Alexandria Engineering Journal (2014) 53, 169184

Alexandria University

Alexandria Engineering Journal

www.elsevier.com/locate/aej
www.sciencedirect.com

ORIGINAL ARTICLE

Reliability of core test Critical assessment

and proposed new approach
Shak Khoury, Ali Abdel-Hakam Aliabdo *, Ahmed Ghazy
Structural Engineering Department, Faculty of Engineering, Alexandria University, Egypt
Received 14 November 2013; revised 21 December 2013; accepted 21 December 2013
Available online 22 January 2014

KEYWORDS
Strength;
Drilled cores;
Interpretation;
Codes;
Regression analysis;
Model

Abstract Core test is commonly required in the area of concrete industry to evaluate the concrete
strength and sometimes it becomes the unique tool for safety assessment of existing concrete
structures. Core test is therefore introduced in most codes. An extensive literature survey on different
international codes provisions; including the Egyptian, British, European and ACI Codes, for
core analysis is presented. All studied codes provisions seem to be unreliable for predicting the in-situ
concrete cube strength from the results of core tests. A comprehensive experimental study was undertaken to examine the factors affecting the interpretation of core test results. The program involves four
concrete mixes, three concrete grades (18, 30 and 48 MPa), ve core diameters (1.5, 2, 3, 4 and 6 in.),
ve core aspect ratios (between 1 and 2), two types of coarse aggregates (pink lime stone and gravel),
two coring directions, three moisture conditions and 18 different steel arrangements. Prototypes for
concrete slabs and columns were constructed. More than 500 cores were prepared and tested in addition to tremendous number of concrete cubes and cylinders. Results indicate that the core strength
reduces with the increase in aspect ratio, the reduction in core diameter, the presence of reinforcing
steel, the incorporation of gravel in concrete, the increase in core moisture content, the drilling perpendicular to casting direction, and the reduction in concrete strength. The Egyptian code provision for
core interpretation is critically examined. Based on the experimental evidences throughout this study,
statistical analysis has been performed to determine reliable strength correction factors that account
for the studied variables. A simple weighted regression analysis of a model without an intercept was
carried out using the SAS Software package as well as Data Fit software. A new model for interpretation of core test results is proposed considering all factors affecting core strength. The model
when calibrated against large number of test data shows good agreement. The proposed model can
effectively estimate the in-situ concrete cube strength from core test results.
2014 Production and hosting by Elsevier B.V. on behalf of Faculty of Engineering, Alexandria
University.

* Corresponding author. Tel.: +20 1001679661.

E-mail address: [email protected] (A.A.-H. Aliabdo).
Peer review under responsibility of Faculty of Engineering, Alexandria University.

Production and hosting by Elsevier

1110-0168 2014 Production and hosting by Elsevier B.V. on behalf of Faculty of Engineering, Alexandria University.
http://dx.doi.org/10.1016/j.aej.2013.12.005

170
1. Introduction
The compressive strength of concrete is a direct requisite of all
concrete structures that need to resist applied forces of whatever nature. Actually, the concrete compressive strength is a
good index of most other properties of practical signicance.
To ensure concrete quality, standard test specimens are examined during construction. These specimens, which give the
potential strength of concrete, are prepared, cured and tested
according to relevant standard specications and codes. On
the other hand, determination of the actual strength of concrete in a structure is not easy because it is dependent on the
history of curing and the adequacy of compaction of concrete.
Therefore, one question that designers frequently ask is
whether or not the standard test specimens can represent
in situ-strength of concrete. The answer to this question becomes even more important when the strengths of standard
test specimens are found to be lower than the specied value.
In this case, either the strength of concrete in the actual structure is low or the specimens are not actually representing the
concrete in the structure. The problem is generally solved by
drilling and testing core specimens from the suspected structural member. Furthermore, it may not be possible to nd
and test standard specimens at a later age and it may be necessary to assess the current strength of a structure to determine
whether the strength and durability are adequate for its future
use when the concrete is doubted or the structure is intended to
be used for higher stress conditions. For these special situations, the core test is the most useful and reliable way to assess
the properties of the concrete in the structure [1]. For these reasons, the common way of determining in-situ strength of concrete is to drill and test cores [110]. Although the method
consists of expensive and time consuming operations, cores
give reliable and useful results since they are mechanically
tested to destruction [2]. However, the test results should be
carefully interpreted because core strengths are affected by a
number of factors such as diameter, l/d ratio and moisture
condition of the core specimen, the direction of drilling, the
presence of reinforcement steel bars in the specimen and even
the strength level of the concrete [1124].
2. Impact of core test
The determination of cube strength is the most common and
simple approach for evaluating the concrete strength during
the construction of new buildings; however, the absence of
cube results or the doubt on the results may raise a critical situations. Furthermore, during the rehabilitation of existing
structures, another approach for evaluating the concrete
strength is of great importance. From this point, testing of
concrete elements in existing structures comes into picture.
From that point, core test is commonly required in the area
of concrete industry and hence is included in most international codes provisions. In fact, core test becomes a must in
many critical circumstances and sometimes becomes the
unique tool for concrete quality assessment. From general
prospective, core test is ultimately needed to assess one or a
combination of the following:
1. The quality of the concrete provided to a construction
(potential strength).

S. Khoury et al.
2. The quality of the concrete in the construction (in-situ
strength), known as actual strength.
3. The ultimate capacity of the structure to carry the imposed
loads; actual loads, design loads, and new additional loads.
4. The deterioration in a structure due to overloading, fatigue
(bridge structures, machine base, etc.), chemical reaction
(ASR or chemical spillage, etc.), re or explosion, and
weathering.

