CONCRETE ADVICE No.

47

Assessment of in-situ concrete strength

using data obtained from core testing

Neil Crook, PhD CEng MICE MICT

When the in-situ strength of the concrete is required, cores are often taken. However
unless they are taken correctly, in the correct place and in sufficient numbers, the
assessment of the in-situ strength becomes less assured and there is less confidence that
the strengths obtained are a true reflection of the concrete strength.

1 Introduction 3 Cube/core relationship

There are a variety of methods that can be
used to determine in-situ concrete strength.
The test procedures associated with these
methods are given in BS EN 12504 parts 1,
(1)
2, 3 and 4 . This advice sheet considers
the assessment of in-situ concrete strength
using data obtained from core testing. The
other non-destructive tests methods can be
used on their own or in conjunction with core
testing.
2 Publications
Until recently the procedures for taking and
testing cores and assessing the in-situ
strength of concrete has been based on BS
(2) (3)
1881:120 , BS 6089: 1981 and Concrete
(4)
Society Technical Report TR 11 .
With the publication of BS EN 12504-1 and
(5)
BS EN 13791 , the standard BS 6089:
(6)
2010 has been completely re-written to
complement BS EN 13791 and provide
information on other techniques and the use
of established statistical principles when
investigating unknown structures. As a result
TR 11 is no longer regarded as best
practice.
There is no unique relationship between
the core and cube strengths of the same
concrete. The relationship is a function of
many variables including:
1 Temperature
2 Compaction
3 Curing
4 Sampling location – top, middle or
bottom
5 Element type and thickness –
horizontal or vertical
In most structures the in-situ strength, as
measured by a core, will generally be
lower than the standard cube strength of
the same concrete.
4 Current Standards
In the UK, the method considered most
relevant for assessing the concrete
strength in the structure is the
determination of the estimated in-situ
cube strength (EICS) from cores as it
exists at the sampling location, without
correction for the effect of curing history,
age or degree of compaction.
 Page 2 of 8
(5) (6)
BS EN 13791 , together with BS 6089 , gives
procedures for;
Assessing whether concrete conforms to the
specification (CASE A)
This is where cores are taken in order to resolve a
dispute over the strength of the concrete supplied.
Despite confirming that the concrete supplied has
satisfied the strength, there may be situations when
the structural element is shown to be inadequate
(7)
when assessed using BS EN 1992-1-1 Annex A as
it is the minimum valid in-situ strength that is used in
the calculation.
Determination of characteristic in-situ strength
(CASE B)
This is undertaken when:
a) there is no information about the concrete or
b) the producer has declared the concrete as
nonconforming with respect to compressive
strength.
The objective in this case is to find the characteristic
strength based on the mean in-situ strength.
5 Where and how many cores
Before embarking on core testing it is important to
ascertain the reason for the testing as it will determine
where and how many cores are required. In all cases
highly stressed sections, reinforcement and
prestressing steel and ducts should be avoided.
The test locations should be chosen such that after
cutting and end-preparation, the length/diameter ratio
of the core should not be less than 1.0 and ideally not
greater than 1.2 and the core does not contain:
• Concrete from within 50mm of the surface (could
be difficult with thin slabs e.g. composite
construction).
• Concrete from within the top 50mm or 20%
(whichever is greater) of the lift in sections where
height of depth of the element is not more than
1.5m
• Concrete from the top 300mm of the lift where
the height or depth is 1.5m or more.
When deciding where and how many cores are
required, it is important to have clarity and agreement
on the reason for carrying out core testing and
whether or not they are to be used in conjunction with
NDT.
CONCRETE ADVICE NO. 47
The strength of the concrete in a vertical element will
vary over its height. If the core data is to be used for
the determination of the quality of the concrete i.e.
cases of suspect concrete from identity test data
(CASE A), the cores should be taken from the middle
third of the vertical element. When assessing in-situ
characteristic strength, cores should be taken from
the upper third of the element (excluding the top 300
mm). In both cases, for thin horizontal sections the
test locations should be selected at random.
The number of test locations will depend on the
purpose of the testing and if it is to be used in
conjunction with data obtained from NDT. When
assessing for conformity with the specification
(CASE A) a minimum group of 4 cores (2 cores from
2 locations) need to be taken to represent a test
region. For the assessment of a test region
containing a large volume of suspect concrete
(CASE A) or for in-situ characteristic strength
determination (CASE B), where the assessment is
based on core data alone, it is recommended that a
minimum of 15 cores, preferably more to allow for
outliers, are taken. When a long core is taken which
is subsequently cut to produce 2 core samples for
testing it is the average of the two strengths which is
considered to be the test result from the location.
The definition of a test region is ambiguous and
unclear. It is usually associated with a batch of
concrete which could be made up of a single or
several loads of concrete. It could also be related to
the volume of concrete identified as failing the criteria
