Professional Documents
Culture Documents
To cite this article: S. Papaioannou, R. Argyropoulou, C. Tachos & N. Markantonis (2015) The
Adhesion Properties of Mortars in Relation with Microstructure, The Journal of Adhesion, 91:7,
505-517, DOI: 10.1080/00218464.2014.931228
1. INTRODUCTION
Masonry construction exists since ancient times and remains a large part of
the building stock in the world as it presents a number of advantages such
505
506 S. Papaioannou et al.
as durability, thermal and sound insulation, and great flexibility of plan form
[1]. Among the various parameters for the construction of a healthy masonry,
the interaction between masonry units and mortar is considered to be of
major importance. The most important factor is the adhesion between
masonry units and mortar, as poor adhesion leads to defective strength
and impermeability [2]. It has been reported that the initial shear strength
of the masonry depends on the mortar=unit adhesion [3] while the flexural
strength is directly related to bond strength [4]. Therefore, the understanding
of the interface behavior is crucial for design and assessment purposes [5].
The adhesion bond results from complex physicochemical processes during
the curing of the binding agent. Qualitatively, the adhesion is described by
the specific work per unit area needed to separate the joint elements [6].
There are a number of factors that influence the mortar=unit adhesion, i.e.,
the kind of units (surface characteristics, absorbency, etc.), the kind of mortar
(workability, water retentivity, etc.), and the application technique (filling of
joints, degree of pressure on units, etc.) [2,3]. With regard to mortars and
units, it is reported that their mineralogical composition and internal structure
affect the masonry joint strength. Thus, it is apparent that jointing materials
with different compositions and porous system will produce different physi-
cal–mechanical behavior [4]. Up to now, considerable work has been done
on the characterization of mortars from historic buildings [7,8] or for resto-
ration purposes [9], while the majority of research on the mechanical beha-
vior of modern masonry is mainly focused on the evaluation of shear and
flexural strength [3,4,6,10,11] with various mortars and units and on the mod-
eling of interface behavior and fracture mechanism [5,12–15], while limited
results concerning the microscopic level have been published [4].
In this work different cement–lime mortars were produced. The work-
ability of the fresh mortars and the compressive and flexural strength of the
hardened mortars were measured. The adhesion of these mortars to perfor-
ated clay bricks was qualitatively estimated with a simplified tensile=tear
testing measurement. The mortars were characterized by means of X-ray dif-
fraction (XRD) and scanning electron microscopy (SEM). The mortar=unit
interface was studied with stereoscopy. Finally, the results of the microstruc-
ture observations were associated with those from the adhesion test.
2.1. Materials
For the preparation of mortars cement CEM IV=B (P-W) 32.5 R (Lafarge
Industry, Volos, Greece), lime putty and sand were mixed in different
ratios. Two types of aggregates were used: washed river sand from the area
(Serres, Macedonia, Greece) and crushed stone sand for mortars provided
from a quarry in the area. The choice not to use ready mixed mortars was
Adhesion Properties of Mortars and Microstructure 507
based on the fact that there are not such production units in many places in
Greece.
The moisture content was 2.8% for the river sand and 2.7% for the
crushed stone. The grading of the sand is presented in Table 1.
#4 100 100
#2 93 67.5
#1 71.6 39.4
# 0.25 3.7 17.9
63 mic 0.08 1
508 S. Papaioannou et al.
the mortar=unit interface. The ultimate strength is the maximum tensile force
required for the debonding per unit area of contact. This idea was followed
for the estimation of the adhesion. Metallic cylinders were inserted in the
holes of the bricks. The cylinders were then adapted to the grips of the testing
machine (INSTRON KN1200). Tensile forces were applied perpendicular to
the mortar=unit contact area. As samples were brittle, in order to obtain accu-
rate data, loading was kept slow by keeping the elongation rate constant at
1 mm=min. A photo of the test setup is presented in Fig. 1(a) and a simple
drawing of the test setup is depicted in Fig. 1(b). All measurements were
performed in a period of 32–35 days after the preparation of the mortars.
The qualitative characterization of adhesion test was as following:
FIGURE 1 (a) Photo of the adhesion test setup. (b) Simple scheme of the adhesion test setup.
