materials

Article
Study on the Compressive Properties of Magnesium
Phosphate Cement Mixing with Eco-Friendly Coir
Fiber Considering Fiber Length
Zuqian Jiang, Liwen Zhang * , Tao Geng, Yushan Lai, Weile Zheng and Min Huang
Department of Civil Engineering, Guangzhou University, Guangzhou 510006, China;
zuqianjiang1996@163.com (Z.J.); gengtaoltthl@163.com (T.G.); Ln3maize@163.com (Y.L.);
wlzhengaurora@163.com (W.Z.); hm1132628255@163.com (M.H.)
* Correspondence: lwzhang@gzhu.edu.cn; Tel.: +86-139-0223-1755




Received: 26 June 2020; Accepted: 15 July 2020; Published: 17 July 2020


Abstract: Coir fiber (CF), an eco-friendly and renewable natural fiber, was introduced into magnesium
phosphate cement (MPC) mortar to improve its crack resistance. A total of 21 specimens were
employed to investigate the failure pattern, compressive strength, stress–strain curve, and energy
absorption of MPC with varying CF lengths (0, 5, 10, 15, 20, 25, and 30 mm) after a curing
period of 28 days through a static compressive test. The results demonstrated that compressive
strength, elastic modulus, and secant modulus decreased with the increase in CF length. However,
energy absorption presented a convex curve, which increased to the maximum value (77.0% relative
to the value of the specimen without CF) with a CF length of 20 mm and then declined. A series of
modern micro-tests were then carried out to analyze the microstructure and composition of specimens
to explain the properties microscopically.

Keywords: magnesium phosphate cement; coir fiber; fiber length; compressive properties;
energy absorption; micro-morphology

1. Introduction
Magnesium phosphate cement (MPC), a chemically bonded ceramic [1,2], is a new inorganic
cementitious material produced by a neutralization reaction between phosphate and magnesium
oxide, and has been widely used in concrete bridge structure reinforcement and road repairment [3,4].
However, its decreased toughness compared to older building materials can result in secondary
damage to these reinforced bridges and repaired roads. To overcome this problem, many studies
were conducted on incorporating various synthetic fibers, such as steel fiber, glass fiber, and basalt
fiber into MPC to improve its toughness. Feng et al. [5] found that the incorporation of micro steel
fiber can significantly improve the compressive strength and flexural toughness of MPC composites.
Ahmad et al. [6] confirmed that MPC has higher compressive strength and superior ductility when
basalt fiber is involved in MPC mortars. Fang et al. [7] reported that glass fiber can effectively
prevent the cracking of MPC. However, the application of these synthetic fibers usually results in
excessive carbon dioxide emissions and energy consumption during their production and waste
treatment, which contribute to the creation and severity a series of environmental issues, such as global
warming [8]. An eco-friendly material instead of synthetic fibers must be found to improve these
drawbacks of MPC.
Compared with synthetic fibers, coir fiber (CF), as a natural plant fiber, has the advantages of
being fully biodegradable, abundantly available, renewable, and environmentally friendly [9,10].
In addition to these remarkable eco-friendly characteristics, CF has remarkable toughness among

Materials 2020, 13, 3194; doi:10.3390/ma13143194 www.mdpi.com/journal/materials

Materials 2020, 13, 3194 2 of 12

natural fibers [11–13], showing a high potential to effectively improve the toughness and mechanical
properties of cement composites. As Thanushan et al. [14] reported, CF can improve the residual
strength, ductility, and energy absorption of earth cement blocks. Li et al. [15] found that cementitious
composites reinforced by CF have better flexural strength, higher energy absorption, and increased
ductility. These cementitious composites are also lighter than other conventional cementitious materials.
Thus, CF is thought to have the ability to improve the drawbacks of MPC because it significantly
improves other cement composites. As existing studies mostly focused on the effect of other fibers
on MPC or the effect of CF on common concrete, a study on MPC reinforced by CF is needed to
understand the mechanical properties of this new cement-based composite material for its application
in practical works.
In this study, a compressive test was conducted to clarify the effect of CF length on the compressive
properties of MPC. A total of 21 specimens were employed in the test to investigate the failure patterns,
compressive strength, stress–strain behavior, and energy absorption of MPC with different CF lengths.
Modern micro-testing techniques, such as scanning electron microscope (SEM) and energy dispersive
X-ray detector (EDX), were adopted to analyze the microstructure and composition of specimens to
explain the properties microscopically.

2. Experimental Program

2.1. Raw Materials

The MPC binder employed in this test was composed of dead-burned magnesium oxide (MgO),
potassium dihydrogen phosphate (KH2 PO4 ), and fly ash (FA), at the KH2 PO4 –MgO mass ratio of 0.68,
the FA–MgO mass ratio of 0.25, and the water–MPC binder mass ratio of 0.15. Dead-burned MgO
powder (calcined at >1500 ◦ C for 6 h) with a specific surface area of 228 m2 /kg, a density of 2650 kg/m3 ,
and an average particle size of 45 µm, was supplied by Zhengyang Casting Material Company (Xinmi,
China). Its chemical composition is shown in Table 1. American Society of Testing Materials (ASTM)
C618 F class FA was provided by Hengnuo Co. Ltd. (Gongyi, China). Its chemical composition is
listed in Table 2. Industrial-grade KH2 PO4 , having a purity of 99.5% and an average particle size of
350 µm, was provided by Jiang Hua Chemical Glass Instrument Co. Ltd. of Nanjing, Jiangsu, China.
In addition, borax (Na2 B4 O7 ·10H2 O), used as a retarder, with a mass of 10% MgO, a purity of 99.5%,
and an average particle size of 350 µm, was supplied by Jiang Hua Chemical Glass Instrument Co. Ltd.
of Nanjing, Jiangsu, China. The mixing proportion of these materials is shown in Table 3.

