6 mm and
6 mm and an adhesive filling ratio of 100% (Figure 13b). In general, specimens with an
an adhesive filling ratio of 100% (Figure 13b). In general, specimens with an adhesive fill-
adhesive filling ratio of 100% were subjected to a uniform tensile load; however, those with
ing ratio of
samples were 100% were subjected
subjected to a uniform
to a nonuniform tensile
tensile loadload;
at thehowever,
location those
wherewith samples
the FSB bent.
were subjected to a nonuniform tensile load at the location where
The second fracture type occurred in the middle of the gauge length of the specimens, the FSB bent. The second
with
fracture type occurred
a level difference in the middle
and adhesive of the(Figure
vacancies gauge 13c).
length Inofthis
thecase,
specimens, with
the tensile a level
load was
difference
concentrated andinadhesive vacancies (Figureareas
the nonadhesive-covered 13c).ofInthe
thisFSB
case, the adhesive
(if the tensile load
waswas notconcen-
applied
trated
to the in the nonadhesive-covered
entire areas of the FSB
surface) because the impregnated epoxy(if the adhesive
adhesive wasasnot
acted appliedcoating
a surface to the
entire surface) because the
that provided a reinforcement effect. impregnated epoxy adhesive acted as a surface coating that
provided a reinforcement effect.
The location of the fracture was determined on the basis of the level difference and
The location
adhesive of theoffracture
filling ratio was determined
the specimens. For theon the basis of
specimens thefeatured
that level difference and
both a level
adhesive filling ratio of the specimens. For the specimens that featured
difference and an adhesive filling ratio, the fracture occurred in the middle of the gauge both a level differ-
ence andTherefore,
length. an adhesive the filling
fracture ratio, the fracture
location was more occurred
strongly ininfluenced
the middleby ofthe
theadhesive
gauge length.
filling
Therefore,
ratio than by thethe
fracture location was more strongly influenced by the adhesive filling ratio
level difference.
than by the level difference.
Materials 2022,16,
Materials2023, 15,121
x FOR PEER REVIEW 13 of
14 of 15
16

(a)

(b) (c)
Figure 13. Fracture location in specimens according to the level difference and adhesive filling ratio:
Figure 13. Fracture location in specimens according to the level difference and adhesive filling ratio:
(a) images of tensile test specimens; schematic diagrams of a fracture (b) in the middle and (c) at the
(a) images of tensile test specimens; schematic diagrams of a fracture (b) in the middle and (c) at the
end of the gauge length.
end of the gauge length.

4.4.Conclusions
Conclusions
Thisstudy
This studyevaluated
evaluatedthethestructural
structuralvulnerability
vulnerabilityand
andfailure
failurecharacteristics
characteristicsof
ofglass
glass
fabric–reinforced composites at cryogenic temperatures (−120 ◦°C) while factoring
fabric–reinforced composites at cryogenic temperatures (−120 C) while factoring in the in the
influenceof
influence ofthe
thelevel
leveldifference
differenceandandadhesive
adhesivevacancies,
vacancies,using
usingaanewly
newlydeveloped
developedtesttest
procedurewith
procedure withslip-prevention
slip-preventionfeatures.
features.The
Themain
mainresults
resultsare
aresummarized
summarizedasasfollows:
follows:
•• Weproposed
We proposeda amechanical
mechanicaltest
testmethod
method forfor textile
textile and
and glass
glass fiber–reinforced
fiber–reinforced materi-
materials
als under
under cryogenic
cryogenic conditions.
conditions. TheThe
jigjig was
was machinedbolted
machined boltedand
andclamped
clampedwith
withthe
the
specimen,preventing
specimen, preventingthe
therotation
rotationofofthe
thegrip
gripsection
sectionofofthe
thespecimen.
specimen.
• Slippage and the premature fracture of the adhesive joints prior to the specimen were
prevented by using our newly proposed test method. The curing time was reduced
to 1 day, and the rate of experimental failure decreased to 20%.
Materials 2023, 16, 121 14 of 15

• Slippage and the premature fracture of the adhesive joints prior to the specimen were
prevented by using our newly proposed test method. The curing time was reduced to
1 day, and the rate of experimental failure decreased to 20%.
• As a result of the tensile test with the proposed jig and specimen, fracturing was more
likely to occur at the end of the gauge length in specimens with a level difference.
Parallelly, the maximum tensile load and displacement decreased with the increasing
level difference.
• Fracturing was more likely to occur at the middle of the gauge length in specimens
with adhesive vacancies because of the general load concentration. Concurrently, the
maximum tensile load and displacement increased and decreased, respectively, with
the increasing adhesive filling ratio.
• A microstructural analysis revealed that the fracture load increased with the increasing
adhesive filling ratio because the impregnated epoxy adhesive absorbed the tensile
load rather than the FSB and hardened the glass fibers on the surface.
Consequently, our proposed test method is highly suited for conducting tensile tests
on textile and glass fiber–reinforced materials under cryogenic conditions, by significantly
reducing the experimental failure rates and preventing slippage to offer accurate measure-
ments. Furthermore, we observed that compared with the level difference, the adhesive
filling ratio of the specimens more prominently deteriorated the fracture load and deter-
mined the fracture location. Therefore, scrupulous inspections for adhesive vacancies
should be performed during the manufacturing management stage to ensure the cryogenic
reliability of the FSB.