Adhesion Properties of Mortars and Microstructure 509
In Table 2 the mixing ratios and the properties of mortars are presented. In the
table the standard deviation (s) for the flexural and compression test for each
mortar, is also depicted. The mixing ratio is in the form of the % weight
cement=lime=sand. Mortars 1–6 are made with river sand. Mortars 7–11 are
made with crushed stone sand. The water=cement (w=c) ratio was 0.7 for
the Mortars M1–M11. Preliminary tests with mortars, namely, Mortar A and
TABLE 2 Mixing Ratio and Properties of Mortars
Mixing Type of Flow Initial Final set Flexural Compressive Adhesion Adhesion
Mortar ratio aggregates (mm) set time (h) time (h) strength (MPa) strength (MPa) (qualitative data) strength (MPa)
Mortar A 1:1:6 River sand 100 2 24 2.5 s ¼ 0.22 12.5 s ¼ 0.83 Zero 0
Mortar B 1:1:6 River sand 117 1:45 24 2.3 s ¼ 0.14 8.1 s ¼ 0.40 Zero 0
Mortar 1 1:1:5 River sand 153 2:20 24 2.6 s ¼ 0.06 8.4 s ¼ 0.53 Excellent >0.15
Mortar 2 1:1:6 River sand 129 2 24 1.7 s ¼ 0.01 5.9 s ¼ 0.55 Good 0.09
Mortar 3 1:1:8 River sand 105 1:55 24 1.5 s ¼ 0.02 6.9 s ¼ 0.48 Poor 0.005
510
Mortar 4 1:2:6 River sand 155 3:25 24 1.0 s ¼ 0.08 3.1 s ¼ 0.37 Very good 0.092
Mortar 5 1:2:7 River sand 152 3:15 24 1.0 s ¼ 0.04 3.6 s ¼ 0.49 Good 0.061
Mortar 6 1:2:8 River sand 130 3:05 24 1.0 s ¼ 0.03 3.9 s ¼ 0.63 Good 0.061
Mortar 7 1:1:5 Crushed stone sand 169 2:20 24 2.0 s ¼ 0.02 9.4 s ¼ 0,72 Good 0.059
Mortar 8 1:1:6 Crushed stone sand 157 2:10 24 2.6 s ¼ 0.04 8.9 s ¼ 0.36 Good enough 0.052
Mortar 9 1:1:7 Crushed stone sand 150 1:45 24 1.5 s ¼ 0.05 6.1 s ¼ 0.45 Poor 0.028
Mortar 10 1:2:8 Crushed stone sand 155 2:30 24 2.7 s ¼ 0.01 12.0 s ¼ 0.63 Very good 0.108
Mortar 11 1:2:9 Crushed stone sand 148 2:45 24 2.9 s ¼ 0.06 13.4 s ¼ 0.70 Very good 0.123
s: Standard deviation.
Adhesion Properties of Mortars and Microstructure 511
Mortar B, showed that with ratios <0.7 the mortars had not adequate
workability resulting in zero adhesion with bricks. In particular, Mortar A
and Mortar B were prepared with a w=c ratio of 0.5 and 0.6, respectively.
From the results shown in Table 2, it can be said that in general for the
same type of sand and for the same cement=lime ratio an increase in the
aggregates content results in deterioration of workability, decrease of the
initial setting time, and deterioration of adhesion. With river sand and for
1:1 cement=lime ratio an increase of the sand content leads to a decrease
of flexural strength, but no evidence for compressive strength correlation is
found. More specifically, in comparison with Mortar 1, Mortar 2 presents a
35% decrease in flexural strength and a 30% decrease in compressive
strength. For Mortar 3 there is a 42% decrease in flexural strength and an
18% decrease in compressive strength, in comparison with Mortar 1. On
the contrary, with river sand and for the 1:2 cement=lime ratio an increase
of the sand content leads to an increase of the compressive strength (16%
for Mortar 5 and 26% for Mortar 6, in comparison with Mortar 4) while the
flexural strength remains constant. On the other hand, with crushed stone
sand and for 1:1 cement=lime ratio an increase of the sand content leads
to a decrease of compressive strength, while for the flexural strength such
relationship is not observed. More specifically, compared to Mortar 7, Mortars
8 and 9 showed a 5% and 35% decrease in compressive strength, respect-
ively. Concerning the flexural strength, Mortar 8 presented a 30% increase
in comparison with Mortar 7, while Mortar 9 showed a 25% decrease in com-
parison with Mortar 7. Therefore, it is evident that not only the mixing ratio of
raw materials, but also the microstructure of mortars, e.g., grain size, porosity,
and porous size, affect the mechanical behavior of mortars. For the same
cement=lime ratio use of crushed stone sand results generally in higher
values of strength than those obtained with river sand. For example, Mortar
8 presents 53% higher flexural strength and 51% higher compressive strength
than Mortar 2. The increase in strength can be attributed to the fact that the
crushed stone sand has grains with higher roughness than those of river sand
and it also presents a continuous grain-size distribution [2], while the river
sand has a large amount of material with grain size <1 mm and a small
amount of fine material (<0.25 mm) as depicted in Table 1. It has been
reported [3] that adhesion is highly affected by the workability of mortars
and the water absorbance of masonry units. This is evident for Mortars A
and B, taking into account that in this work moistened bricks of the same
quality were used. Moreover, by comparing workability and adhesion for
mortars made with the same raw materials and same cement=lime ratio,
e.g., Mortars 1–3, it is clear that a decrease in workability results in deterio-
ration of adhesion. On the other hand, mortars made with different raw
materials or mixing ratios as, for example, Mortars 1, 8, and 10 presented
similar values of workability but different adhesion. Therefore, the value of
workability alone is not an indicator for adhesion.