3. Codes provisions for core analysis; state-of-the-art

To determine in-situ cube strength, core test is available but
most codes give different results depended on the factors
used. It is generally agreed that the compressive strength of
extracted core can be obtained by dividing the ultimate load
by the cross-sectional area of the core, calculated from the
average diameter; however, the critical problem is actually
to translate this result to cube/cylinder strength as mentioned
earlier. In fact, the core test results must be carefully interpreted since core strength is affected by a number of factors
such as diameter, aspect ratio (l/d), moisture condition of
the core specimen, direction of drilling, the presence of reinforcement steel bars in the core, type and size aggregate
and even the strength level of the concrete. Factors considered in codes provisions for core interpretation are different.
Table 1 summarizes the factors considered in Egyptian code,
American code, European code, British code and Concrete
society to interpret core strength to in site concrete strength.
The table clearly implies that for a specic core strength, the
obtained in-situ strength is different based on the considered
code. The approaches adopted in some of the listed codes/for
core interpretation are summarized below. More details
about other codes approaches for core analysis may be seen
elsewhere [14].
3.1. Current Egyptian Code (2008)/standard and British Code
(2003)
According to these codes, the estimated in-situ concrete cube
strength (fcu) may be calculated from the measured compressive strength of core (fcore) according to the following
expression:
fcu Fl=d  FReinf: fcore

where the factors (Fl/d) and (FReinf.) account for the effect of l/d
and the presence of reinforcing steel, respectively. These
factors are given in the codes as follows:
F1=d

D
1:5 1=k

FReinf: 1 1:5

Ur  d
Uc 

where D is the equal to 2.5 for core drilled horizontally

(perpendicular to casting direction); or 2.3 for cores drilled
vertically (parallel to casting direction), k is length to diameter
ratio (l/d) after ends preparation, Ur is diameter of the reinforcing bar, Uc is core diameter, d is distance of axis of bar
from the nearest core end, and is the core length after end
preparation.

Reliability of core test Critical assessment

Table 1
List

1
2
3

171

Factors involved in interpretation of core results by different codes.

Code/standard

Edition

Egyptian Code/Standard Specication

British Code/Standard Specication
American Concrete Institute ACI

European Standard Specication

5
6

Japanese Standard
Concrete Society

2008
2003
1998
2012
1998
2009
1998
1987

Factors Considered
Aspect ratio
p
p
p
p
p
p
p
p

In addition, for core specimen containing two bars no further apart than the diameter of the larger bar, only the bar corresponding to the higher value of (Ur  d) is considered. If the
bars are further apart, their combined effect P
should be assessed
by replacing the term (Ur  d) by the term ( Ur  d).
It should be pointed out that above equations used to interpret the core concrete strength to the in-situ concrete cube
strength have been developed based on a set of assumptions
and through many converting process. It is also of interest to
note that the damage effect is considered in the development
of the formulas in indirect way. The subject derivation and detailed formulas may be seen elsewhere [14].
3.2. American Concrete Institute (ACI)
3.2.1. Former ACI Code (2002) & Current ASTM (2009)
The methodology of core interpretation given in the former
ACI code was remained without changes for decades and up
to Year (2003). The in-place strength of concrete cylinder at
the location from which a core test specimen was extracted
can be computed using the equation:
fcy Fl=d  fcore

where fcy is the equivalent in-place concrete cylinder strength,

fcore is concrete core strength, and Fl/d is the strength correction
factor for aspect ratio.
The former ACI code does not include any equation to calculate the correction factor (Fl/d); however, the code gives different values for this term that is associated with different
aspect ratios (l/d) as given in Table 2. It should also be noted
that the approach of current ASTM is similar to that mentioned above. The only considered variable is the aspect ratio
(l/d). It should be noted that identical approach to that mentioned above is still effective in ASTM C42/C42M-03 [10].

Diameter

p
p

Table 3
03.
List
b

(1)

(4)

0.87

0.93

0.96

0.98

Factors

Mean values

Fl/d:l/d ratio
As-received


2
1  {0.130  afcore} 2  dl

2
1  {0.117  afcore} 2  dl

2
1  {0.144  afcore} 2  dl

Air drieda

a
b

Fl/d

Strength correction factors according to ACI 214.4R-

Soaked 48 h

Table 2 Mean values for factor Fl/d according to ACI Code

(1998) and ASTM.
1.75

p
p

Direction
p
p

where fc is the equivalent in-place concrete cylinder strength,

fcore is concrete core strength, Fl/d is strength correction factor
for aspect ratio, Fdia is strength correction factors for diameter,
Fmc is strength correction factor for moisture condition of core
sample, and FD is the strength correction factor that accounts
for effect of damage sustained during core drilling including
micro-cracking and undulations at the drilled surface and cutting through coarse-aggregate particles that may subsequently
pop out during testing.
The ACI committee considered the correction factors presented in Table 3 for converting core strengths into equivalent
in-place strengths based on the work reported by Bartlett and
MacGregor [6]. It should be noted that the magnitude of

(3)

1.50

p
p
p

fc Fl=d  Fdia  Fmc  FD  fcore

Starting from Year 2003, signicant changes have been made

to the relevant ACI Code provisions regarding the interpreta-

1.25

Damage

tion of core strength test results. New factors have been considered. These include core diameter, moisture content of
core sample, core damage associated with drilling, in addition
to the effect of aspect ratio that was previously considered in
the former ACI edition (1998). According to the ACI
214.4R-03, the in-place concrete strength can be computed
using the equation:

(2)

1.00

Moisture

3.2.2. Current ACI Code (2012) [15]

Specimen length-to-diameter ratio, l/d

Reinforcing
p
p

Fdia: core diameter

50 mm
100 mm
150 mm

1.06
1.00
0.98

Fmc: core moisture content

As-received
Soaked 48 h
Air drieda

1.00
1.09
0.96

FD: damage due to drilling

1.06

Standard treatment specied in ASTM C 42/C 42M.

Constant a equals 4.3(104) 1/MPa for fcore in MPa.

