Magazine of Concrete Research, 2008, 60, No.

5, June, 329–337
doi: 10.1680/macr.2007.00024

Development of a laser roughness analyser to

predict in situ the bond strength of concrete-to-
concrete interfaces
P. M. D. Santos and E. N. B. S. Júlio

University of Coimbra—Polo II

The bond strength at the interface between concrete layers cast at different times is important to ensure the
monolithic behaviour of reinforced concrete composite elements, for example precast beams with cast-in-place slabs
and bridge decks strengthened with a new concrete layer. Previous studies by the current authors stated that the
bond strength is highly influenced by the surface roughness of the concrete substrate. Further studies proved that it
is possible to quantify the surface roughness and to correlate this with the bond strength of the interface. Neverthe-
less, the adopted methodology is work intensive and partially destructive. This paper describes a subsequent study
using a laser roughness analyser that was specifically developed to characterise the roughness of the concrete
substrate. The proposed new method presents four major advantages: (a) increased accuracy; (b) it is simple and
fast to execute; (c) it implies a really non-destructive procedure; and (d) results can be assessed in situ.

Introduction
Several precast reinforced concrete (RC) elements
include cast-in-place parts. Rehabilitation of RC struc-
tures generally implies repairing and/or strengthening
operations that comprise adding new to old concrete. In
both cases, the behaviour of the resulting composite
RC element depends on the bond strength of the con-
crete-to-concrete interface.
Design codes, such as Eurocode 2,1 ACI 3182 and
CSA A23.3,3 present design expressions to determine
the longitudinal shear strength of the interface between
parts of elements cast at different times. These expres-
sions are based on the shear-friction theory and the
longitudinal shear strength is evaluated considering ba-
sically four fundamental parameters: (a) compressive
strength of the weakest concrete; (b) normal stress at
the interface; (c) amount of reinforcement crossing the
interface; and (d) roughness of the substrate surface.
Common to all these design codes is the qualitative
evaluation of the roughness, being the surface rough-
ness usually classified as very smooth, smooth, rough
Department of Civil Engineering, University of Coimbra—Polo II,
3030-290 Coimbra, Portugal
(MACR-D-07-00024) Paper received 16 February 2007; last revised
17 May 2007; accepted 8 June 2007
or very rough. Depending on the roughness of the con-
crete surface, the values of two factors—friction and
cohesion—are given to be used in the design expres-
sions. This approach is clearly inaccurate since it de-
pends on a subjective assessment of the roughness.
To help in the classification of the roughness, some
design codes link this to typical finishing of concrete
surfaces, which include: surface left as-cast against
steel, plastic or wooden formwork; (free) surface left
without further treatment after vibration; and surface
intentionally roughened, using different methods. This
aid presents obvious limitations and can lead to incor-
rect values of friction and cohesion. For instance, the
reference to ‘surface intentionally roughened’, as in
ACI 318,2 is evidently insufficient since different tech-
niques can be used with that purpose. The exhaustive
consideration of all these is not the development needed
since different results can be obtained with the same
method. Considering the sand-blasting technique to il-
lustrate this idea, it is evident that the obtained surface
roughness depends on: the substrate concrete; the type
and fineness of the adopted sand; the characteristics of
the equipment used; the distance between the needle
and the substrate; the treatment application time; etc.
Therefore, an improved design expression has to con-
sider the quantification of the surface roughness instead
of considering the methods used to produce it.
329

www.concrete-research.com 1751-763X (Online) 0024-9831 (Print) # 2008 Thomas Telford Ltd