Table 1. Chemical composition of MgO.

Composition MgO Al2 O3 Fe2 O3 CaO SiO2 Loss

wt % 96.25 0.29 1.09 1.18 1.16 0.03

Table 2. Chemical composition of Fla ash (FA).

Composition SiO2 Al2 O3 Fe2 O3 CaO TiO2 Loss

wt % 56.74 24.58 6.55 4.87 1.86 5.4
ASTM C618 (%) ≥70 - - ≤6

Table 3. Coir Fiber (CF) physical and mechanical properties.

Tensile Tensile
Diameter Density Moisture Length Elongation
Strength Modulus
(µm) (kg/m3 ) Content (%) (mm) at Break (%)
(MPa) (GPa)
150–350 1200 10–12 5–30 128–157 3.86–5.60 21.2–40.7
Materials 2020, 13, 3194 3 of 12

The CF (Jiagaocheng import and Export Trade Co. Ltd., Shangrao, Jiangxi, China), came from Sri
Lanka, and had a diameter range of 150–350 µm and a density of 1200 kg/m3 . Before the CF was mixed
with the MPC mortar, it was boiled, dried, and cut into different lengths. Six CF lengths (5, 10, 15, 20,
25, and 30 mm) were used to investigate the effects of CF length on the compressive properties of MPC.
The detailed physical and mechanical properties of CF are shown in Table 4.

Table 4. Specimen parameters.

MPC Binder (Mass Ratio) Volume Fiber

Specimen B/M 1 W/BM 2
Group Fraction 3 Length
Number MgO KH2 PO4 Fly Ash (%) (%)
(Vf , %) (mm)
SC-CFL0-1
CFL0 SC-CFL0-2 0.0
SC-CFL0-3
SC-CFL5-1
CFL5 SC-CFL5-2 5.0
SC-CFL5-3
SC-CFL10-1
CFL10 SC-CFL10-2 10.0
SC-CFL10-3
1.0 0.68 0.25 10 15 3.0
SC-CFL15-1
CFL15 SC-CFL15-2 15.0
SC-CFL15-3
SC-CFL20-1
CFL20 SC-CFL20-2 20.0
SC-CFL20-3
SC-CFL25-1
CFL25 SC-CFL25-2 25.0
SC-CFL25-3
SC-CFL30-1
CFL30 SC-CFL30-2 30.0
SC-CFL30-3
1 B/M is the mass ratio of borax to MgO; 2 W/BM is the mass ratio of water to binder; 3 3.0% volume fraction refers
to existing research [16–18].

To prepare the specimens, KH2 PO4 , Na2 B4 O7 ·10H2 O, and FA dry powder were first put into a
shaft mixer and stirred uniformly for about 30 s. Second, MgO was added into the mixture and stirred
continuously for another 30 s. Then, water was poured into the mixture and stirred for 1 min until
MPC mortar was produced. After, CF of different fiber lengths was distributed into the MPC mortar
and mixed quickly for 30 s. Finally, the well-mixed mortar was pumped into molds and demolded
1 h later to be cured in a room with approximately relative humidity (RH) of 45 ± 5% at 2 ± 3 ◦ C for
28 days.

2.2. Specimens and Test Setup

In total, 21 specimens were employed in the test, each with dimensions of 40 × 70 × 40 mm,
as shown in Figure 1. The specimens were divided into seven groups according to CF length. One of
the groups was designated as the control specimen, which is without CF. Each specimen was labeled
test method-CF length-specimen number. For example, SC-CFL10-2 represented the second specimen
with 10 mm long CF, which was used to investigate compressive performance. Table 4 offers detailed
parameters of all specimens.
Materials 2020, 13, 3194 4 of 12
Materials 2020, 13, x FOR PEER REVIEW 4 of 12

Figure1.1. Test
Figure Testsetup.
setup.