172

S. Khoury et al.

damage factor suggested in the table is based on data for normal weight concrete with strengths between 14 and 92 MPa.

fpot Fl=d  FReinf: fcore

3.3. European Code (CEN)

3.3.1. Former CEN Code (1998)
The main factors considered in the former European code are
the size and geometry of the core, coring direction, the presence of reinforcing bars or other inclusions and the effect of
drilling damage. This approach was originally taken from
the German Standards by that time. The proposed relationship
to convert the strength of a core specimen fcore into the equivalent in-situ concrete value fc as recommended by the code and
reported elsewhere [16] is as follows:
fc CH=D  Cdia  Ca  Cd  fcore

where fc is the equivalent in-place concrete strength, fcore is

concrete core strength, CH/D is correction factor for aspect ratio and is calculated according to the following formula where
H and D are, respectively, the core height and diameter:
CH=D

2
1:5 D=H

from the measured compressive strength of core (fcore) according to the following expression:

Cdia. is the correction factors for core diameter, that is equal to

1.06, 1.00 and 0.98 for core diameter, respectively, equal to 50,
100 and 150 mm. Ca is correction factors for the presence of
reinforcing steel, equal to 1 for no bars, and varying between
1.03 for small diameter bars (U = 10 mm) and 1.13 for large
diameter bars (U = 20 mm) and Cd is the correction factors
for damage due to drilling,
The factor Cd is proposed to be 1.10 providing that the core
extraction is carefully carried out by experienced operators.
However, taking into account that the lower the concrete
quality the larger the drilling damage, it appears more suitable
to put Cd = 1.20 for fcore < 20 MPa, and Cd = 1.10 for
fcore > 20 MPa, as suggested by Dolce et al. [16].
3.3.2. Current CEN Code (2009)
The approach proposed by current European Code for interpreting the core strength to in-situ concrete cube strength
(fcu) is similar to that suggested previously by the former British Code that presented earlier. The damage and diameter factors considered in the former CEN Code were eliminated in the
more recent edition. Factors considered are only aspect ratio
and the presence of reinforcing steel as given in Eqs. (2) and
(3). The main difference in core interpreting between recent
CEN Code and British Code seems to be in the magnitude
of the Term D in Eq. (3). The CEN Code considers
D = 2.5 for cube strength and 2.0 for cylinder strength instead
of the values of 2.3 and 2.5 that is mentioned in the British
Code. In addition, the aspect ratio should be limited between
1.0 and 1.2
3.4. Concrete Society (CS)
The approach proposed by Concrete Society CS [23] is similar
to that suggested by the former British Standard presented
earlier. However, the Concrete Society estimates the so-called
Potential concrete strength that is different than the in-situ
concrete cube strength. According to Concrete Society, the
estimated potential concrete strength (fpot) can be calculated

where the factors (Fl/d) and (FReinf.) account for the effects of
l/d and the presence of reinforcing steel, respectively, and they
can be determined using Eqs. (2) and (3) presented earlier.
Again, the difference seems to be in the magnitude of the
term D that is considered 2.3 or 2.5; depending on the drilling
direction; in the current Egyptian Code, but it is proposed to
be 2.0 in the current European Code. However; according to
Concrete Society, the term D is equal to 3.25 for cores drilled
horizontally (perpendicular to casting direction); or 3.0 for
cores drilled vertically (parallel to casting direction). The derivation of these numbers may be seen elsewhere [14].
From general prospective, it has been shown that the potential strength is approximately 1.5 times the strength of a core
providing that the core length/diameter ratio = 2, the drilling
direction is vertical, the core is free of reinforcement, and concrete is well compacted and does not include the weaker material near the top of a lift. Actually, the formulae for converting
core strength to cube strength indicate that, if the concrete in
the structure is fully compacted and normally cured, the actual
concrete strength is about 77% of the potential strength.
4. Research signicance
Core tests are generally performed to assess whether suspect
concrete in a new structure complies with strength-based
acceptance criteria or not. In addition, it is critically used
to determine in-place concrete strengths in an existing structure for the evaluation of structural capacity. It is generally
agreed among engineers, contractors, consultants and
researchers that the results of core tests are very reliable to assess the strength of concrete elements. Unfortunately, this
statement may not be totally true and may lead to a misleading in the assessment of structural safety. Actually, the available test information on cores is full of contradictions and
confusions. The conict between different codes regarding
the interpretation and conversion of the core results to the
in site concrete strength raises a critical debate. It seems that
the core test itself may be reliable but the core analysis is
questionable. This comprehensive and costly program; which
includes drilling about 500 cores, is undertaken to reconsider
and focus on this critical aspect.
5. Experimental program
A special testing program on cores was constructed to investigate the inuence of different variables that may affect the
interpreting of core test results into in-situ cube strength. Variables considered are as follows; 1 core length-to-diameter ratio (L/d), 2 core diameter (d), 3 direction of drilling, 4 the
presence of reinforcing steel, 5 moisture content of core specimen, and 6 damage due to drilling. Four different concrete
mixtures were considered throughout the program using two
types of aggregate; crushed aggregate (pink lime stone) and
natural aggregate (gravel). Ordinary Portland cement CEM
I-42.5N (ASTM Type I) was used. Three levels of concrete
grades were considered; 18, 30, and 48 MPa. Concrete mix
proportions are given in Table 4.

Reliability of core test Critical assessment

Table 4

Concrete mixes used throughout the program.

Mix proportion (kg/m3)

Mix.

Mix
Mix
Mix
Mix

173

1
2
3
4

Some properties

Coarse agg.

Fine agg.

Cement

Total water

Admixture (L)

w/c

Aggregate type

1080
1070
1080
1200

760
705
760
690

300
400
450
400

170
175
160
142

5
4
6
5

0.57
0.44
0.36
0.36

Pink lime stone

The program comprises prototypes of slabs and columns.

Actually,
the
examined
structural
elements
were
1.20 2.20 m-concrete slabs with different thicknesses; 10,
12, 15, 17.5, 20, and 30 cm. Slabs were made of the three concrete grades mentioned above with and without reinforcing
steel. Also, concrete columns having cross-section of
20 80 cm and 1.2 m height were constructed using highstrength concrete (480 kg/cm2). Special wooden forms were designed and fabricated. Slabs and columns were cast and cured
in the laboratory condition simultaneously with corresponding
15 cm-standard cubes and 15 30 cm standards cylinders.
After hardening, cores with different diameters (1.5, 2, 3, 4
and 6 in.) were extracted from slabs and columns using a diamond-tipped core drilling machine then trimmed to give overall aspect ratios (l/d) between 1.0 and 2.0. The aspect ratios of
capped core specimens were 1.0, 1.25, 1.5, 1.75 and 2.0. In
addition, non-standard cylinders comparable to extracted
cores with respect to diameter, aspect ratio, concrete grade
and reinforcement were introduced to evaluate the damage effect due to drilling. The compressive strengths of cores and