Santos and Júlio

There are no generally accepted methods to measure structive technique that can be used in situ. Following
the roughness of a concrete surface. Some attempts of this decision, and considering that there is not a speci-
standardisation include: the definition of standard pro- fic commercial device for assessing the roughness of
files of increasing roughness by the International Con- concrete surfaces, a portable laser roughness analyser
crete Repair Institute (ICRI); and the sand patch test by was developed, Fig. 1. Results from this study are
the ASTM E 965.4 In the first case, the roughness of described in the present paper.
the concrete surface is compared with the roughness of
nine standard profiles. In the second case, the diameter Evaluation of surface roughness using the laser
of a circle produced with a defined quantity of cali- roughness analyser
brated sand placed over the substrate surface serves as
The laser roughness analyser uses two laser sensors
roughness parameter. More accurate methods to assess
with a range of 30–50 mm and 30–130 mm with a
the roughness of the substrate surface are therefore
resolution of 10 ìm and 60 ìm respectively. A linear
needed and several alternatives have been proposed.
displacement table allows the laser sensors to perform
Almost all of them are destructive, or partially destruc-
a maximum evaluation length of 220 mm. In Fig. 1(a)
tive, but there are also non-destructive techniques that
these and the remaining components can be observed
usually use optical methods.
in the interior of the device. The hardware is protected
Abu-Tair et al.5 proposed a technique that uses a set
by an aluminium box, Fig. 1(b). This portable equip-
of needles placed over the substrate surface to define
ment just needs to be connected to a laptop to assess
the roughness profile and Issa et al.6 used a modified
the roughness profile of a concrete substrate, Fig. 1(c),
version of the slit-island method using dyed water
in situ or at laboratorial or industrial facilities.
instead of a filling material. Non-destructive techniques
The control software was developed with National
using a mechanical stylus and laser profilometry were
Instruments LabView 7.1, running on Microsoft Win-
also proposed by Garbacz et al.7 and Maerz et al.8
dows 2000 or XP operating system on a laptop con-
respectively.
nected to the laser roughness analyser using a USB
Once the profile of the substrate surface has been
obtained, it can be numerically defined using roughness
and waviness parameters,9 which are based in the sur-
face characteristics, such as the height and spacing
between peaks and valleys.

Experimental study
Previous studies by the authors
The first study by the present authors10 proved that (a)

the bond strength between concrete layers cast at differ-

ent times is highly influenced by the surface roughness
of the concrete substrate. In a subsequent study,11 the
authors developed a technique to assess the roughness
profile of the substrate surface based on digital image
processing. First, extracted cores from the substrate are
cleaned with acetone; then, an epoxy resin is applied
over the surface; next, the cores are sawed crossing the
axis and the obtained surface is polished; subsequently,
(b)
a digital image of the latter is produced by scanning;
and, finally, this image is digitally processed to obtain
the roughness profile.11 Eleven roughness parameters
were determined from this profile and correlated with
the bond strength in shear and in tension. High correla-
tion coefficients were obtained, mainly with three of
the roughness parameters considered. Nevertheless, the
method used was not entirely satisfactory since it is
work intensive, partially destructive and needs to be
processed at the laboratory.
(c)
Owing to the good results obtained with the method
used in the preceding study, it was decided to develop a Fig. 1. Laser roughness analyser: (a) interior view,
new technique to overcome the disadvantages referred (b) aluminium cover box, (c) equipment, connected to a
to. The main goal consisted in developing a non-de- laptop, being used
330 Magazine of Concrete Research, 2008, 60, No. 5
 Development of a laser roughness analyser to predict in situ the bond strength of concrete-to-concrete

interface. The software was developed to control the
equipment, to assess the data and to generate an output
text file containing the coordinates of the roughness
profile. Control operations include the following possi-
bilities, Fig. 2
(a) definition of the evaluation length, useful when
readings are made on small concrete specimens
(b) selection of the laser sensor
(c) positioning the laser sensor at the start position,
necessary at the beginning of the reading opera-
tions
(d) aborting of the reading process, in case of emer-
gency.
The roughness profile is recorded and graphically re-
presented, in real time, during the readings of the con-
crete surface, Fig. 2. The graphic shows the distance
between the laser sensor and the concrete surface for
each position along the evaluation path, Fig. 2.
The obtained results are stored in a text file, with
tab-separated values, compatible with any spreadsheet
application, for future editing and treatment. Software
was developed easily to compute essential information
from these data files, Fig. 3. This application allows
the user to (a) import a text file with the coordinates of
the roughness profile, obtained from the laser rough-
ness analyser; (b) remove invalid coordinates, whenever
detected; (c) calculate the roughness parameters, to
determine the mean line and to draw the roughness
profile, Fig. 3.