Author Contributions: Conceptualization, J.-M.L. and Y.-C.J.; investigation, J.-M.L., Y.-C.J., D.-H.L.
and S.-K.K.; writing—original draft preparation, Y.-C.J. and J.-M.L.; data curation, J.-H.K.; writing—
review and editing, J.-H.K., D.-H.L. and S.-K.K.; visualization, D.-H.L.; supervision, J.-M.L. All
authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by the Korean Institute of Energy Technology Evaluation and
Planning (KETEP) and the Ministry of Trade, Industry, and Energy (MOTIE) of the Republic of
Korea (20224000000090). This work was supported by the Materials/Parts Technology Develop-
ment Program (20017530, Development of Cryogenic Insulation Materials and Container Applica-
tion/Evaluation Technology for Liquid Hydrogen Storage Containers for Hydrogen Commercial
Vehicles) funded by the Ministry of Trade, Industry, and Energy (MOTIE, Korea). This research was
supported by the Development and Demonstration of Onboard Marine Debris Disposal Modules
program of the Korean Institute of Marine Science and Technology Promotion (KIMST), funded by
the Ministry of Oceans and Fisheries (KIMST-20220494).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Conflicts of Interest: The authors declare no conflicts of internet.

References
1. De Carvalho, J.R.; Arnaldo, V. Natural gas and other alternative fuels for transportation purposes. Energy 1985, 10, 187–215.
[CrossRef]
2. Cha, S.J.; Kim, J.D.; Kim, S.K.; Kim, J.H.; Oh, H.K.; Kim, Y.T.; Park, S.B.; Lee, J.M. Effect of temperature on the mechanical
performance of plywood used in membrane-type LNG carrier insulation systems. J. Wood Sci. 2020, 66, 28. [CrossRef]
3. International Code for the Construction and Equipment of Ships Carrying Liquefied Gases in Bulk (I.G.C. Code); International Maritime
Organization (IMO) Publications: London, UK, 1986.
4. Bae, J.H.; Hwang, B.K.; Kim, J.H.; Kim, S.K.; Lee, J.M. Cumulative damage of hollow glass microsphere weight fraction in
polyurethane foam in response to cryogenic temperatures and repeated impact loading. Cryogenics 2020, 107, 103057. [CrossRef]
5. Kim, J.H.; Choi, S.W.; Park, D.H.; Park, S.B.; Kim, S.K.; Park, K.J.; Lee, J.M. Effects of cryogenic temperature on the mechanical
and failure characteristics of melamine-urea-formaldehyde adhesive plywood. Cryogenics 2018, 91, 36–46. [CrossRef]
Materials 2023, 16, 121 15 of 15