Gravel

associated standard specimens were determined at comparable

ages using a compression testing machine with a fully automatic press. The rate of load application for all specimens
was 0.25 MPa/s as recommended by most codes. Each strength
value is the average of at least three core specimens. Layout of
the current research plan is presented in Fig. 1 while some
stages of the experimental work are shown in the photographs
given in Figs. 28.
A total number of about 500 cores in addition to 300 standard molded cubes and cylinders were examined in the current
research. In fact, this work was very exhausting, costly and
time consuming.
6. Results and discussions
Results of about 500 cores were examined to evaluate the effects of studied variables on the relations between core
strength and strength of corresponding standard molded cubes
and cylinders. Some results are to be presented; however, comprehensive obtained data may be seen elsewhere [14]. The compressive strength of standard specimens is listed in Table 5.
6.1. Effect of studied variables on core strength
6.1.1. Effect of length/diameter ratio (l/d)

Figure 1

General layout of the research program.

Figure 2

The length to diameter ratio (l/d) of core specimens has long

been recognized as a prime factor that inuences the failure
load. This effect depends on various conditions such as
strength of concrete, elastic modulus and certainly on (l/d)
ratio. Strength correction factors for (l/d) ratio (Fl/d) are determined herein by converting the strength of a core with an (l/d)
between 1 and 2 to the strength of an equivalent standard core
with (l/d) equal 2. Standard cylinder having a length/diameter
ratio of 2.0 is considered herein as a benchmarking. As the ratio increases, the measured strength decreases due to the effect

Fabrication of molds for columns and slabs.

174

S. Khoury et al.

Figure 3

Figure 4

Reinforced slabs with different congurations of reinforcement and depth.

Some molds for standard cubes and cylinders.

of specimen shape on stress distributions. The current study

has revealed that the correction factor for (l/d) ratio depends
on level of strength, diameter of specimen and moisture.
Fig. 9 illustrates the relation between the aspect ratio (l/d)
and correction factor (Fl/d) according to level of strength and
core diameter. The gure clearly demonstrates that the factor
(Fl/d) is signicantly reduced for high strength concrete. As
the concrete strength increases, a correction factor becomes
closer to 1.0. Fig. 10 presents comparison between the
obtained data with two different recognized code and society

Figure 5

(ACI [15] and Concrete society CS [23]). The current correction factor (Fl/d) given by ACI 214.4R-03 ts the data for high
strength level (48 MPa); however, in low strength concrete
(18 MPa), the (l/d) ratio becomes so effective that the ACI approach appears irrational. On the other hand, the ACI factor
(Fl/d) has been found out to agree with Chungs equation
(1989) for low strength concrete but does not coincide with
that equation in high strength concrete. Disregarding the effect
of concrete strength level on (Fl/d) seems to be responsible to
large extent to this conict.
On the basis of data obtained herein through testing large
number of cores, statistical analysis has been performed to
determine a reliable factor (Fl/d). A simple weighted regression
analysis of a model without an intercept was carried out using
the SAS Software package (SAS Institute, 2008) as well as
Data Fit software. Figs. 11 and 12 show the multi-dimension relationship among the subject factor (Fl/d), aspect ratio,
and concrete core strength derived from the advanced statistical analysis programs.
The advanced performed analysis with the results of large
numbers of cores (about 150 data) gives that the correction
factor that accounts for aspect ratio (Fl/d) can be given by
the formula:

Non-standard molded reinforced cylinders comparable to extracted cores.

Figure 6

Some reinforced concrete slabs used throughout the program.

Reliability of core test Critical assessment

Figure 7

175

Core drilling at pre-specied locations and some extracted cores.

where fcore is the core strength in kg/cm2; and l/d is length to

diameter ratio for the same unit.
6.1.2. Effect of core diameter (D)

Figure 8

Fl=d
1

Table 5
Mixture

h

General view for columns and slabs.

1
 
i
l 2

2

1:15
d
f

75

core

The core diameter plays an important role in affecting core

strength results. It is generally agreed for molded concrete that
the concrete strength is decreased as the specimen size increases. However, it is of great interest to explore that the inverse trend is evident in case of drilled cores. As the diameter
decreases, the ratio of cut surface area to volume increases,
and hence the possibility of strength reduction due to cutting
damage increases. For core diameters above 100 mm this effect
may not be signicant be also it cannot be neglected as discussed later. Fig. 13 shows the mean ratio between strength
of cores having different diameters and drilled at comparable
locations in a specic structural element. From overall
prospective, the gure shows that smaller cores have lesser
strengths. In smaller cores the effect of core diameter on

Compressive strength of standard specimen.

Compressive strength (MPa)
Cylinder specimen (150 300 mm)

Mix
Mix
Mix
Mix

1
2
3
4

Cube specimen (150 mm)

7 days

28 days

7 days

28 days

10.5
19.3
18.8
18.8

13.4
24.1
39.6
23.8

13.1
23.5
39.1
23.2

19.3
29.8
49.1
29.6

Concrete strength = 18 MPa, d = 100 mm

Concrete strength = 48 MPa, d = 100 mm
Concrete strength = 48 MPa, d = 150 mm

Concrete strength = 30 MPa, d = 100 mm

Concrete strength = 30 MPa, d = 150 mm

Aspect ratio l/d

Figure 9

Relation between core aspect ratio and correction factor Fl/d.

176

S. Khoury et al.
ACI

Current research

CS

CS
Concrete strength = 48 MPa
Core diameter = 100 mm

Aspect ratio l/d

ACI

Current research

CS

ACI
Current research

CS

Concrete strength = 18 MPa

Core diameter = 100 mm

Aspect ratio l/d

Figure 10

Figure 11

Relation between correction factor Fl/d and aspect ratio as compared with ACI Code and CS.