512 S. Papaioannou et al.
By the XRD measurements it was found that the river sand consists
mainly of quartz with the presence of albite in a small percentage, while
the crushed stone sand consists mainly of calcite with the presence of anker-
ite in a small percentage. In mortars made with river sand, quartz was the
main phase, followed by calcite, portlandite, C2S, and traces of other phases.
In mortars with crushed stone sand calcite was the main phase detected. In
Fig. 3 the XRD spectra of Mortar 9 is presented.
In Fig. 4(a)–(d) the SEM micrographs of Mortars 1, 2, 9, and 11 are
demonstrated. In these figures dark spots and bright spots are indicated with
white and black arrows, respectively. The EDX analysis of mortars is pre-
sented in Table 3 (data shown as wt%). Traces of Na, P, S, Cl, and Ti were also
detected in the samples. Mortar 1 (Fig. 4(a)) presents a large grain-size distri-
bution with grain size varying from 10 to 50 mm. Mortars 9 and 11 have more
dense and uniform grains with size 1–10 mm. In general, mortars made with
river sand have larger grain size than those with crushed stone sand and this
can explain why mortars with crushed stone sand presented higher values of
strength. The results are in agreement with findings from historic mortars,
where mortars with higher quartz content were more coarsely crystallized [7].
SEM observations of the brick=mortar interfaces are presented in
Fig. 5(a)–(d). In these figures the bright grains are indicated with black arrows
and the gray matrix is indicated with white arrows. As it can be observed, on
the surface of the bricks grains of the mortar have been embedded. This is evi-
dent for Mortar 1=brick and Mortar 11=brick interfaces (Fig. 5(a) and (d),
respectively), where many bright crystals can be observed on the brick
FIGURE 4 (a) SEM micrograph of Mortar 1. White arrow indicates a dark spot and black
arrow indicates a bright spot. (b) SEM micrograph of Mortar 2. White arrow indicates a dark
spot and black arrow indicates a bright spot. (c) SEM micrograph of Mortar 9. White arrow
indicates a dark spot and black arrow indicates a bright spot. (d) SEM micrograph of Mortar
11. White arrow indicates a dark spot and black arrow indicates a bright spot.
surface and these mortars presented the best adhesion. On the contrary, on
the Mortar 9=brick interface it can be seen that bright grains are sparsely
embedded on the brick surface. The EDX analysis (wt%) of the grains is
presented in Table 4.
In a study with concrete units [4] it was found that strong bond formations
are attributed to complex Ca–Al–Si phases within the block binder. Moreover,
in historic mortars the mechanical properties can be due to alkali–silicate reac-
tions at the brick fragment–lime interface [8]. On the other hand, in a study of
the adhesion of mortars made with cement, organic additives, and different
514 S. Papaioannou et al.
kind of aggregates [19] it resulted that mortars with no CaO presented lower
strength. The above studies indicate that regardless of the type of masonry
unit and mortar used, adhesion is mainly attributed to Ca–Si complex reac-
tions. These results are in agreement with the results of the present study.
The bright crystals that seem to contribute to the adhesion consist mainly of
Ca–Si–Al, while adhesion improves by the number of these crystals. From
the results in Table 4 it results that the matrix on the brick=mortar interface
comprises Al–Si compositions with low Ca content and this is in agreement
with [4]. From the brick=Mortar 1 surface observations with stereoscopy it
was found that almost the whole surface of the brick is covered with mortar.