6. Kim, B.C.; Soon, H.Y. Pressure resistance of the corrugated stainless steel membranes of LNG carriers. Ocean. Eng. 2011, 38,
592–608. [CrossRef]
7. Jeong, Y.J.; Kim, H.T.; Kim, S.K.; Kim, J.H.; Kim, Y.T.; Heo, W.S.; Lee, J.M. Evaluation of the pressure-resisting capability of
membrane-type corrugated sheet under hydrodynamic load. Thin-Walled Struct. 2021, 162, 107388. [CrossRef]
8. Lee, D.H.; Bae, J.H.; Hwang, B.K.; Kim, J.H.; Kim, S.K.; Lee, J.M. Impact behavior of hollow glass bubble reinforced foam core
LNG insulation panel in cryogenic temperature. J. Compos. Mater. 2021, 55, 1995–2010. [CrossRef]
9. Kim, S.K.; Kim, J.D.; Lee, D.H.; Kim, J.H.; Lee, J.M. Characteristics of pre-strained polyisocyanurate foam: Deformation recovery
and compressive mechanical behavior at cryogenic temperature. J. Cell. Plast. 2021, 58, 357–376. [CrossRef]
10. Kim, K.H.; Yoon, S.H. Vibration isolation of LNG containment systems due to sloshing with glass fiber composite. Compos. Struct.
2012, 94, 469–476. [CrossRef]
11. Kim, J.H.; Park, D.H.; Lee, C.S.; Park, K.J.; Lee, J.M. Effects of cryogenic thermal cycle and immersion on the mechanical
characteristics of phenol-resin bonded plywood. Cryogenics 2015, 72, 90–102. [CrossRef]
12. Bang, C.S.; Kim, J.G.; Lee, D.G. Performance improvement by glass fiber of adhesively bonded metal joints at the cryogenic
temperature. Compos. Struct. 2013, 96, 321–331. [CrossRef]
13. Yoon, S.H.; Lee, D.G. Design of the composite sandwich panel of the hot pad for the bonding of large area adhesive films. Compos.
Struct. 2011, 94, 102–113. [CrossRef]
14. Karthik, K.; Rajamani, D.; Manimaran, A.; Udayaprakash, J. Evaluation of tensile properties on Glass/Carbon/Kevlar fiber
reinforced hybrid composites. Mater. Today Proc. 2021, 39, 1655–1660. [CrossRef]
15. Maciel, N.D.; Ferreira, J.B.; da Silva Vieira, J.; Ribeiro, C.G.; Lopes, F.P.; Margem, F.M.; Monteiro, S.N.; Vieira, C.M.F.; da Silva,
L.C. Comparative tensile strength analysis between epoxy composites reinforced with curaua fiber and glass fiber. J. Mater. Res.
Technol. 2018, 7, 561–565. [CrossRef]
16. Rahmanian, S.; Thean, K.S.; Suraya, A.R.; Shazed, M.A.; Salleh, M.M.; Yusoff, H.M. Carbon and glass hierarchical fibers: Influence
of carbon nanotubes on tensile, flexural and impact properties of short fiber reinforced composites. Mater. Des. 2013, 43, 10–16.
[CrossRef]
17. Tamaki, N.; Kimiyoshi, N.; Junro, K.; Yutaka, K. Influence of prepreg conditions on the void occurrence and tensile properties of
woven glass fiber-reinforced polyimide composites. Compos. Sci. Technol. 2009, 69, 2428–2433. [CrossRef]
18. Oh, D.J.; Lee, J.M.; Chun, M.S.; Kim, M.H. Reliability evaluation of a LNGC insulation system with a metallic secondary barrier.
Compos. Struct. 2017, 171, 43–52. [CrossRef]
19. Bang, C.S.; Park, C. Optimum glass fiber volume fraction in the adhesive for the Al-SUS adhesively bonded joints at cryogenic
temperatures. Compos. Struct. 2014, 108, 119–128. [CrossRef]
20. Ribeiro, F.M.; Campilho, R.D.; Carbas, R.J.; Da Silva, L.F. Strength and damage growth in composite bonded joints with defects.
Compos. Part B Eng. 2016, 100, 91–100. [CrossRef]
21. Lee, C.S.; Chun, M.S.; Kim, M.H.; Lee, J.M. Debonding failure characteristics of multi-laminated bonding system under cryogenic
temperature. Int. J. Adhes. Adhes. 2011, 31, 226–237. [CrossRef]
22. Yoon, S.H.; Kim, K.H.; Lee, D.Y. Cryogenic strength of adhesive bridge joints for thermal insulation sandwich constructions.
Compos. Struct. 2014, 111, 1–12. [CrossRef]
23. Yoon, S.H.; Kim, K.H. Improvement of the adhesive peel strength of the secondary barrier with level difference for LNG
containment system. Compos. Struct. 2013, 95, 528–538. [CrossRef]
24. Stefanov, T.; Ryan, B.; Ivankovic, A.; Myrphy, N. Dynamic mechanical analysis of carbon black filled, elastomer-oughened ethyl
cyanoacrylate adhesive bulk films. Int. J. Adhes. Adhes. 2020, 101, 102630. [CrossRef]
25. ISO P 1421; Rubber-or Plastics-Coated Fabrics–Determination of Tensile Strength and Elongation at Break. ISO: Geneva,
Switzerland, 2016.
26. Wang, B.; Shin, Y.S.; Eric, N. Hull Deformation Effect on Membrane-Type LNG Containment Systems. In Proceedings of the
International Conference on Offshore Mechanics and Arctic Engineering, Busan, Republic of Korea, 19–24 June 2016; p. 49941.
[CrossRef]
27. Lu, J.; Xu, S.; Deng, J.; Wu, W.; Wu, H.; Yang, Z. Numerical prediction of temperature field for cargo containment system (CCS) of
LNG carriers during pre-cooling operations. J. Nat. Gas Sci. Eng. 2016, 29, 382–391. [CrossRef]
28. Chen, P.; Wang, Y.; Li, J.; Wang, H.; Zhang, L. Adhesion and erosion properties of epoxy resin composite coatings reinforced with
fly ash cenospheres and short glass fibers. Prog. Org. Coat. 2018, 125, 489–499. [CrossRef]
29. Botelho, E.C.; Silva, R.A.; Pardini, L.C.; Rezende, M.C. Evaluation of adhesion of continuous fiber–epoxy composite/aluminum
laminates. J. Adhes. Sci. Technol. 2004, 18, 1799–1813. [CrossRef]
30. Rajak, D.K.; Wagh, P.H.; Moustabchir, H.; Pruncu, C.I. Improving the Tensile and Flexural Properties of Reinforced Epoxy
Composites by using Cobalt Filled and Carbon/Glass Fiber. Forces Mech. 2021, 4, 100029. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.