Advanced statistical analysis for prediction of (Fl/d) using SAS Program (SAS Institute, 2008).

reduction in core strength becomes signicant. Fig. 14 indicates that the factor (Fd) required to compensate strength
reduction varies from 1.05 to 1.08 in averages for core diameter, 75 and 50 mm, and it goes as high as 1.131.17 in smaller
core with diameter 38 mm and also depends on degree of compaction and potential defects in the concrete.
The coefcient of variation in the compressive strength decreases for larger core diameter as shown in Fig. 15. It is depend mostly on the variance of the thickness of the damaged
region. It is imperative that a large number of cores to be

drilled in case of core diameter less than 100 mm to maintain

comparable within-test variation as for large diameter cores.
It should be pointed out that Both Egyptian and British Codes
allow the use of cores with diameters either 100 or 150 mm
without suggesting any correction factor for this study; however, the ACI Code allows the use of smaller cores down to
50 mm diameter and under some precautions 38 mm-diameter
core may be used according to ACI Code [14].
The statistical analysis has been performed to determine a
reliable factor (Fd) accounting for core diameter. Again, the

Reliability of core test Critical assessment

Figure 12

177

Advanced statistical analysis for prediction of (Fl/d) using Data Fit software.

30 MPa
48 MPa
30 MPa
48 MPa

48 MPa
30 MPa

Aspect ratio l/d

Figure 13

Mean Ratio between 100 mm- and 150 mm-core strengths for different aspect ratios (l/d) and concrete strength.

d = 38 mm

d = 38 mm

d = 38 mm

Factor
Fd
d = 50 mm
d = 50 mm
d = 75 mm

d = 75 mm

d = 50 mm
d = 75 mm

Aspect ratio l/d

Figure 14

Effect of core diameter (d) on core strength for different aspect ratios (l/d) (d = 38, 50, 75 mm).

SAS Software package (SAS Institute, 2008) as well as

Data Fit software yields the following expression for the
correction due to diameter effect (Fd):
Fd 1:507  0:11 lnd

10

where (d) is the core diameter in mm.

6.1.3. Effect of damage due to drilling
Non-standard cylinders comparable to extracted cores with
respect to diameter, aspect ratio, concrete grade and reinforcement were prepared to evaluate the damage effect due to

drilling. Samples were divided into groups and each consists

of two identical samples but one was obtained by drilling
and the other by concrete pouring. Results of laboratory tests
reported herein and elsewhere [14] conrm that the damage occurred in core samples due to cutting process is directly related
to concrete strength and type of aggregate. Results have also
revealed that the concrete strength level has a pronounced effect on the damage factor (Fdmg) and is considered to be a main
factor. In addition, aspect ratio (l/d), core diameter, and the
cutting surface area of core specimen should also be given
attention. Analysis of test result has shown that the effect of

178

S. Khoury et al.

38 mm
38 mm
50 mm
75 mm
50 mm
75 mm

50 mm
38 mm
100 mm75 mm

100 mm

100 mm

Aspect ratio l/d

Figure 15

Effect of core diameter on the precision of core test results for different aspect ratios (l/d) (d = 38, 50, 75, 100 mm).

Figure 16

Effect of aggregate type on damage factor for different aspect ratios (l/d).

(l/d) ratio on core damage is minimal. It is evident from Fig. 16

that the presence of natural aggregate (gravel) has more harmful effect on the extracted core as compared to crushed aggregate (pink lime stone). Strength reduction up to 1420% was
noticeable in cores made with crushed and natural aggregates,
respectively.
In fact, the drilling operation weakens the bonds between
the aggregate and the surrounding matrix. In high strength
concrete where the matrix-aggregate bond is higher and the
transition zone is more cohesive, the damage in core specimen
is low. Actually, during the coring operation pronounced
shearing forces between the coring bit and the concrete surface
are developed which would cause greater damage to low
strength concrete as compared to higher strength concrete

Concrete strength = 18 MPa

Figure 17

(Fig. 17) illustrates the relation between the damage factors

according concrete strength level.
The statistical analysis has been performed using SAS
and Data Fit software and a signicant number of core test
results as shown in Fig. 18. The formula for correction due to
damage effect (Fdmg) is as follows:
Fdmg: 2:4 

l=d0:006
d0:1  fcore a 

11

where fcore is the core strength in kg/cm2; d is the core diameter

in mm, l/d is the core aspect ratio, and a is a constant depending on type of aggregate that is equal to 0.06 in crushed pink
lime stone.

Concrete strength = 30 MPa

Concrete strength

Effect of concrete strength on damage factor for different aspect ratios (l/d).

Reliability of core test Critical assessment

179

Figure 18 Advanced statistical analysis for formulating the damage factor (Fdmg) using Data Fit Software & 250 results of cores as
affected by multiple variables (crushed agg.-concrete).

6.1.4. Effect of coring direction

According to ACI 214.4R-03 [15], cores drilled vertically (in
the direction of placement and compaction) can be stronger
than cores drilled horizontally (perpendicular to the placements direction. The difference in core strength due to drilling
direction is generally attributed to bleeding in fresh concrete,
which creates a weak paste pockets underneath coarse aggregate particles resulting in weak paste-to-coarse aggregate
bond. Results indicate that the mean ratio between the
strength of core drilled vertically to that of core drilled horizontally varies between 1.075 and 1.08 as shown in Fig. 19.
The gure implies that the effect of coring direction is almost
undependable on aspect ratio or concrete strength level.
On the basis of the argument mentioned above, it may be
reasonable to assume that the correction factor (Fdir) that accounts for coring direction is constant and can be considered
equal to 1.0 and 1.075 for horizontal and vertical coring,
respectively.
6.1.5. Effect of reinforcing steel
It is strongly believed that a reduction in measured core
strength occurs for a core containing reinforcing steel (other
than along its axis) depending on concrete strength level.
The extent of this reduction also depends upon many other
variables, and may range from 0% to 20% (when large diameter or multiple bars are present and associated with low level
of concrete strength). The correction factor to account for this