On the brick=Mortar 11 surface, it was observed that the mortar is spread to a
lesser extend on the brick surface. The same remarks hold for the
brick=Mortar 2 interface which presented less adhesion. For historic mortars
it is reported [7] that lime penetrates into the ceramic and causes reactions that
transform the microstructure of the ceramic imparting high strength. In [4] the
Ca–Al–Si phases were attributed to the use of fly ash in concrete blocks. From
the results of the present work, the adhesion seems to be highly affected by
the presence of Ca–Al–Si phases in the brick=mortar interface. These could
be formed from complex reactions among cement, lime, and aggregates or
could be partly attributed to fly ash in the pozzolanic cement used. By com-
paring the results shown in Table 3, it can be noticed that Mortars 9 and 11
are richer in Ca content than Mortars 1 and 2, but exhibit lower concentration
in Si and Al than Mortars 1 and 2 and this can be expected as Mortars 9 and 11
are made with crushed stone sand. On the other hand, in the brick=mortar
interface the grains of Mortars 9 and 11 have higher concentrations in Al
and Si, which can be attributed to the presence of the brick. Moreover, from
the results in Table 4, it is noticed that in bright grains of Mortars 1, 2, and 11
there is a similar Ca=Si=Al ratio. More specifically the Ca=Si=Al ratio in Mortars
1, 2, and 11 is 1=0.5=0.108, 1=0.49=0.09, and 1=0.45=0.14, respectively. In
Mortar 9 which presented a poor adhesion the bright grains on the interface
are deficient in Si and the Ca=Si=Al ratio is 1=0.3=0.16. Moreover, dark grains
in Mortars 1 and 11 present similar Si=Al=Ca ratio (1=0.49=0.14 and
Adhesion Properties of Mortars and Microstructure 515
FIGURE 5 (a) SEM micrograph of Mortar 1=brick interface. Black arrow indicates a bright
grain and white arrow indicates a dark grain. (b) SEM micrograph of Mortar 2=brick interface.
Black arrow indicates a bright grain and white arrow indicates a dark grain. (c) SEM micro-
graph of Mortar 9=brick interface. Black arrow indicates a bright grain and white arrow indi-
cates a dark grain. (d) SEM micrograph of Mortar 11=brick interface. Black arrow indicates a
bright grain and white arrow indicates a dark grain.
ii. the formation of fine Ca–Al–Si phases which can penetrate in the brick
providing adhesion.
4. CONCLUSIONS
The adhesion bond of different mortars has been tested and it was associated
with their microstructure. The adhesion bond between mortar and masonry
units results from physicochemical processes taking place during the mortar
curing and the extend of these processes affect the extend of the bond. The
adhesion depends greatly on the formation of Ca–Al–Si phases in a Si–Al
matrix in the masonry unit=mortar interface. In the matrix, the Si=Al ratio that
favors adhesion is in the range 1=0.5, with a low Ca content (Si=Ca: in the
range 1=0.15). The fine Ca–Al–Si phases that penetrate into the brick should
have a Ca=Si=Al ratio in the range 1=0.45–0.5=0.09–0.15. The more fine
Ca–Al–Si phases are formed in a well-structured matrix the stronger adhesion
is achieved.
FUNDING
The authors would like to thank the Research Committee of the Technologi-
cal Educational Institute of Central Macedonia for the financial support under
the Grant SAT=CE=050613-104=01.
REFERENCES
[5] Ghiassi, B., Oliveira, D. V., Lourenco, P. B., and Marcari, G., Compos. Part B 46,
21–30 (2013).
[6] Nemiroskii, B. A., Gur’ev, V. V., and Sporykhin, B. B., Corros. Non-metallic
Mater. 46 (7), 33–36 (2010).
[7] Moropoulou, A., Bakolas, A., and Bisbikou, K., Constr. Build Mater. 14, 35–46
(2000).
[8] Moropoulou, A., Tsiourva, Th., Bisbikou, K., Biscontint, G., Bakolas, A., and
Zendrit, E., Constr. Build Mater. 10 (2), 151–159 (1996).
[9] Lanas, J., Sirera, R., and Alvarez, J. I., Cem. Concr. Res. 36, 961–970 (2006).
[10] Alecci, V., Fagone, M., Rotunno, T., and De Stefano, M., Constr. Build Mater. 40,
1038–1045 (2013).
[11] Abdou, L., Ami Saada, R., Meftah, F., and Mebarki, A., Mech. Res. Commun. 33,
370–384 (2006).
[12] Augenti, N. and Parisi, F., Constr. Build Mater. 25, 1612–1620 (2011).
[13] Foucha, F., Lebon, F., and Titeux, I., Constr. Build Mater. 23, 2428–2441 (2009).
[14] Giambanco, G., Rizzo, S., and Spallino, R., Comput. Methods Appl. Mech. Eng.
190, 6493–6511 (2001).
[15] Naderi, M. and Ghodousian, O., J. Adhes. 88 (10), 848–865 (2012).
[16] EN 1015–3: Methods of test for mortar for masonry-Part 3: Determination of con-
sistence of fresh mortar (by flow table), 1999.
[17] EN 196–3: Methods of testing cement-Part 3: Determination of setting times and
soundness, 2005.
[18] EN 1015–11: Methods of test for mortar for masonry–Part 11: Determination of
flexural and compressive strength of hardened mortar, 2000.
[19] Ozkahraman, H. T. and Isik, E. C., Constr. Build Mater. 19, 251–255 (2005).