A
A static
static compression
compression testtest was
was carried
carried out
out using
using an an MTS-E45.305
MTS-E45.305 electronic
electronic universal
universal testing
testing
machine
machine in in reference to the
the ASTM
ASTMC109,C109,asasshown
shownininFigure
Figure1.1. First,
First, specimens
specimens werewere preloaded
preloaded at
at the
the
raterate of 1.0
of 1.0 mm/min
mm/min until
until the the loading
loading force
force reached
reached 0.1 0.1
kN.kN. Then,
Then, loading
loading started
started at the
at the raterate of
of 2.0
2.0 mm/min.
mm/min. TheThe load–displacement
load–displacement (L–D)
(L–D) curvedata
curve datawere
wererecorded
recordedin inaacomputer
computer by sensors installed
installed
at
atthe
theload
loadcell.
cell.When
Whenthetheloading
loadingforce dropped
force dropped to to
50%50%of of
thethe
peak load,
peak thethe
load, testtest
waswas
stopped, and and
stopped, the
specimen
the specimenwas was
considered to betoabe
considered failure. The The
a failure. compressive
compressive strength of specimens
strength of specimens waswas
calculated by
calculated
Equation (1) according to ASTM C109-16
by Equation (1) according to ASTM C109-16 [19]: [19]:
P
f Puu (1)
fcu cu= A (1)
A
where fcu is the compressive strength (MPa), Pu is the maximum load (kN), and A is the area under
where fcu is the compressive2 strength (MPa), Pu is the maximum load (kN), and A is the area under
compression (A = 2800 mm ). The elastic modulus of specimens was calculated using Equation (2) in
compression (A = 2800 mm2 ). The elastic modulus of specimens was calculated using Equation (2) in
accordance with ASTM C469-14 [20]:
accordance with ASTM C469-14 [20]:
σ −σ
EE = 2 2 1 1 (2)
(2)
ε2 −
2 
ε11
whereEEisisthe
where theelastic
elasticmodulus (GPa),σσ1 1isisthe
modulus(GPa), stressatatε1ɛ1== 0.5
thestress 0.5 uε (MPa), σσ22 is
uɛ (MPa), is the
the stress
stress increased
increased toto
40% of peak stress (MPa), and ε is the longitudinal strain corresponding to stress σ
40% of peak stress (MPa), and ɛ2 is the longitudinal strain corresponding to stress σ2. For specimen
2 2 . For specimen
failure,
failure,aahigh-definition
high-definitioncamera
camerawas
wasused
usedto tocapture
capturethethefailure
failurepattern.
pattern. Samples
Samples werewerealsoalsocollected
collected
from
frombroken
brokenspecimens, undergoing
specimens, analyses
undergoing by SEM,
analyses by EDX,
SEM,and X-ray
EDX, diffraction
and (XRD) to investigate
X-ray diffraction (XRD) to
specimens’
investigate micro-morphology and hydration
specimens’ micro-morphology products
and hydrationto clarify the to
products mechanisms
clarify the of CF behavior.
mechanisms of CF
behavior.
3. Results and Discussion
3. Results
3.1. Failure and Discussion
Patterns
The failure
3.1. Failure patterns (Figure 2) from the compression test were divided into four types as displayed
Patterns
in Figure 3. Type I (Figure 3a) was a typical brittle failure, in which specimens failed almost
The failure patterns (Figure 2) from the compression test were divided into four types as
instantaneously when cracks occurred. The MPC matrix broke into pieces after the loading reached
displayed in Figure 3. Type I (Figure 3a) was a typical brittle failure, in which specimens failed almost
the maximum value. Type II (Figure 3b) was similar to Type I, but separated portions did not fall off
instantaneously when cracks occurred. The MPC matrix broke into pieces after the loading reached
the maximum value. Type II (Figure 3b) was similar to Type I, but separated portions did not fall off
from the specimens due to the connection of CF in the cracks. Compared with Types I and II, Type
Materials 2020, 13, 3194 5 of 12

Materials 2020, 13, x FOR PEER REVIEW 5 of 12

Materials 2020, 13, x FOR PEER REVIEW 5 of 12
from the specimens due to the connection of CF in the cracks. Compared with Types I and II, Type III
III displayed
displayed an
an an initial
initial signal
signal of
of of ductile
ductile failure. AsininFigure
failure. Figure 3c,specimens
specimens stillhad had a resistance to to load
III displayed initial signal ductile failure.AsAs in Figure3c, 3c, specimensstill
still had aa resistance
resistance load
after the
after the loading
loading force
force reached
reached its
its peak
peak value.
value. With thethe loading
loading process,
process, cracks
cracks were
were observed
observed moremore
With
gradually
gradually in in specimens,
in specimens,
specimens,and and often
andoften smaller
oftensmaller in width
smallerininwidth
width than
thancracks
cracksin Types I
in Types and II.
I and Specimens
II. Specimens were not
were
than cracks in Types I and II. Specimens were not
penetrated
not penetratedby cracks
by cracksbecause
becauseCF effectively restricted the development of these cracks. Figure 3d
penetrated by cracks because CFCF effectively
effectively restrictedthe
restricted thedevelopment
developmentofofthese these cracks.
cracks. Figure
Figure 3d
depicts the
depicts the finaltype
type of failurepattern,
pattern, Type IV, which showed a similar mode to Type III, except
the final
final typeofoffailure
failure pattern,TypeType IV,IV,
which
whichshowed
showed a similar mode
a similar to Type
mode III, except
to Type that
III, except
that
some some
CF CF intertwisted
intertwisted or or agglomerated
agglomerated in in cracks.
cracks.
that some CF intertwisted or agglomerated in cracks.

Figure 2. Failure patterns.

Figure 2. Failure patterns.

Figure
Figure 3.
3. Failure
Failurepatterns
patternstype
typeof
ofthe
thespecimen:
specimen: (a)
(a)Type
Type I,I, (b)
(b) Type
Type II, (c) Type III, and (d) Type
Type IV.
IV.
Figure 3. Failure patterns type of the specimen: (a) Type I, (b) Type II, (c) Type III, and (d) Type IV.