Concrete strength = 18 MPa

Figure 19

reduction (Freinf) depends on several variables including level

of concrete strength, number and diameter of reinforcing bars,
bars location with respect to core axis nearest end of core.
To get a reliable expression for the factor (Freinf) was very
complicated process. Results of about 230 cores were considered with 18 steel congurations that are different with respect
to steel bars diameter, number of bars, bars locations with respect to vertical and horizontal axis, and others. In conducting
these comparisons, it was attempted to keep other parameters
comparable to evaluate only one variable at a time. A number
of different comparisons are shown in Table 6. For example,
the main variable between Specimens A1 and A8 is the diameter of steel. Results are presented in full details elsewhere [1];
however, some points are outlined herein. Results of this research conrm that the presence of steel in drilled cores reduces core strength. Large diameter-bar adversely affects
core strength. The reduction is more signicant for low
strength concrete (18 MPa). The presence of one 16 mm-bar
in core samples with l/d = 1 reduces the core strength by about
9%. Strength reduction on the order of 25% is associated with
the presence of one 22 mm-bar. The core strength is affected by
the horizontal location of the bar. A 25 mm-deviation of
10 mm-bar from the vertical axis of the core causes additional
strength reduction; up to 20%, as compared to comparable
core specimen but without bars deviation providing other
variables are comparable. The noticeable strength reduction
is due to the damage through cutting operation and the developed stress concentration around steel bars.

Concrete strength = 30 MPa

Concrete strength = 48 MPa

Effect of coring direction on core strength for different aspect ratios (l/d).

180
Table 6
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
A12
A13
A14
A15
A16
A17
A18

S. Khoury et al.
List of comparisons between tested cores to determine.
A18

A17

A16

A15

A14

nd

nd

A13

A12

A11

A10

A7

A6

A5

A4

A3

A2

A1

nd

nmd
nmd

A8
d

nd

A9

nmd
nmd

nm

d
nmd

d
d

d Diameter of steel bar.

m Distance of steel bar from nearly end of core.
n Number of steel bars and spacing between bars.
Distance of steel bar from vertical axis of specimen.

This brief review indicated that the various proposed relationships for correction factors are all nonlinear. It should
be noted that the equations given by the Egyptian Code takes
into account most variables that may affect the interpretation
of the results; however, the code ignores the deterioration of
steelconcrete bond that may occur and also the position of
the reinforcement from vertical axis of core specimens.
Weighted nonlinear regression analysis has been performed
to determine the factor (Freinf) with the use of the software
SAS package and Data Fit. This shows that the correction factor for reinforcement (Freinf) is given by the following
expression:
 For cores containing a single bar:


Ur  r Ur  S=10
1:13
 0:015
Freinf 1 1:5
Uc  L
fcore

12

 For core specimen containing two bars no further apart

than the diameter of the larger bar, only the bar corresponding to the higher value of (Ur  d) is considered. If
the bars are further apart, their combined
effect is assessed
P
by replacing the term (Ur  r) by ( Ur  r as follows:

Figure 20

P


Ur  r Ur  S=10
1:13
 0:015
Freinf 1 1:5
Uc  L
fcore

13

where Freinf is the correction factor for reinforcement, Ur is the

diameter of the reinforcement, Uc is the diameter of the concrete specimen, r is the distance of axis of bar from nearer
end of specimen, S is the distance of axis of bar from axis of
core specimen, L is the length of the specimen after end preparation by grinding or capping, and fcore is the concrete core
strength (kg/cm2).

6.1.6. Effect of moisture condition of core

Results of about 100 cores indicate that the strength of cores
left to dry in air for 7 days is on average 13% greater than that
of cores soaked at least 40 h before testing. The strength of
cores with negligible moisture gradient and tested after cutting
is found to be 79% larger than that of soaked cores as shown
in Fig. 20. The authors strongly recommend to use a correction
factor accounting for moisture condition (Fm) equals to 1.09
and 0.96, respectively, for cores tested after 48 h soaked in
water and for those tested after 7 days dry in air.

Effect of core moisture condition on core strength for different aspect ratios (l/d).

Reliability of core test Critical assessment

181

Table 7 Effect of strength level on the compressive strength of

drilled cores.
Strength level of cast
specimens (Standard cube), MPa

Strength
reduction (%)

18
30
48

1216
712.5
67

48 MPa

30 MPa
Group 'C'
of cores

Column with w/c = 0.4

Column with w/c = 0.6

18 MPa
Group 'B'
of cores

Top
Group 'A'
of cores

3/4

Results of 500 cores

Mid

Figure 23 Reliability of approach included in the Egyptian Code

for core analysis.

1/4
Bottom

Relative strength, %

Figure 21

Strength variation through columns height.

6.2. Concrete characteristics

6.2.1. Level of concrete strength
On the basis of the collected data, the effect of strength level on
the compressive strength of drilled cores is presented in Table 7.
It is evident from the table that the percentage of strength
reduction in cores reduces with the increase in concrete
strength level. It seems that in high strength concrete on the order of 50 MPa, the damage effect becomes negligible.
6.2.2. Strength variation through columns height
Concrete strength varies within the same element as a result of
bleed water especially in vertical members. To explore this
point, cores were drilled at different levels in columns.
Fig. 21 shows the strength variation throughout the column

Figure 22

length based on core tests. Results indicate that cores drilled at

the top portion of column are found to be 1020% weaker
than cores extracted from the middle or lower portions, respectively, depending on watercement ratio (w/c). The gure reveals a change in concrete strength with changing location,
but the change seems to be not pronounced and can be neglected in the lower 75% of the columns height. It is therefore
advisable not to drill cores in the upper quadrant of concrete
column or walls to avoid incorrect judgment.
6.2.3. Type of aggregate
The relationship between compressive strength of drilled cores
and comparable molded cylinders is shown in Fig. 22 for core
specimen made with pink limestone and gravel. The cores were
drilled from concrete blocks and were tested immediately after
drilling. The blocks had been moist cured for 7 days and were
then left to dry under lab condition for 28 days. The molded
cylinders (4 8 in.) were also moist-cured. As expected,
concrete made with pink limestone achieves strength higher
than that obtained by concrete made with natural aggregate
(gravel) although the w/c ratio of the gravel-concrete was

Effect of steel bars on the strength of poured and drilled cores.