Fig. 2. Graphical user interface

Fig. 3. Evaluation of roughness parameters

Magazine of Concrete Research, 2008, 60, No. 5 331
Santos and Júlio

To test the laser roughness analyser and compare this (a) the average roughness Ra
new method with the method used in the previous (b) mean peak-to-valley height Rz(DIN)
study,11 the same three different types of surface rough- (c) the maximum peak-to-valley height Rmax
ness were adopted (d ) the mean third highest peak-to-valley height R3 z
(e) the maximum third highest peak-to-valley height
(a) smooth: surface left as-cast against steel formwork,
R3 zmax
considered to serve as reference
(f) the ten points’ height Rz(ISO)
(b) rough without exposing the aggregates: surface
(g) the total roughness height R y
treated with a wire brush
(h) the root-mean-square (RMS) profile height Rq
(c) rough with aggregate exposure: surface prepared
(i) the mean peak height Rpm
with sand-blasting.
( j) the maximum peak height Rp
Ten records were produced for each surface type and (k) the mean valley depth Rvm
the corresponding roughness profiles were generated. (l ) the maximum valley depth Rv .
In Figs 4 to 6 one record is represented for each case.
Subsequently, based on these profiles, the surface
roughness was quantitatively evaluated by determining Figure 7 plots the values that the 12 roughness para-
12 roughness parameters, including the 11 considered meters assume for each surface condition, assessed
in the previous study11 with the laser roughness analyser. These and the corre-
0·06

0·04

0·02
Y: mm

0·00

⫺0·02

⫺0·04

⫺0·06
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
X: mm

Fig. 4. Roughness profile of the surface left as-cast

0·5
0·4
0·3
0·2
Y: mm

0·1
0·0
⫺0·1
⫺0·2
⫺0·3
⫺0·4
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
X: mm

Fig. 5. Roughness profile of the wire-brushed surface

0·8
0·6
0·4
Y: mm

0·2
0·0
⫺0·2
⫺0·4
⫺0·6
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
X: mm

Fig. 6. Roughness profile of the sand-blasted surface

332 Magazine of Concrete Research, 2008, 60, No. 5
 Development of a laser roughness analyser to predict in situ the bond strength of concrete-to-concrete

sponding values determined with the procedure devel-
oped in the previous study are presented in Table 1.
It should be noted that the definition of roughness
parameters such as the mean third highest peak-to-
valley height and maximum third highest peak-to-
valley height implies the manual selection of the third
highest peak-to-valley height. In the previous study,11
the present authors performed this manual selection but
concluded that this procedure is not completely objec-
tive and that further work is needed to automate this
operation. Table 1 presents the values of these rough-
ness parameters, assessed in the previous and in the
present studies, without manual selection.
The roughness profile of the smooth surface was
assessed using both laser sensors, with 10 ìm and
60 ìm of accuracy, to check if the latter is sufficiently
accurate to analyse smooth surfaces, such as those
obtained with concrete cast against steel formwork.
Results are presented in Fig. 8. A comparison between
the coefficients of variation of the roughness para-
meters obtained with each sensor is shown in Fig. 9.
Evaluation of concrete-to-concrete bond strength
The slant shear test and the pull-off test were
adopted to assess the bond strength, respectively in
shear and in tension, of the interface between the old
and the new concrete layers. In the slant shear test, a
200 3 200 3 400 mm3 prismatic specimen, with the
interface at approximately 308 to the vertical, is tested
under compression. In the pull-off test, a 200 3 200 3
200 mm3 cubic specimen is used, with the interface at
mid-height, and a core of 75 mm diameter, drilled into
the added concrete and extending 15 mm beyond the
interface into the concrete substrate, is tested in ten-
sion.
Five slant shear specimens and five pull-off speci-
mens were considered for each type of surface rough-
ness: smooth; rough without exposing the aggregates;
and rough with aggregate exposure. After casting the
concrete substrate, the surface of each specimen was
prepared with the corresponding method: no treatment;
wire brushing; and sand-blasting, respectively. A period
of 84 days was adopted between casting the substrate
and placing the added concrete. All specimens were
stored in laboratory conditions, without control of tem-
perature and relative humidity. Tests were performed 28
days after placing the added concrete.
The average compressive strength of the substrate
concrete and of the added concrete, assessed with cubic
specimens at 28 days of age, was 50.40 MPa and

1·8
1·6
1·4
Profile height: mm

1·2
1·0
0·8
0·6
0·4
0·2
0·0
Ra Rz(DIN) Rmax R3z R3zmax Rz(ISO) Ry Rq Rpm Rp Rvm Rv

⫺6 ⫺6 ⫺6
Left as-cast (10 ⫻ 10 m) Wire-brushing (60 ⫻ 10 m) Sand-blasting (60 ⫻ 10 m)