for identity testing. Sequential identity testing group
failures making up many batches could still constitute
a single test region.
The decision to assess each test region separately
or collectively is open to discussion. However, the
greater number of sequential group failures, the
greater the number of cores that are needed. For a
test region containing a large volume of suspect
concrete, it is possible that the number of cores may
be 15 or more and here an additional check needs to
be made on the statistical certainty that a conforming
concrete has been supplied.
Assessment of the data will involve checking that the
lowest individual core result (for suspect concrete) or
the mean value of a group (for characteristic
(6)
strength) is valid. BS 6089 gives guidance on a
simple way of testing for statistical outliers and
should be applied for each data set. As there is a risk
that an outlier may be found it is advisable that
additional cores are taken to ensure that the analysis
can be carried out with either a minimum of 15 or 4
results depending on which assessment is being
carried out.
 Page 3 of 8
Note: Although not stated, the number of iterations to
assess outliers should be restricted to 2 passes. If
more are required to satisfy the criterion, the cores
may have come from more than one population i.e.
different concretes.
6 Influences on the testing result
Size
Once the core has been taken it needs to be prepared
(1)
and tested in accordance with BS EN 12504-1 . The
diameter of the core should have been chosen such
that it is at least 3.5 times the declared value of the
coarsest fraction of aggregates actually used in the
concrete. Smaller ratios can have a significant effect
on the strength of the core specimen. For example,
tests have indicated that for 20mm aggregate a
100mm diameter core was approximately 7% stronger
than 50mm diameter core which in turn was
approximately 20% stronger than a 25mm diameter
core. In the absence of any information on the
aggregate size, the default core diameter should be
100mm.
Moisture condition
The condition of the core at the time of test is
important as the assessment of in-situ strength is
based on the premise that the concrete is in a dry
state. If the core is to be capped it will probably be wet
at the time of testing and have a moisture content
between 8% - 12%. Preparation of the ends by
grinding will give the option for the core to be classed
as ‘dry’ when tested. Dry is considered to be after the
core has been exposed to air for at least 3 days.
Testing ‘wet’ will give an apparent reduced strength
from one tested ‘dry’ of between 10-15%. If the cores
are tested immediately after water storage, the UK
recommendation is to increase the measured strength
by 10%. The moisture condition from wet i.e. water
storage to ‘dry’ i.e. exposed to air for 3 days is a
judgement and an estimate of its condition should be
made at the time of testing and an appropriate factor
between 0 and 10% should be applied.
Reinforcement
It is preferential to avoid reinforcement when taking
cores as it will have an influence on the recorded
strength but this is often not possible. When present,
factors need to be applied based on location and size
of the reinforcing bar(s) within the core so that the
strength obtained is an estimate of the concrete
alone. Results adjusted in this way should be treated
with some degree of suspicion.
CONCRETE ADVICE NO. 47
Shape factor
In previous standards there has been an adjustment
for the direction of drilling. However recent
(8)
research has found that this is not appropriate.
Factors are applied to convert the actual size of the
tested specimen (including any capping) to either an
equivalent 1:1 cube or 2:1 cylinder.
Result
The outcome of applying these factors to the
measured tested strength is to produce an Estimated
In-situ Cube or Cylinder Strength (EICS). The symbol
(6)
f is is used in the formulas to be found in BS 6089 .
Visual assessment and density
In addition to recording the density of the core, which
can give a good indication if the sample is
representative of the concrete being tested, there
needs to be an estimation of the voidage of the
tested concrete in excess of the normally expected
value of 0.5% representing well compacted concrete
in standardised cubes or cylinders. The effect of
excess voidage is to reduce the strength of the
concrete in-situ and can give an indication on the
quality of workmanship employed in placing and
compacting the concrete.
7 Assessment
In the case of assessing suspect concrete arising out
of identity test failures (CASE A), valid results for
each test region are checked against the criteria
based on the relationship between the characteristic
in-situ strength and characteristic strength (f ck ) from
(9)
standardised specimens. As BS EN 206-1 permits
a single acceptable batch of concrete to have a
compressive strength of (f ck – 4) MPa the minimum
in-situ strength from a test region f is,lowest needs to be
greater than 0.85(f ck, spec – 4) to show that a local
area satisfies the conformity criterion for minimum
compressive strength. If there are at least 15 valid
core results an additional check is required whereby
the mean of the core strengths f m(n)is needs to be at
least 0.85(f ck, spec + 1.48s), where s is the standard
deviation. If both criteria are satisfied the test region
may be deemed to contain concrete with adequate
strength and the concrete in the region conformed to
(9)
BS EN 206-1 .
For assessing an unknown concrete (CASE B), BS
6089 does not consider the procedures given in
(5)
BS EN 13791 for calculating in-situ characteristic
strength as applicable.
 Page 4 of 8 CONCRETE ADVICE NO. 47