Failure patterns of each group are summarized in Figure 2. Type I patterns were found in group
Failure patterns of each group are summarized in Figure 2. Type I patterns were found in group
CFL0, in which nothing was observed in the initial initial stage
stage ofof loading.
loading. However, when loading force
CFL0, in which nothing was observed in the initial stage of loading. However, when loading force
was close to its
its peak
peak value,
value, an
an obvious
obvious crack occurred in the the specimens
specimens andand rapidly
rapidly extended
extended after
was close to its peak value, an obvious crack occurred in the specimens and rapidly extended after
the peak load until specimens were penetrated. Then, specimens failed instantaneously with a harsh
the peak load until specimens were penetrated. Then, specimens failed instantaneously with a harsh
rock-fissuring sound. Group
Group CFL5
CFL5 presented
presented the
the Type II failure. It performed like CFL0 in the test
rock-fissuring sound. Group CFL5 presented the Type II failure. It performed like CFL0 in the test
process, but did not break up into separated pieces
process, but did not break up into separated pieces duedue to to
CFCF
connecting
connectingthese pieces
these together.
pieces together.With CF
With
process, but did not break up into separated pieces due to CF connecting these pieces together. With
length increasing,
CF length specimens
increasing, displayed
specimens Type III
displayed Typefailure patterns.
III failure CFL10, CFL10,
patterns. CFL15, CFL15,
and CFL20 andcracked
CFL20
CF length increasing, specimens displayed Type III failure patterns. CFL10, CFL15, and CFL20
slowly
crackedwith frequent
slowly sounds sounds
with frequent of CF rupture and pulled
of CF rupture out CFout
and pulled when
CF loading exceeded
when loading the peak
exceeded the
cracked slowly with frequent sounds of CF rupture and pulled out CF when loading exceeded the
load. These specimens were greatly improved in axial deformation compared
peak load. These specimens were greatly improved in axial deformation compared with the with the specimens
peak load. These specimens were greatly improved in axial deformation compared with the
that incorporated
specimens CF shorter than
that incorporated 10 mm.than
CF shorter When CF length
10 mm. Whenexceeded
CF length20exceeded
mm, i.e., 20
CFL25
mm,andi.e., CFL30,
CFL25
specimens that incorporated CF shorter than 10 mm. When CF length exceeded 20 mm, i.e., CFL25
specimens
and CFL30,showed
specimensthe showed
Type IVthefailure
Typepattern. In pattern.
IV failure this type, Insome agglomerated
this type, CF was observed
some agglomerated CF was
and CFL30, specimens showed the Type IV failure pattern. In this type, some agglomerated CF was
in cracks due
observed to the due
in cracks overlong
to the CF,overlong
resulting CF,
in poor fluidity
resulting in ofpoor
MPCfluidity
mortar of
andMPCinhomogeneous
mortar and
observed in cracks due to the overlong CF, resulting in poor fluidity of MPC mortar and
distribution
inhomogeneous of CF.distribution
This meant of thatCF.
no This
CF was found
meant in some
that no CFcracks, although
was found specimens
in some stillalthough
cracks, showed
inhomogeneous distribution of CF. This meant that no CF was found in some cracks, although
greater ductility.
specimens still showed greater ductility.
specimens still showed greater ductility.
Materials 2020, 13, x FOR PEER REVIEW 6 of 12
Materials 2020, 13, 3194 6 of 12

3.2. Stress–Strain Behavior

3.2. Stress–Strain Behavior
Figure 4a displays the average compressive strength (ACS) of specimens. We observed that ACS
gradually
Figuredecreased as the
4a displays CF length
average increased. When
compressive CF length
strength was
(ACS) of30 mm, ACS We
specimens. declined to 53.65
observed that MPa,
ACS
decreasingdecreased
gradually by 6.27%asrelative to the
CF length value ofWhen
increased. specimens without
CF length wasCF
30 (57.24
mm, ACSMPa), as shown
declined in Figure
to 53.65 MPa,
4b.
decreasing by 6.27% relative to the value of specimens without CF (57.24 MPa), as shown in Figure 4b.

Figure 4. Compressive behavior: (a) Compressive strength and (b) relative reduction.
Figure 4. Compressive behavior: (a) Compressive strength and (b) relative reduction.
This effect of CF length on MPC compressive strength could be explained by the following
This effect of CF length on MPC compressive strength could be explained by the following
reasons: In MPC, the main hydration product is a compound called potassium phosphate magnesium
reasons: In MPC, the main hydration product is a compound called potassium phosphate magnesium
(MgKPO4 ·6H2 O, MKP), also known as K-struvite, which surrounds the unreacted MgO to form a
(MgKPO4·6H2O, MKP), also known as K-struvite, which surrounds the unreacted MgO to form a
crystal product, i.e., the dominant component providing the mechanical strength of MPC [21]. Figure 5
crystal product, i.e., the dominant component providing the mechanical strength of MPC [21]. Figure
presents its micro-morphology through SEM and micro-morphological schematics. Then, as shown in
5 presents its micro-morphology through SEM and micro-morphological schematics. Then, as shown
Figure 5, the MPC matrix exhibited a resistance to external compression with the help of the space
in Figure 5, the MPC matrix exhibited a resistance to external compression with the help of the space
network constituted by these interconnecting crystal products. Although adding CF can restrict the
network constituted by these interconnecting crystal products. Although adding CF can restrict the
expansion of cracks in the MPC matrix by enhancing the connections of these crystal products through
expansion of cracks in the MPC matrix by enhancing the connections of these crystal products
the bridge effect, as shown in Figure 6, it is of little use in improving the compressive strength due to
through the bridge effect, as shown in Figure 6, it is of little use in improving the compressive strength
the negligible width of the cracks before the peak load. This phenomenon is also reported in other
due to the negligible width of the cracks before the peak load. This phenomenon is also reported in
studies about fibers [14,22]. Adding CF can also lead to the decline in MKP. As shown in Figure 7,
other studies about fibers [14,22]. Adding CF can also lead to the decline in MKP. As shown in Figure
the maximum intensity of MKP decreased from 894.3 to 854.8 a.u. when CF length increased from 0
7, the maximum intensity of MKP decreased from 894.3 to 854.8 a.u. when CF length increased from
to 30 mm, allowing the MPC to lose a certain compressive strength. Additionally, the increase of CF
0 to 30 mm, allowing the MPC to lose a certain compressive strength. Additionally, the increase of
length reduced CF amount due to the constant proportion of CF in specimens. As a result, the CF per
CF length reduced CF amount due to the constant proportion of CF in specimens. As a result, the CF
unit volume in the MPC matrix decreased with increased CF length, leading not only to decreased
per unit volume in the MPC matrix decreased with increased CF length, leading not only to decreased
effectiveness of CF, but to nonuniform distribution, which lowered the compactness of the MPC matrix
effectiveness of CF, but to nonuniform distribution, which lowered the compactness of the MPC
and led to the agglomeration of overlong CF. Figure 8 offers the fracture surface of specimens, in which
matrix and led to the agglomeration of overlong CF. Figure 8 offers the fracture surface of specimens,
the aperture and porosity of specimens both enlarge with the increase of CF length. These phenomena
in which the aperture and porosity of specimens both enlarge with the increase of CF length. These
had an adverse effect on the mechanical properties of the MPC, resulting in the decrease in ACS.
phenomena had an adverse effect on the mechanical properties of the MPC, resulting in the decrease
Figure 9a depicts the stress–strain curves of specimens in three stages: an approximate elastic
in ACS.
stage (O–A), a crack’s stable propagation stage (A–B), and a softening descending stage (B–C, B–D),
as described in Figure 9b.
Materials 2020, 13, 3194 7 of 12
Materials
Materials2020,
2020,13,
13,x xFOR
FORPEER
PEERREVIEW
REVIEW 7 7ofof1212