182

S. Khoury et al.
improve compressive strength while the presence of the same
steel in drilled core has harmful effect on core strength due
to damage effect.
7. Critical assessment

Figure 24 Reliability of approaches included in different codes

for core analysis.

lower than that of limestone-concrete. This attributes to the

weaker interfacial transition zone between the aggregate and
the cement paste due to the smooth surface of the natural
aggregate. It is clear from the gure that all drilled cores yield
compressive strength less than that achieved by similar cylinders that have comparable aspect ratio, diameter, and identical
number and location of steel bars. Again, the graph points out
that the presence of steel in poured cylinders may slightly

Table 8

A total of 500 cores covering all studied variables are considered herein to critically asses the approach given by the Egyptian Code. Fig. 23 correlates between the measured core
strength and the estimated in-situ cube strength using the code
approach. Unfortunately, the Egyptian Code approach fails to
predict actual concrete strength. The error varies from 5% to
65%. Fig. 24 conrms that none of the examined approaches
(Egyptian Code, ACI Code, European Code) can give promising prediction. It should be noted that Fig. 24 was constructed
using only 240 cores that do not contain steel bars. The ACI
approach seems to be the closest one to the experimental data;
however, if cores with reinforcement are to be introduced the
ACI Code may give unsatisfactory prediction since it does
not include any equations for to account for the presence of
steel in cores. At his point, more dependable approach for core
interpretation seems to be ultimately required.
Line of Equality
8. Proposed model for interpreting core test results
Based on experimental evidences and the use of two powerful
softwares as mentioned earlier, the following model is proposed and may be considered as a modication to the current
code provisions for core interpretation. According to the new

Proposed strength correction factors for interpretation of core test results.

Factors

Proposed formula/magnitude

Fl/d: aspect ratio

Fl=d
1

h

75
f1:15
core

1

2dl

Fd: core diameter (50, 75, 100 and 150 mm)

Fd 1:507  0:11 lnd

Fdir: coring direction

Parallel of casting
Perpendicular of casting

1.0
1.075

Frienf: reinforcing steel

Cores containing a single bar:
Cores containing two bars further apart than the diameter of the larger bar:

h
i
r S=10
 f1:13
Freinf 1 1:5 Ur rU
0:015
Uc L

 core
P
Ur rUr S=10
Freinf 1 1:5
 f1:13
0:015
Uc L

Fm: core moisture condition

After cutting
Socked 48 h
Air dried for 7 days

1.00
1.09
0.96

Fdmg: damage due to drilling

Fdmg: 2:4  dl=d

0:1
f

Notations:
fcore: core strength (kg/cm2).
d: core diameter (mm).
Ur: diameter of reinforcement (mm).
Uc: diameter of the concrete specimen (mm).
r: distance of axis of bar from nearer end of specimen (mm).
S: distance of axis of bar from axis of core specimen (mm).
L: length of the specimen after end preparation by grinding or capping (mm).
fcore: concrete core strength (kg/cm2).
l/d: aspect ratio after capping.

core

0:006
core

Reliability of core test Critical assessment

183

procedure, the in-situ concrete cube strength (fc,is) can be

predicted by converting the strength of non-standard cores
(fc,NS) into equivalent strength of standard cores (fc,S), and
then by converting the last (fc,S) into an equivalent in-situ concrete strength (fc,is). The core test result may be biased if the
test specimen differs from a standard test specimen that is
4 in. (100 mm) in diameter by 8 in. (200 mm) long, does not
contain any pieces of reinforcing bar, and the cutting direction
is parallel to casting direction.
8.1. General form
The general form of the model may be written as follows:
fc;S Fl=d Fd Frienf Fdir fc;NS

14

where fc,S and fc,NS are the strengths of the standard and
Non-standard core specimens respectively. The strength
correction factors (Fl/d), (Fd), (Fdir) and (Frienf) account for
the effects of length to diameter ratio, core diameter, coring
direction, and the presence of reinforcing bar pieces on the
strength of non-standard core. These factors can be calculated
using Equations from (9) to (13) presented earlier and is summarized in Table 8.
The in-situ concrete cube strength (fc,is) is modeled herein as
follows:
fc;is Fm Fdmg fc;S

Figure 26

Calibration of proposed approach.

the proposed approach is more reliable for interpreting core

results than other available methods (Fig. 24). In fact, the proposed model considers large number of variable and is based
on comprehensive experimental program that included about
500 cores in addition to more than 300 specimens.

15

where (fc,is) is the in-situ concrete cube strength and factors

(Fm) and (Fdmg) account for the effects of the moisture condition of core and damage sustained during drilling, respectively,
on the strength of the standard core specimen.
8.2. Calibration of proposed model
The proposed approach is calibrated against large number of
experimental data in Fig. 25 where the estimated in-situ cube
strength using the approach is plotted against actual cube
strength. The gure shows good agreement indicating the
validity of the proposed method. Also, Fig. 26 conrms that

Figure 25 Comparison between the actual and estimated cube

concrete strength.

9. Conclusions
Based on the comprehensive experimental study reported herein the following conclusions can be drawn:
1. Core test is commonly required in the area of concrete
industry for assessment of concrete quality. In fact, core
test becomes a must in many critical circumstances and
sometimes it is the unique tool for safety assessment in
existing structure.
2. Core test is included in most international codes provisions including Egyptian Code ECP203, ACI Code,
British Code, European Code and others. Core test
may be reliable; however the interpretation of the results
to in-situ concrete strength is questionable. Actually,
extensive literature survey indicates that different codes
give different in-situ strength from one core test result.
3. A comprehensive study was undertaken in this research
to examine the factors affecting the interpretation of
core test results. More than 500 cores were prepared
and tested as well as more than 300 concrete specimens.
Actually, the program was very exhausted, costly and
time consuming. Variables studied are core aspect ratio,
core diameter, concrete strength level, the presence of
reinforcing steel, coring direction, core damage due to
drilling, type of coarse aggregate, core moisture condition, as well as the core location in vertical members
with respect to height.
4. Results indicate that the core compressive strength
increases with the decrease in the core aspect ratio
(l/d); however, this effect becomes negligible for high
strength concrete.
5. The effect of core diameter on core strength is completely different that in case of molded concrete cylinder.
It is generally agreed for molded concrete that the

184

6.

7.

8.

9.

10.

11.

12.

13.

14.