The procedure is replaced by the well-established

t-statistic method applied to the mean of all the valid
results. The in-situ characteristic strength calculated is
checked against the lowest individual corrected core
strength +4MPa. The lower of the two values is then
used as the characteristic in-situ strength for structural
assessment. On obtaining this value reference should
(7)
be made to BS EN 1992-1-1 Annex A for the design
or redesign of the structure.
9 Example 1: Individual corrected in-situ cube
strength (f is ) calculation
On the assumption that the core has been taken,
prepared and examined in accordance with BS EN
(1)
12504-1 , Table 1 shows an example of an
individual corrected in-situ cube strength (EICS or f is )
calculation. The full requirements of the core test
report are given in BS EN 12504-1.

8 Potential Strength
(6)
Potential strength, discussed in BS 6089 Annex A
should not to be confused with the in-situ strength
related to assessments for suspect or unknown
concrete. Potential strength forms the basis for
discussion between the supplier and constructor when
lower than expected strengths are obtained. Potential
strength will also be dependent on a number of other
factors i.e. curing, strength gain, which have been
based on concrete containing CEM I some 35 years
ago. For this reason acceptance of potential strength
in dispute resolutions can be problematical.
The test report should also give an estimation of the
excess voidage of the core i.e. voidage in excess of
0.5%. Excess voidage up to 1.5% is not unusual.
Greater values would suggest inadequate
compaction of the in-situ concrete.
This might be important in a dispute resolution
between the supplier and contractor when the
specified strength class has not been proven from
the core data.