MgO
MKP

FA

--MgO --MgKPO4•6H2O

Figure
Figure 5.5.Fracture
Figure5. Fracturesurface
surfacemicro-morphology
micro-morphologyand structure
andstructure Magnesium
structureofofMagnesium Phosphate
MagnesiumPhosphate Cement
PhosphateCement (MPC).
Cement(MPC).
(MPC).

MPC matrix

Coir fibers Coir fibers Coir fibers

Bonding
MPC stress
Tensile Tensile
matrix MPC matrix stress
stress

Crack

Bridge effect failed

CF debonding CF rupture
MPC matrix
MPC matrix MPC matrix
Slip

Coir fiber

Coir fiber Coir fiber

Pulling out

Figure
Figure 6.6.Scanning
Figure6. ScanningElectron
ElectronMicroscope
Microscope(SEM)
(SEM)images
imagesand
andsketches
sketchesshowing
showingthe
thebridge effect
bridgeeffect during
effectduring
during
crack
crack propagation.
propagation.
crack propagation.
 Materials 2020, 13, x FOR PEER REVIEW 8 of 12

Materials 2020, 13, 3194 8 of 12

Materials
Materials 2020,
2020, 13, x13,
FORx FOR PEER
PEER REVIEW
REVIEW 8 of 812of 12

Figure 7. X-ray diffraction (XRD) results for specimens: (a) XRD results and (b) intensity of
MgKPO4·6H2O (MKP).

Figure
Figure
Figure 7. X-ray
7. X-ray X-ray diffraction
diffraction
diffraction (XRD)
(XRD)(XRD) results
results for for
results for specimens:
(a) (a)
specimens:
specimens: (a) XRD
XRD XRD results
andand
results
results (b) (b)
and (b) intensity
of of
intensity
intensity of
MgKPO
MgKPO
MgKPO 4 ·
6H
4·6H42·6H O (MKP).
2 O (MKP).
O (MKP).
2

Figure 8. Fracture surface morphology.

Figure 9a depicts the stress–strain curves of specimens in three stages: an approximate elastic
stage (O–A), a crack’s stable propagation
Figure 8. stage (A–B), and a softening descending stage (B–C, B–D),
Figure
Figure 8. Fracture
8. Fracture surface
surface
Fracture morphology.
morphology.
surface morphology.
as described in Figure 9b.
Figure
Figure 9a depicts
9a depicts thethe stress–strain
stress–strain curves
curves of specimens
of specimens in three
in three stages:
stages: an an approximate
approximate elastic
elastic
stage (O–A), a crack’s stable propagation stage (A–B), and a softening descending stage (B–C,
stage (O–A), a crack’s stable propagation stage (A–B), and a softening descending stage (B–C, B–D), B–D),
as described
as described in Figure
in Figure 9b. 9b.

Figure 9. (a) Stress–strain curve and (b) curve configuration.

Figure 9. (a) Stress–strain curve and (b) curve configuration.

Figure
Figure 9. Stress–strain
9. (a) (a) Stress–strain curve
curve andand (b) curve
(b) curve configuration.
configuration.
Materials 2020, 13, 3194 9 of 12
Materials 2020, 13, x FOR PEER REVIEW 9 of 12

The
The elastic
elastic stage (O–A) can
stage (O–A) can be be considered
consideredas asan
anapproximately
approximatelylinear linearascending
ascending stage,
stage, in in which
which all
all specimens
specimens displayed
displayed a similar
a similar stress–strain
stress–strain behavior behavior under compression.
under compression. However, However,
the elasticthe elastic
modulus
modulus (E) decreased
(E) decreased from 14.74 from 14.74
to 8.89 GPato 8.89
as CFGPa as CF
length length increased
increased from 0 to from30 mm,0 toshown
30 mm, in shown
Figure in10.
Figure 10. The result is attributed to the low elastic modulus of CF which
The result is attributed to the low elastic modulus of CF which will compromise the stiffness of MPC will compromise the
stiffness
accordingoftoMPC according
the complex to the
material complex
theory [23]. Asmaterial theory
described [23]. As
in Figure described
8, the in Figure
compactness 8, the
of specimens
compactness of specimens decreased as CF length increased due to the increase
decreased as CF length increased due to the increase of porosity, which also resulted in the decline of E. of porosity, which
also
When resulted in the decline
stress increased of E.AWhen
to point as thestress increased to
load increased, point
some A as the load
micro-cracks increased,
occurred some
in the micro-
matrix due
cracks occurred in the matrix due to the local stresses at the cracks exceeding
to the local stresses at the cracks exceeding the matrix’s crack-resistance. However, the matrix still the matrix’s crack-
resistance. However,
kept its capacity due the matrix
to the still kept
cohesion its capacity
in these due to theitscohesion
cracks, although in these cracks,
elastic modulus degraded although its
gradually
elastic
as the modulus degraded
load increased. gradually
In this as the for
stage (A–B) loadtheincreased.
specimens In this
withstage (A–B)
CF, the for the specimens
propagation with
of the cracks
CF, the propagation of the cracks was delayed by CF through its bridge
was delayed by CF through its bridge effect, which depended on CF’s tensile capacity and its bondingeffect, which depended on
CF’s tensile capacity and its bonding performance with the matrix,
performance with the matrix, as shown in Figure 6. This resistance was more obvious for longer as shown in Figure 6. This
resistance was more
CF. As in Figure 9, theobvious
range offor A–B longer
and the CF.strain
As in Figure 9, theto range
corresponding the peakof stress
A–B and the strain
increased with
corresponding to the peak stress increased with the increase of CF length.
the increase of CF length. The variation of secant modulus (Es ) could also be used to demonstrate The variation of secant
modulus (Es) could
the restrictive effectalso
of CFbe length
used toon demonstrate
micro-cracks. the Erestrictive effect of CF length on micro-cracks. Es
s decreased from 14.14 to 8.40 GPa as CF length
decreased from 14.14 to 8.40 GPa
increased from 0 to 30 mm (Figure 10). as CF length increased from 0 to 30 mm (Figure 10).