S. Khoury et al.
concrete strength is decreased as the specimen size
increases. However, in case of drilled cores, as the diameter
decreases, the ratio of cut surface area to volume increases,
and hence the possibility of strength reduction due to cutting damage increases. Strength reduction up to 17% was
recorded in cores with diameter less than 100 mm. On the
other hand, for cores with larger diameter than 100 mm,
this damage effect is minimal but should be considered.
The damage effect due to core drilling is signicant for
low strength concrete (18 MPa). In fact, the drilling
operation weakens the bonds between the aggregate
and the surrounding matrix. In addition, concrete made
of gravel is subjected to damage during core drilling
much more than concrete made with pink lime stone.
The measured strength of cores drilled vertically (parallel
to casting direction) is greater than that for a horizontally
drilled core (normal to casting direction) providing other
variables are comparable. The difference is about 8%.
The presence of reinforcing steel in the core samples
reduces the measured core strength. A strength reduction up to 25% was recorded for core contained
22 mm-bar. The noticeable strength reduction is due to
the damage through cutting operation and the developed stress concentration around the existing steel bars.
The moisture condition of the core by the time of testing
affects its strength. Contradictory, the strength of core
specimen left in air for 7 days to dry achieved about
12% increase in the core strength.
Comparison between the actual concrete cube strength
and the estimated values using different approaches recommended by many codes is unsatisfactory.
The ACI approach seems to be the closest one to the
experimental data providing that steel bars do not exist
in core samples. More dependable approach for core
interpretation seems to be ultimately required.
On the basis of comprehensive collected data obtained in
this research through testing large number of cores, statistical analysis has been performed to determine reliable
strength correction factors that account for the studied
variables. A simple weighted regression analysis of a
model without an intercept was carried out using the
SAS Software package (SAS Institute, 2008) as well
as Data Fit software.
A performance-based model for interpretation of core
test results is proposed in this research. The new
approach considers all factors that may affect core
strength. The model when calibrated against large number of test data shows good agreement.
Based on experimental evidences, it can be stated that
the proposed model to estimate the in-situ concrete cube
strength from the result of core test seems to be very
promising.

References
[1] O. Arioz, M. Tuncan, K. Ramyar, A. Tuncan, Assessing
concrete strength by means of small diameter cores, Constr
Build Mater 22 (2008) 981988.

[2] I.M. Nikbin, M. Eslami, D.S.M. Rezvani, An experimental

comparative survey on the interpretation of concrete core
strength results, Eur. J. Sci. Res. 37 (3) (2009) 445456, ISSN
1450-216.
[3] A.M. Neville, Properties of Concrete, Addison-Wesley, UK, 1995.
[4] J.H. Bungey, Determining concrete strength by using small
diameter cores, Mag. Concr. Res. 31 (107) (1979) 9198.
[5] F.M. Bartlett, J.G. MacGregor, Effect of core diameter on
concrete core strengths, ACI Mater. J. 91 (5) (1994) 460470.
[6] F.M. Bartlett, J.G. MacGregor, Effect of core length-todiameter ratio on concrete core strengths, ACI Mater. J. 91 (4)
(1994) 339348.
[7] F.M. Bartlett, J.G. MacGregor, Effect of moisture condition on
concrete core strengths, ACI Mater. J. 91 (3) (1993) 227236.
[8] F.M. Bartlett, Precision of in-place concrete strengths predicted
using core strength correction factors obtained by weighed
regression analysis, Struct. Saf. 19 (4) (1997) 397410.
[9] F.M. Bartlett, J.G. MacGregor, Cores from high-performance
concrete beams, ACI Mater. J. 91 (6) (1994) 567576.
[10] ASTM C 42/C 42M-90, Test for obtaining and testing drilled
cores and sawed beams of concrete, Annual Book of ASTM
Standards, 2008.
[11] BS 1881: Part 120: 1983, Method for Determination of the
Compressive Strength of Concrete Cores, British Standards, 1983.
[12] TS EN 12504-1, Testing Concrete in Structures-part 1: Cored
Specimens-taking, Examining and Testing in Compression,
Turkish Standards Institute, Ankara, 2002.
[13] CSTR 11, Concrete core testing for strength, Technical Report
No.11, The Concrete Society, London, 1976, p. 44.
[14] A. Ghazi, Reliability of core test, M.Sc. Thesis, Faculty of
Engineering, University of Alexandria, Alexandria, Egypt, 2011,
151pp.
[15] ACI Committee 214.4-03, Guide for Obtaining Cores and
Interpreting Compressive Strength Results American Concrete
Institute, Farmington Hills, Mich., 2003, 16pp [2013].
[16] M. Dolce, A. Masi, M. Ferrini. Estimation of the actual in-place
concrete strength in assessing existing RC structures, The
Second International b Congress, June 58, 2006, Naples,
Italy, 2006.
[17] BS 1881: Part 120: 1983, Method for Determination of the
Compressive Strength of Concrete Cores, British Standards, 1983.
[18] ACI Committee 214, Recommended Practice for Evaluation of
Strength Test Results of Concrete (ACI 214-77), American
Concrete Institute, Farmington Hills, Mich., 1977, 14pp.
[19] ASTM Committee C 39-68, Standard Test Method for
Compressive Strength of Cylindrical Concrete Specimens,
Annual Book of ASTM Standards, V. 04.02, ASTM,
Philadelphia, 1990, pp. 2023.
[20] BS EN 12504-1, Testing Concrete in Structurespert !: CoredTaking, Examining ar.u Testing in Compression/ 3ri:ish
Standards, 2000, 12pp.
[21] British Standards Institution, BS 1881: Part 120, Method for
Determination of the Compressive Strength of Concrete Cores,
London, 1983, 6pp.
[22] Comite Europeen de Normalisation (CEN), Design of
structures Part 3: Assessment and retrotting of building, EN
1998-3, 2005.
[23] Concrete Society, Concrete Core Testing for Strength, Technical
Report No. 11, The Concrete Society, London, 1987, 44pp.
[24] TS 10465, Test Methods for Concrete-Obtaining Samples and
Determination of Compressive Strength in Structures
(Destructive Method), Turkish Standards Institute, Ankara,
Turkey, 1992, 16pp.