Table 1: Example of an individual corrected in-situ cube strength (EICS or f is ) calculation

Example
Reference Notes
figures
Core length after capping or grinding (mm) Measured (l) 110
Core diameter (mm) Measured (d) 100
Maximum load (kN) Measured (L) 215.0
2
Area (mm ) Calculated (A) 7855
Core compressive strength, f core (MPa) Calculated (L/A) 27.4

λ Calculated core length/diameter ratio (l/d) 1.10

Shape correction factor for cubes,* K is K is = 2.5/(1.5 + [1/ λ]) 1.04

Diameter of the bar (mm) Measured (φr) 12

Diameter of the core (mm) Measured (φc) 100
Distance of axis of bar from the nearer end of
Measured (h) 45
core (mm)
Multiple bar correction factor, K s K s =1.0+1.5((∑[φr x h])/([φc x l]) 1.07

Core tested wet or dry or unknown condition Wet

K m = 1.0 if tested dry or moisture condition
Strength correction for cores tested wet
unknown 1.1
or dry or in an unknown condition, K m
K m = 1.1 if tested wet

f is corrected for shape, rebar and

f is = f core x K is x K s x K m = EICS 33.5
moisture condition, f is,corr (MPa)
* If determining the in-situ cylinder strength K is = 2.0/(1.5 + [1/ λ])
 Page 5 of 8 CONCRETE ADVICE NO. 47

For 4 cores, if R t > 0.202 the f is, lowest result is

10 Example 2: Assessment of the in-situ
strength of a test region where identity
testing has not confirmed conformity with
the strength class (CASE A)
Small volume
Where a test region contains a small volume of
concrete comprising of one or a few batches is under
suspicion, due to a set of cube data not achieving the
identity test criteria, at least 4 cores should be taken
(1)
and tested in accordance with BS EN 12504-1 .
suspicious.
If R t > 0.298 the f is, lowest result should be rejected
Therefore as R t = 0.090 < 0.202, the lowest result
(S3) is valid.
(5)
From BS EN 13791 Cl. 9, for a small test region
with less than 15 valid core results, the region may be
deemed to contain concrete with adequate strength if
the lowest corrected in-situ cube or cylinder
compressive strength test result complies with:

An example is given in Table 2 for 4 cores S1 to S4. (f is,lowest ) ≥ 0.85(f ck – 4)

The validity of the lowest core result (S3) needs to be Where f ck = characteristic compressive strength of
checked by calculating R t ; standard specimens (i.e. strength class).

R t = (Mean of other core results − Lowest result) For f ck = 40

Mean of other core results 0.85(f ck – 4) = 30.6
(6)
If R t exceeds the value given Table 5 of BS 6089 , From Table 2, f is,lowest = 29.4 i.e. < 30.6
the core result should be treated as being suspicious
or rejected as appropriate. Therefore the region tested has failed to meet
strength criteria and suggests the concrete has not
For the data set in Table 2; come from a conforming population. Thus the design
assumptions are not valid and the structure should be
f is, lowest (result S3) = 29.4 assessed for structural adequacy by using a reduced
Mean of other f is core results = 32.3 strength class or by calculating the in-situ
characteristic strength, see CASE B.
Hence:
R t = 0.090

Table 2: Data and calculations from cores taken from a small test region
Core reference S1 S2 S3 S4
Length (l) 110 99 100 101
Diameter (d) 100 100 100 100
Maximum load (L) 215 250 210 190
Core compressive strength - f core 27.4 31.8 26.7 24.2
λ (l/d) 1.10 0.99 1.00 1.01
Shape correction - K is 1.04 1.00 1.00 1.00
Bar 1 diameter (φr 1 ) 12 - - 12
Distance from nearest end (h 1 ) 45 - - 45
Bar 2 diameter (φr 2 ) - - - 12
Distance from nearest end (h 2 ) - - - 60
Reinforcement correction - K s 1.07 1.00 1.00 1.19
Tested wet or dry or unknown condition Wet Dry Wet Wet
Moisture correction - K m 1.1 1.0 1.1 1.1

f is = f core x K is x K s x K m 33.5 31.8 29.4 31.7

f is,lowest 29.4
Note. It is not typical for wet and dry cores to be in the same data-set. This has been given as an example.
 Page 6 of 8 CONCRETE ADVICE NO. 47