Figure 10.
Figure Elasticand
10. Elastic andsecant
secantmodulus.
modulus.

After point B, the stress–strain behavior entered the softening stage. This stage showed different
After point B, the stress–strain behavior entered the softening stage. This stage showed different
configurations at different CF lengths. For specimens with CF less than 5 mm, stress is dropping
configurations at different CF lengths. For specimens with CF less than 5 mm, stress is dropping
sharply to point C until the specimens failed. When the CF was longer than 10 mm, the stress of
sharply to point C until the specimens failed. When the CF was longer than 10 mm, the stress of
specimens decreased slowly to point D, and the specimens still showed notable residual strength and a
specimens decreased slowly to point D, and the specimens still showed notable residual strength and
higher ultimate strain. These phenomena suggest that a shorter CF (less than 5 mm in this study) has a
a higher ultimate strain. These phenomena suggest that a shorter CF (less than 5 mm in this study)
negligible effect on the stress–strain behavior of MPC due to its slight bonding capacity. Figure 7 shows
has a negligible effect on the stress–strain behavior of MPC due to its slight bonding capacity. Figure
that most CF was pulled out quickly rather than being ruptured for group CFL5, allowing cracks to
7 shows that most CF was pulled out quickly rather than being ruptured for group CFL5, allowing
extend rapidly until the breakdown of the specimens. However, as CF length increased, the bonding
cracks to extend rapidly until the breakdown of the specimens. However, as CF length increased, the
capacity improved, so that the CF had sufficient bonding force with the MPC, to allow the CF to stay in
bonding capacity improved, so that the CF had sufficient bonding force with the MPC, to allow the
the matrix and restrict these expanding cracks through its bridge effect. Those specimens still had an
CF to stay in the matrix and restrict these expanding cracks through its bridge effect. Those specimens
ability to balance load, although their capacity declined with continuous loading. These phenomena
still had an ability to balance load, although their capacity declined with continuous loading. These
also illustrated that the ductility of specimens increased in accordance with the increase of CF length.
phenomena also illustrated that the ductility of specimens increased in accordance with the increase
of CFEnergy
3.3. length.Absorption

FigureAbsorption
3.3. Energy 11 shows the energy absorption of specimens with different CF length, which was defined
as the strain energy equaling the area under the stress–strain curve shown in Figure 9b. In this study,
Figure 11 shows the energy absorption of specimens with different CF length, which was defined
the strain at point D was limited to 7000 µε for calculating the area enclosed by stage O–B–D. Figure 11a
as the strain energy equaling the area under the stress–strain curve shown in Figure 9b. In this study,
shows that energy absorption first increased and then decreased with the increase in CF length.
the strain at point D was limited to 7000 µ ε for calculating the area enclosed3 by stage O–B–D. Figure
When the CF length was 20 mm, energy absorption increased to 230.0 kJ/m , by 77.0% relative to the
11a shows that energy absorption first increased and then decreased with the increase in CF length.
When the CF length was 20 mm, energy absorption increased to 230.0 kJ/m3, by 77.0% relative to the
Materials 2020, 13, 3194 10 of 12
Materials
Materials2020,
2020,13,
13,x xFOR
FORPEER
PEERREVIEW
REVIEW 1010ofof1212

value
value ofofthe
valueof thespecimens
the specimens without
specimenswithout CF,
withoutCF,CF,shown
shownin
shown ininFigure
Figure11b.
Figure 11b. This
11b.This phenomenon
Thisphenomenon
phenomenonwas was attributed
wasattributed
attributedto totothe
the
the
bridge
bridge effect
bridgeeffect of CF,
effectofofCF, mentioned
CF,mentioned in Section
mentionedininSection 3.2.
Section3.2. As
3.2.As in Figure
AsininFigure 6, the
Figure6,6,the matrix
thematrix was
matrixwas strengthened
wasstrengthened
strengthenedafter after adding
afteradding
adding
CF,
CF,benefiting
CF, benefitingfrom
benefiting fromthe
from theconnection
the connection
connectionof ofofthe
theCF
the CFfor
CF forits
for itsinternal
its internalcrystal
internal crystalproducts.
crystal products.
products.IfIfIfcracks
cracks
cracksinininthe
the matrix
thematrix
matrix
developed
developed continuously,
developedcontinuously,
continuously,thethe CF
theCF eventually
CFeventually slipped
eventuallyslipped
slippedoutout of the
outofofthe matrix,
thematrix, whether
matrix,whether through
whetherthrough breaking
throughbreaking their
breakingtheir their
bonding
bonding force
force or
or being
being ruptured,
ruptured, as
as shown
shown in
in Figure
Figure 6.
6. Whatever
Whatever resulted,
resulted,
bonding force or being ruptured, as shown in Figure 6. Whatever resulted, debonding or rupture, debonding
debonding or
or rupture,
rupture,
energy
energy was
energywas consumed
wasconsumed
consumedand and thus
andthus would
thuswould increase
wouldincrease with
increasewith withthethe increase
theincrease ininCF
increasein CF length,
CFlength, due
length,due totothe
dueto the increased
theincreased
increased
bonding
bonding force.
bondingforce.
force.