Fig. 7. Roughness parameters assessed with the laser roughness analyser

Table 1. Roughness parameters

Roughness parameter: mm Digital image treatment Laser roughness analyser

Left as-cast Wire- Sand- Left as-cast Wire- Sand-

brushing blasting brushing blasting

Average roughness, Ra 0.032 0.095 0.202 0.019 0.129 0.227

Mean peak-to-valley height, Rz(DIN) 0.214 0.413 0.818 0.106 0.391 1.015
Maximum peak-to-valley height, Rmax 0.371 0.749 0.999 0.146 0.555 1.553
Mean third highest peak-to-valley height, R3 z 0.206 0.403 0.781 0.097 0.376 0.998
Maximum third highest peak-to-valley height, R3zmax 0.370 0.707 0.994 0.136 0.541 1.537
Ten points height, Rz(ISO ) 0.356 0.853 1.192 0.146 0.846 1.697
Total roughness height, R y 0.371 0.887 1.235 0.156 0.864 1.710
Root-mean-square (RMS) profile height, Rq 0.049 0.119 0.254 0.024 0.160 0.309
Mean peak height, Rpm 0.084 0.162 0.411 0.054 0.209 0.623
Maximum peak height, Rp 0.133 0.282 0.529 0.079 0.413 1.130
Mean valley depth, Rvm 0.131 0.252 0.406 0.052 0.182 0.392
Maximum valley depth, Rv 0.239 0.605 0.706 0.078 0.451 0.580

Magazine of Concrete Research, 2008, 60, No. 5 333

Santos and Júlio
0·30

0·25

Profile height: mm
0·20

0·15

0·10

0·05

0·00
Ra Rz(DIN) Rmax R3z R3zmax Rz(ISO) Ry Rq Rpm Rp Rvm Rv
⫺6 ⫺6
Left as-cast (10 ⫻ 10 m) Left as-cast (60 ⫻ 10 m)

Fig. 8. Roughness parameters obtained with both laser sensors for the left as-cast surface

70

60
Coefficient of variation: %

50

40

30

20

10

0
Ra Rz(DIN) Rmax R3z R3zmax Rz(ISO) Ry Rq Rpm Rp Rvm Rv
Left as-cast (10 ⫻ 10⫺6 m) Left as-cast (60 ⫻ 10⫺6 m) Wire-brushing (60 ⫻ 10⫺6 m) Sand-blasting (60 ⫻ 10⫺6 m)

Fig. 9. Coefficient of variation of the adopted roughness parameters

46.22 MPa respectively. All slant shear specimens were Max ⫽ 16·11
Bond strength in shear: MPa

18 Avg ⫽ 14·13
tested, presenting an adhesive failure. For the pull-off Min ⫽ 12·93
16
specimens, debonding occurred when drilling the core Max ⫽ 11·57
14 Avg ⫽ 10·67
in all five specimens with the substrate surface left as- 12 Min ⫽ 9·07
cast. All the remaining pull-off specimens were tested, 10
8
also presenting an adhesive failure. Test results are 6 Max ⫽ 1·31
shown in Figs 10 and 11. 4 Avg ⫽ 1·30
Min ⫽ 0·71
2
Correlation between roughness and bond strength 0
Left as-cast Wire-brushing Sand-blasting
Table 1 presents the results obtained with the pro-
posed new methodology, using the developed laser Fig. 10. Bond strength in shear
roughness analyser, and those obtained with the
methodology adopted in the previous study,11 based on
digital image processing. Comparing both methods, it Max ⫽ 2·83
Bond strength in tension: MPa

3·50 Avg ⫽ 2·65

can be observed that, for the rough without exposing Max ⫽ 2·37 Min ⫽ 2·38
3·00 Avg ⫽ 1·92
the aggregates surface type, Fig. 12, the roughness 2·50 Min ⫽ 1·70
parameters present, in general, similar values when 2·00
calculated using each of the methods. For the rough 1·50
with aggregates exposure surface type, Fig. 13, most 1·00
of the roughness parameters considered present higher 0·50
values when assessed with the laser roughness analyser 0·00
Left as-cast Wire-brushing Sand-blasting
than when determined by digital image processing.
For the smooth surface type, Fig. 14, most of the Fig. 11. Bond strength in tension
334 Magazine of Concrete Research, 2008, 60, No. 5
 Development of a laser roughness analyser to predict in situ the bond strength of concrete-to-concrete
1·0
0·9
0·8
0·7
Profile height: mm

0·6
0·5
0·4
0·3
0·2
0·1
0·0
Ra Rz(DIN) Rmax R3z R3zmax Rz(ISO) Ry Rq Rpm Rp Rvm Rv

Digital image treatment Laser roughness analyser

Fig. 12. Roughness parameters assessed with the laser roughness analyser and the digital image treatment for the wire-brushed
surface