Large volume
Where a larger volume of concrete is under suspicion
as a result of many continuous cube data sets failing
to meet the identity test criteria, 15 valid cores results
are required. To allow for rejection of any invalid
results it is suggested at least 17 cores are taken to.
Note: If failing data sets are not continuous, the
assessment of in-situ strength should be based on a
small volume test region principle for each data set.
The dataset in Table 3 of estimated in-situ cube
strength values for 40 core results is taken as an
example calculation. For this data set;
For 39 cores, if R t > 0.099 the f is, lowest result is
suspicious.
If R t > 0.118 the f is, lowest result should be rejected
Therefore as R t = 0.093 < 0.099, the lowest result
(A17) is not suspicious and should be accepted.
If this result had also been categorised as being
rejected, then further investigation needs to be
carried out as there may be two populations present
suggesting two different concretes have been
supplied.

Lowest EICS (f is, lowest ) result (A11) = 32.0 Therefore

Mean of other f is core results = 37.4 f is, lowest = 34.0
f m(39),is = 37.4
Hence: s = 1.518
R t = (37.4 – 32.0)/37.4 = 0.144
where s is the standard deviation of the 39 results
(6)
If R t exceeds the value given Table 5 of BS 6089 ,
(5)
the core result should be treated as being suspicious From BS EN 13791 Cl. 9, for a test region
or rejected as appropriate. comprising many batches of concrete with 15 or more
valid core results may be deemed to contain concrete
For 40 cores, if R t > 0.099 the f is, lowest result is with adequate strength if complies with both:
suspicious.
f m(n),is ≥ 0.85(f ck +1.48s) Criterion 1
If R t > 0.118 the f is, lowest result should be rejected and
f is,lowest ≥ 0.85(f ck – 4) Criterion 2
Therefore as R t = 0.144 > 0.118, the lowest result
(A11) should be removed from the dataset. Therefore for C32/40 strength class f ck = 40

The procedure is repeated on the next lowest f is 0.85(40 +1.48 × 1.518) = 35.9 Criterion 1
value (A17) 0.85(40 – 4) = 30.6 Criterion 2

Lowest EICS (f is, lowest ) result (A17) = 34.0 Hence:

Mean of other f is core results = 37.5 f m(n),is and f is,lowest conform to criteria 1 and 2
respectively and as both criteria have been met, the
Hence: region tested is deemed to contain concrete
R t = (37.5 – 34)/37.5 = 0.093 conforming to the specified strength class.

Table 3: Estimated in-situ cube strength from cores taken from large test region (40 results low to high)
Ref f is Ref f is Ref f is Ref f is
A11 32.0 A28 36.2 A07 37.1 A23 39.0
A17 34.0 A34 36.3 A15 37.1 A35 39.0
A24 35.7 A37 36.3 A29 37.1 A36 39.0
A06 35.8 A32 36.4 A38 37.1 A04 39.1
A16 35.9 A39 36.4 A40 37.5 A14 39.2
A03 36.0 A20 36.6 A01 37.6 A09 39.3
A27 36.0 A33 36.7 A19 37.9 A12 39.4
A10 36.1 A05 36.9 A22 38.0 A21 39.9
A25 36.1 A18 36.9 A31 38.3 A08 40.0
A13 36.2 A26 36.9 A02 39.0 A30 41.0
 Page 7 of 8 CONCRETE ADVICE NO. 47