Figure
Figure 11.
Figure11. (a)
11.(a) Energy
(a)Energy absorption
Energyabsorption and
absorptionand (b)
and(b) relative
(b)relative increase
relativeincrease
increaseofofthe
theenergy
energyabsorption.
absorption.

However,
However,
However,as asasdescribed
described
describedin ininSection
Section 3.2
Section3.2 (Figure
3.2(Figure
(Figure8),8), the
8),the CF
theCF per
CFper unit
unitvolume
perunit volume decreased
volumedecreased with
decreasedwithwiththethe
the
increase
increase in
in CF
CF length
length due
due to
to the
the constant
constant CF
CF proportion
proportion in
in the
the MPC,
MPC, leading
leading
increase in CF length due to the constant CF proportion in the MPC, leading to the decrease in its to
to the
the decrease
decrease in
in its
its
effectiveness.
effectiveness. Overlong
effectiveness.Overlong
OverlongCF CF (longer
CF(longer than
(longerthan 20
than20 mm
20mm
mminininthis
this study)
thisstudy) often
oftentwined
study)often twinedtogether
twined togetherand
together andweakened
and weakened
weakened
mortar
mortar fluidity,
mortarfluidity, which
fluidity,which had
whichhadhadanan adverse
anadverse effect
adverseeffect on
effecton CF
onCF effectiveness,
CFeffectiveness, matrix
effectiveness,matrix compactness,
matrixcompactness,
compactness,and and CF–MPC
andCF–MPC
CF–MPC
bonding
bonding performance. CL30 obviously presented a looser structure around
bonding performance. CL30 obviously presented a looser structure around CF than CL20 (Figure12).
performance. CL30 obviously presented a looser structure around CF
CF than
than CL20
CL20 (Figure
(Figure 12).
12).
For
For that
Forthat reason,
thatreason, the
reason,the energy
theenergy absorption
energyabsorption of specimens
absorptionofofspecimens declined
specimensdeclined
declinedwhenwhen
whenCFCF was
CFwas longer
waslonger than
longerthan a threshold
thana athreshold
threshold
(20 3 , and its relative increment
(20mm
(20 mmmmininthis study).
inthis
this study).Energy
study). Energy absorption
Energy decreased
absorption
absorption from 230.0
decreased
decreased fromto
from 206.2tokJ/m
230.0
230.0 to206.2
206.2 kJ/m
kJ/m3,3 and its relative
, and its relative
dropped
increment from 77.0% to 58.7% as CF length increased from 20 to 30 mm
increment dropped from 77.0% to 58.7% as CF length increased from 20 to 30 mm (Figure10).
dropped from 77.0% to 58.7% as CF length increased from 20(Figure
to 30 mm 10).
(Figure 10).

MPC
MPCmatrix
matrix
Coir
Coirfiber
fiber
Loose
Loosestructure
structure
Compacter
Compacterinterface
interface
Coir
Coirfiber
fiber
MPC
MPCmatrix
matrix

(a)
(a) (b)
(b)
Figure
Figure12.
12.SEM
SEMresults
SEM resultsofofboding
bodinginterface:
interface:(a)
(a)CFL20
(a) CFL20and
CFL20 and(b)
and (b)CFL30.
(b) CFL30.
CFL30.

4.4.Conclusions
Conclusions
AAtypical
typicalcompressive
compressivetest
testwas
wasperformed
performedtotostudy
studythe
theeffect
effectofofCF
CFlength
lengthon
onthe
thecompressive
compressive
strength,
strength, failure pattern, stress–strain curve, and energy absorption of MPC. Our conclusionsare
failure pattern, stress–strain curve, and energy absorption of MPC. Our conclusions are
summarized
summarizedasasfollows:
follows:
Materials 2020, 13, 3194 11 of 12

4. Conclusions
A typical compressive test was performed to study the effect of CF length on the compressive
strength, failure pattern, stress–strain curve, and energy absorption of MPC. Our conclusions are
summarized as follows:
(1) Adding CF to MPC could effectively increase ductility. When the length of the CF was longer
than 10 mm, the failure pattern changed from brittle to ductile.
(2) The addition of CF caused a 6.27% decrease in compressive strength when CF length increased
from 0 to 30 mm. The elastic and secant modulus of specimens presented a similar trend as
compressive strength, and the plastic properties of MPC showed remarkable improvement.
(3) Adding CF also improves the energy absorption of the MPC, which means increasing the
toughness of the MPC. However, this improvement is restricted by CF length. When the length
increased to 20 mm, energy absorption reached the maximum value of 230.0 kJ/m3 and increased
by 77.0% relative to the value of specimens without CF. Then, it decreased in turn as CF length
increased continuously.