1·8

1·6

1·4
Profile height: mm

1·2

1·0

0·8

0·6

0·4

0·2

0·0
Ra Rz(DIN) Rmax R3z R3zmax Rz(ISO) Ry Rq Rpm Rp Rvm Rv

Digital image treatment Laser roughness analyser

Fig. 13. Roughness parameters assessed with the laser roughness analyser and the digital image treatment for the sand-blasted
surface

0·40

0·35

0·30
Profile height: mm

0·25

0·20

0·15

0·10

0·05

0·00
Ra Rz(DIN) Rmax R3z R3zmax Rz(ISO) Ry Rq Rpm Rp Rvm Rv

Digital image treatment Laser roughness analyser

Fig. 14. Roughness parameters assessed with the laser roughness analyser and the digital image treatment for the left as-cast
surface

roughness parameters considered present smaller va- methodology, more precise roughness profiles are
lues when assessed with the laser roughness analyser obtained.
than when determined by digital image processing. Table 2 presents the values of the coefficient of
This analysis confirms that with the proposed new correlation assuming linear correlations between the
Magazine of Concrete Research, 2008, 60, No. 5 335
Santos and Júlio

Table 2. Correlation between roughness parameters and bond strength

Roughness parameter Coefficient of correlation (R-squared)

Digital image treatment Laser roughness analyser

Shear Tension Shear Tension

Average roughness, Ra 0.9178 0.9203 0.9748 0.9762

Mean peak-to-valley height, Rz(DIN) 0.8985 0.9013 0.8908 0.8936
Maximum peak-to-valley height, Rmax 0.9898 0.9907 0.8788 0.8817
Mean third highest peak-to-valley height, R3z 0.9049 0.9075 0.8891 0.8919
Maximum third highest peak-to-valley height, R3zmax 0.9773 0.9786 0.8780 0.8810
Ten points height, RzISO 0.9886 0.9895 0.9504 0.9523
Total roughness height, R y 0.9890 0.9900 0.9520 0.9539
Root-mean-square (RMS) profile height, Rq 0.9043 0.9070 0.9596 0.9613
Mean peak height, Rpm 0.8509 0.8541 0.8691 0.8721
Maximum peak height, Rp 0.9203 0.9228 0.8927 0.8955
Mean valley depth, Rvm 0.9463 0.9483 0.9240 0.9264
Maximum valley depth, Rv 0.9984 0.9980 0.9999 0.9998