This procedure is valid for 4 or more cores but for

11 Example 3: Assessment of characteristic in- small numbers of test data, the likelihood that an
situ compressive strength of a structure unknown structure may contain (unknown) weaker
where the strength of the concrete is areas increases.
unknown or where the producer has declared
non-conformity (CASE B) In such circumstances, it is advisable to take a more
conservative approach and use a lower core strength
BS 6089
(6)
Clause 6.2 details the alternative for structural calculations, such as f is,lowest .
(5)
approach to the one given in BS EN 13791 for the
determination of characteristic in-situ compressive Using Table 3 containing 39 valid test results;
strength by statistical use of the well-established and
accepted t-statistic to determine the characteristic n = 39
(6)
strength. t 0.05 = 1.69 (Table 6 of BS 6089)
f ck = 40
The in-situ characteristic strength f ck,is is calculated f is, lowest = 34.0
from; f m(39),is = 37.4
s = 1.518
f ck,is = f m(n),is − (t 0.05 x s)
Hence:
where t 0.05 value is based on (n − 1) degrees of f ck,is = 37.4 – (1.69 x 1.518) = 34.8
(6)
freedom (v), taken from Table 6 of BS 6089 , and s (f is,lowest + 4) = 34 + 4 = 38.0
is the sample standard deviation.
As f ck,is is less than (f is,lowest + 4), the in-situ
If the calculated characteristic in-situ strength result is characteristic strength of the area tested is 34.8MPa.
based on a large number of core data, it is
appropriate to use this value for structural If f ck,is were to be greater than 38.0, the in-situ
calculations where the in-situ characteristic strength characteristic strength of the area tested would then
(f ck,is ) is less than (f is,lowest + 4). be the lower of the two results i.e. 38.0MPa

If the in-situ characteristic strength (f ck,is ) is greater

than (f is,lowest + 4), then the lower value should be
used.
Page 8 of 8 CONCRETE ADVICE NO. 47

References
1. BSI. BS EN 12504. Testing concrete in structures. BSI, London
Part 1 Cored specimens - taking, examining and testing in compression (2009)
Part 2 Non-destructive testing. Determination of rebound number (2012).
Part 3 Determination of pull out force (2005). Part 4 Determination of ultrasonic pulse velocity (2004).
2. BSI. BS 1881-120. Testing concrete. Method for determination of the compressive strength of concrete cores,
BSI, London, 1983
3. BSI. BS 6089. Guide to assessment of concrete strength in existing structures, BSI, London, 1981
4. THE CONCRETE SOCIETY. Concrete core testing for strength, Technical Report 11, The Concrete Society,
Camberley, 1987
5. BSI. BS EN 13791: 2007, Assessment of compressive strength in structures and precast concrete component,
BSI, London
6. BSI. BS 6089:2010, Assessment of in-situ compressive strength in structures and precast concrete components
– complementary guidance to BS EN 13791: 2004, BSI, London
7. BSI. BS EN 1992-1-1: 2004, Eurocode 2. Design of concrete structures. General rules and rules for buildings,
BSI, London
8. THE CONCRETE SOCIETY. In-situ concrete strength – an investigation into the relationship between core
strength and standard cube strength, Project Report 3, The Concrete Society, Camberley 2004
9. BS EN 206: 2013. Concrete. Specification, performance, production and conformity, BSI, London

Impartial advice can be sought from The Concrete Society. Members are entitled to Contacts
substantial discounts on services and products including site visits and investigations,
dependent on status. For publications and information, The Concrete Society Bookshop The Concrete Society 01276 607140
holds a wide range of books and pamphlets along with an extensive library stock. We Bookshop 07004 607777
provide many services such as literature searches and notification of new references to
our extensive catalogue. www.concrete.org.uk www.concretebookshop.com
Issued April 2013
Reviewed Oct 2016

CONCRETE Advice Sheets are produced and published by The Concrete Society. The information and advice
contained in the Advice Sheets is based on the experience and knowledge of the Concrete Society’s Technical
Staff. Although The Society does its best to ensure that any advice, recommendation or information it gives is
accurate, no liability or responsibility of any kind (including liability for negligence), howsoever and from
whatsoever cause arising, is accepted in this respect by The Concrete Society, its servants or agents. Readers
should also note that all Concrete Society publications are subject to revision from time to time and should
therefore ensure that they are in possession of the latest version.