Author Contributions: Conceptualization, L.Z. and Z.J.; methodology, L.Z. and T.G.; validation, L.Z., Z.J. and T.G.;
investigation, Z.J., T.G., Y.L., W.Z., and M.H.; data curation, Z.J., Y.L., and W.Z.; writing—original draft preparation,
Z.J.; writing—review and editing, L.Z. and Z.J.; supervision, L.Z.; project administration, L.Z. All authors have
read and agreed to the published version of the manuscript.
Funding: This work was supported by the Graduate Student Innovation Ability Cultivation Fund of Guangzhou
University [grant number 2019GDJC-M38]; the National Natural Science Foundation of China [grant numbers
51608137]; and Innovation Training Project of Guangzhou University [grant number CX2019329].
Acknowledgments: The authors of these works gratefully acknowledge language revision provided by Yufang Rao.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Roy, D.M. New Strong Cement Materials: Chemically Bonded Ceramics. Science 1987, 235, 651–658.
[CrossRef] [PubMed]
2. Abdelrazig, B.E.I.; Sharp, J.H.; El-Jazairi, B. The chemical composition of mortars made from magnesia
phosphate cement. Cem. Concr. Res. 1988, 18, 415–425. [CrossRef]
3. Qiao, F.; Chau, C.K.; Li, Z. Property evaluation of magnesium phosphate cement mortar as patch repair
material. Constr. Build. Mater. 2010, 24, 695–700. [CrossRef]
4. Yang, Q.B.; Zhu, B.R.; Zhang, S.Q. Properties and applications of magnesia-phosphate cement mortar for
rapid repair of concrete. Cem. Concr. Res. 2000, 30, 1807–1813. [CrossRef]
5. Feng, H.; Chen, G.; Gao, D. Mechanical Properties of Steel Fiber-Reinforced Magnesium Phosphate Cement
Mortar. Adv. Civ. Eng. 2018, 3, 1–11. [CrossRef]
6. Ahmad, M.R.; Chen, B. Effect of silica fume and basalt fiber on the mechanical properties and microstructure
of magnesium phosphate cement (MPC) mortar. Constr. Build. Mater. 2018, 190, 466–478. [CrossRef]
7. Fang, Y.; Cui, P.; Ding, Z.; Zhu, J.X. Properties of a magnesium phosphate cement-based fire-retardant coating
containing glass fiber or glass fiber powder. Constr. Build. Mater. 2018, 162, 553–560. [CrossRef]
8. Ahmad, M.R.; Chen, B.; Haque, M.A.; Shah, S.F.A. Development of a sustainable and innovant hygrothermal
bio-composite featuring the enhanced mechanical properties. J. Clean. Prod. 2019, 229, 128–143. [CrossRef]
9. Lecompte, T.; Perrot, A.; Subrianto, A. A novel pull-out device used to study the influence of pressure during
processing of cement-based material reinforced with coir. Constr. Build. Mater. 2015, 78, 224–233. [CrossRef]
10. Paul, A.; Thomas, S. Electrical properties of natural-fiber-reinforced low-density polyethylene composites:
A comparison with carbon black and glass-fiber-filled low-density polyethylene composites. J. Appl. Polym.
Sci. 1997, 63, 247–266. [CrossRef]
11. Ali, M.; Liu, A.; Sou, H.; Chouw, N. Mechanical and dynamic properties of coconut fiber reinforced concrete.
Constr. Build. Mater. 2012, 30, 814–825. [CrossRef]
12. Ali, M.; Li, X.Y.; Chouw, N. Experimental investigations on bond strength between coconut fiber and concrete.
Mater. Des. 2013, 44, 596–605. [CrossRef]
Materials 2020, 13, 3194 12 of 12

13. Baruah, P.; Talukdar, S. A comparative study of compressive, flexural, tensile and shear strength of concrete
with fibres of different origins. Indian Concr. J. 2007, 81, 17–24.
14. Thanushan, K.; Yogananth, Y.; Sangeeth, P.; Coonghe, J.G.; Sathiparan, N. Strength and Durability
Characteristics of Coconut Fiber Reinforced Earth Cement Blocks. J. Nat. Fibers 2019, 1–16. [CrossRef]
15. Li, Z.J.; Wang, L.J.; Wang, X.G. Flexural Characteristics of Coir Fiber Reinforced Cementitious Composites.
Fiber. Polym. 2006, 7, 286–294. [CrossRef]
16. Sekar, A.; Kandasamy, G. Optimization of Coconut Fiber in Coconut Shell Concrete and Its Mechanical and
Bond Properties. Materials 2018, 11, 1726. [CrossRef]
17. Hwang, C.L.; Tran, V.A.; Hong, J.W. Effects of short coconut fiber on the mechanical properties, plastic cracking
behavior, and impact resistance of cementitious composites. Constr. Build. Mater. 2016, 127, 984–992. [CrossRef]
18. Fang, Y.; Chen, B.; Oderji, S.Y. Experimental research on magnesium phosphate cement mortar reinforced by
glass fiber. Constr. Build. Mater. 2018, 188, 729–736. [CrossRef]
19. ASTM C109-16. Standard Test Method for Compressive Strength of Hydraulic Cement Mortars; American Society
for Testing and Materials, ASTM International: West Conshohocken, PA, USA, 2016.
20. ASTM C469-14. Standard Test Method for Static Modulus of Elasticity and Poisson’s Ratio of Concrete in Compression;
American Society for Testing and Materials, ASTM International: West Conshohocken, PA, USA, 2014.
21. Xu, B.W.; Lothenbach, B.; Leemann, A.; Winnefeld, F. Reaction mechanism of magnesium potassium phosphate
cement with high magnesium-to-phosphate ratiod. Cem. Concr. Res. 2018, 108, 140–151. [CrossRef]
22. Estabragh, A.R.; Ranjbari, S.; Javadi, A.A. Properties of Clay Soil and Soil Cement Reinforced with
Polypropylene Fibers. ACI Mater. J. 2017, 114, 195–205. [CrossRef]
23. Swamy, P.A.V.B. Efficient inference in a random coefficient regression model. Econometrica 1970, 38,
311–323. [CrossRef]

© 2020 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 (http://creativecommons.org/licenses/by/4.0/).