bond strength of the interface, both in shear and in
tension, and each of the roughness parameters consid-
ered, determined with the proposed new methodology
and with the methodology adopted in the previous
study.11 It can be observed that values are always high-
er than 0.85 and, for some roughness parameters, high-
er than 0.95. The best result was obtained with the
maximum valley depth, Rv , presenting an almost per-
fect linear correlation with the bond strength in shear
and in tension, using either the laser roughness analyser
or digital image processing, with values always over
0.9990. In relation to the total roughness height, R y ,
high values of coefficient of correlation are registered
using digital image processing, 0.9890 in shear and
0.9900 in tension, and using the laser roughness analy-
ser, 0.9520 in shear and 0.9539 in tension. The maxi-
mum peak-to-valley height, Rmax , exhibits high values
of coefficient of correlation using digital image proces-
sing, 0.9898 in shear and 0.9907 in tension, but signifi-
cantly lower values, using the laser roughness analyser,
0.8788 in shear and 0.8817 in tension. With the average
roughness, Ra , the opposite is observed, with high va-
lues of coefficient of correlation using the laser rough-
ness analyser, 0.9748 in shear and 0.9762 in tension,
and lower values, using digital image processing,
0.9178 in shear and 0.9203 in tension.
Conclusions
This study corroborates the conclusions drawn in
previous studies.10,11
(a) The roughness of the substrate surface influences
significantly the bond strength of concrete-to-con-
crete interfaces.
(b) It is possible to obtain a roughness profile of the
336
substrate surface and, from this, to determine
roughness parameters.
(c) It is possible to define linear correlations between
some of these and the bond strength of the inter-
face, both in shear and in tension, with very high
coefficients of correlation (R2 .0.95).
Moreover, it was demonstrated that the proposed new
method, using the developed laser roughness analyser,
presents all the advantages, with even more accurate
results, and overcomes all the disadvantages of the
digital image processing method, adopted in the pre-
vious study.11 In fact, the proposed new method really
is a non-destructive method, easy to carry out and
allowing the fast and in situ prediction of the bond
strength of concrete-to-concrete interfaces.
The use of the 10 ìm laser sensor for the left as-cast
surface seems to be more adequate, since all roughness
parameters are slightly higher when compared with the
same parameters measured with the 60 ìm laser sensor.
This indicates that for very smooth surfaces, the resolu-
tion of the laser sensor has to be carefully selected.
The coefficient of variation of the roughness para-
meters presents significantly smaller values for the
wire-brushed surface, almost always lower than 20%,
than for each of the remaining two surface types: as-
cast and sand-blasted. This can be justified with the
resulting uniformly rough surface, obtained with this
treatment. For the left as-cast surface, this coefficient
varies between 20% and 40%. This higher value can be
justified with the existence of air holes in the concrete
surface, which have a very significant influence on the
assessed roughness parameters when these are included
in the path of the evaluation length. For the sand-
blasted surface, this coefficient presents the highest
values. This can be justified by taking into account that
this treatment implies the exposure of the aggregates,
which produces an irregularly rough surface.
Magazine of Concrete Research, 2008, 60, No. 5
 Development of a laser roughness analyser to predict in situ the bond strength of concrete-to-concrete
Further work is needed to create a large database,
statistically representative, before proposing a design
expression that replaces the current qualitative ap-
proach of the design codes. Nevertheless, results indi-
cate that this fast, non-destructive, in situ methodology
can quantify the roughness of the substrate surface and
the maximum valley depth roughness parameter to be
adequate to incorporate a design expression of the
longitudinal shear strength of the interface between
parts of elements cast at different times.
Acknowledgements
The authors acknowledge the financial support of the
Portuguese Science and Technology Foundation (FCT),
PhD Grant number SFRH/BD/25510/2005. Acknowl-
edgements are extended to the following companies for
their financial and material support: MAPREL—Em-
presa de Pavimentos e Materiais Pré-Esforçados Lda;
SIKA Portugal SA, AFAssociados—Projectos de En-
genharia SA; WEBER Cimenfix, CIMPOR—Cimentos
de Portugal; BETÃO-LIZ Adémia and Euro-Planning—
Engenharia & Gestão Lda.
References
1. Comité Européen de Normalisation. Eurocode 2: Design of
Concrete Structures. Part 1.1: General Rules and Rules for
Buildings. CEN, Brussels, April, 2003, prEN 1992-1-1.
2. American Concrete Institute. ACI Manual of Concrete
Practice. Part 3: 1994, Use of Concrete in Buildings—Design,
Magazine of Concrete Research, 2008, 60, No. 5
Specifications, and Related Topics. ACI, Detroit, 2002, ACI
318.
3. Canadian Standards Association. Design of Concrete Struc-
tures—Structures Design. CSA, Rexdale, Ontario, 1994, CSA
A23.3-94.
4. American Society for Testing and Materials. ASTM E
965-96: Standard Test Method for Measuring Pavement Macro-
texture Depth Using a Volumetric Technique, Book of Standards,
Vol. 04.03. ASTM, West Conshohocken, PA, USA.
5. Abu-Tair A. I., Lavery D., Nadjai A., Rigden S. R. and
Ahmed T. M. A. A new method for evaluating the surface
roughness of concrete cut for repair or strengthening. Construc-
tion and Building Materials, 2000, 14, No. 13, 171–176.
6. Issa M. A., Islam M. S. and Chudovsky A. Fractal dimension:
a measure of fracture roughness and toughness of concrete.
Fracture Mechanics, 2003, 70, No. 1, 125–137.
7. Garbacz A., Górka M. and Courard L. Effect of concrete
surface treatment on adhesion in repair systems. Magazine of
Concrete Research, 2005, 57, No. 1, 49–60.
8. Maerz N., Nanni A., Myers J. and Galecki G. Laser profilo-
metry for concrete substrate characterization prior to FRP lami-
nate application. Concrete Repair Bulletin, 2001, May/June,
4–8.
9. Mummery L. Surface Texture Analysis: The Handbook. Hom-
melwerke GmbH, Thyssen, Germany, 2000.
10. Júlio E. N. B. S., Branco F. A. B. and Silva V. D. Concrete-
to-concrete bond strength. Influence of the roughness of the
substrate surface. Construction and Building Materials, 2004,
18, No. 9, 675–681.
11. Santos P., Júlio E. and Silva V. D. Correlation between con-
crete-to-concrete bond strength and the roughness of the sub-
strate surface. Construction and Building Materials, 2007, 21,
No. 8, 1688–1695.
Discussion contributions on this paper should reach the editor by
1 December 2008
337
Copyright of Magazine of Concrete Research is the property of Thomas Telford Ltd and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.