h i g h l i g h t s
Concrete type directly affects the response of the repeatability of the rebound index.
Ultra-high strength concrete has the lowest coefficient of variation for rebound hammer test.
Main concrete parameters influencing the response of the Schmidt hammer are examined.
Carbonation dramatically influences the repeatability of the rebound index.
EN 13791 basic curve underestimates the mean compressive strength values.
a r t i c l e i n f o a b s t r a c t
Article history: Diagnostic of concrete structure properties is essential towards assessing the properties of the evaluated
Received 28 July 2016 element. Since the Schmidt hammer is categorized as an economic and effective non-destructive testing
Received in revised form 10 October 2016 tool, an extensive investigation is applied on several hundred of concrete specimens produced from sev-
Accepted 16 November 2016
eral types, in order to assess the compressive strength and understand the limitations and boundaries of
rebound hardness. From normally vibrated concrete to ultra-high strength concrete, water-binder ratio,
water-powder ratio, supplementary cementitious materials and admixtures are relatively affecting the
Keywords:
response of rebound index of the Schmidt hammer in terms of repeatability for compressive strength pre-
Concrete
Compressive strength
diction. Additional porosity measurements were performed to enhance this statement.
Porosity Ó 2016 Elsevier Ltd. All rights reserved.
Rebound index
Repeatability
http://dx.doi.org/10.1016/j.conbuildmat.2016.11.085
0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
318 A. El Mir, S.G. Nehme / Construction and Building Materials 131 (2017) 317–326
aggregate with maximum diameter size of 5 mm. The grading frac- Table 2
tions used are of quartz sand ranging, between 0.2/5 mm. Materials applied for producing the selective concrete type.
Limestone and perlite powder were the fillers used to produce UHSC HPSCC NVC
HPSCC. Grinded fine quartz sand was supplied for the UHSC. The Cement type CEM II A-S 42.5 R CEM III A-N 32.5 CEM III A-N 32.5
term ‘‘powder” is defined by the sum of cement, SCMs and fillers. MSR MSR
According to EN 934-2:2009+A1:2012 Standard [18], high range CEM III A-R 32.5 CEM III A-R 32.5
water reducing admixture and air entraining agent represent the MSR MSR
SCMs MK, SF MK, SF MK, SF
added admixtures in several dosages. Workability tests were per- Filler type Quartz sand Limestone, perlite –
formed according to concrete type and its relevant standard Admixture Sika ViscoCrete 5 Sika ViscoCrete 5, Sika ViscoCrete 5
[19,20]. type Sika Aer
Measurements were carried out at several time offsets: 2-7-56-
90-240-400 days. Specimens were tested under saturated surface
dry conditions at 2 and 7 days. However, for all other time offsets, qb and qp are the bulk and particle densities (g/cm3) respectively.
the specimens were under air dried atmospheric laboratory condi- Water absorption by immersion test was applied until the spec-
tions. All specimens were demolded after 20–24 h of mixing and imens were fully saturated so that it could be checked by no mass
cured under water for 7 days and afterwards under laboratory variation. It was followed by being oven dried for 24 h under
atmospheric conditions. 100 °C. Relative masses were recorded in order to get the water
Depending on the classification and concrete type, w/p and w/b content in V% which corresponds to the atmospheric water satu-
ratios were selected to achieve the desired rheological and hard- rated condition designated by the apparent porosity. Cylindrical
ened properties. specimens with relative dimensions of Ø100 200 mm were used
A total of 104 concrete mixtures data from earlier publications in this evaluation.
[14,21–23], student’s diploma and current research work were col- Carbonation depth (xc) was computed by spraying
lected together and arranged as the following: 12 NVC, 81 HPSCC phenolphthalein-alcohol solution onto the sliced surface of
and 11 UHSC (Tables 1 and 2 summarize the test parameters with 150 mm cubic specimen. The carbonation depth which corre-
their relative materials). sponds to the unchanged color thickness layer was measured at
the center of each exposed surface and the average value was
3.2. Test methods recorded.
Fig. 3. Experimental rebound index – compressive strength results (795 test specimens) with respect to the basic curve given in EN 13791.
An extensive analysis was accomplished on the repeatability of recommendation (Fig. 2) should be reconsidered since no clear
795 data measurements collecting mean rebound indices and com- tendency is observed in the COV over the mean rebound index
pressive strength values. values in comparison with Fig. 4b. The latter clearly indicates
Since there are various concrete types and classifications, sta- the decreasing pattern of COV over the increase of mean rebound
tistical characteristics of the rebound index are essential to under- index. As the evaluated data is not distributed equally concerning
stand the response of the Schmidt hammer to the tested surface. mixture types, testing dates and admixtures, a correlation
Figs. 4a and 4b represent the COV of all experimental mixtures between concrete types and COV cannot be achieved. The paper
with regard to their mean rebound indices and compressive aims to have a generalized view on the behavior of rebound index
strength values. It can be observed the decreasing tendency of and compressive strength over the COV. Moreover, in the follow-
COV with the increase of the mean values of the compressive ing section, concrete mixtures are arranged based on the evalu-
strength, which confirms the tendency in the behavior illustrated ated property, enabling more objective interpretations that
earlier in Fig. 1. Despite that fact, ACI 228.1R-03 Committee would be reached.
Fig. 4a. Coefficient of variation (VR) as a function of the mean rebound index (Rm).
A. El Mir, S.G. Nehme / Construction and Building Materials 131 (2017) 317–326 321
Fig. 4b. Coefficient of variation (VR) as a function of the mean compressive strength (fcm).
4.2. Parameters affecting the repeatability of the Schmidt hammer of 90 days. Differences seem to become more balanced over the
age of 90 days. From a practical point of view, as the studied con-
4.2.1. Concrete type crete types have different rheological properties in terms of
Analysis of batch to batch variation was not the scope of this deformability and viscosity, authors claim that rebound index
study. The goal was rather to evaluate the evolution of COV of con- COV is directly affected by the placement method of concrete.
crete types over time. Therefore, all experimental data was
arranged regarding concrete types. Fig. 5 shows a graphical 4.2.2. Powder content
demonstration of mean within test variation of rebound index val- A comparative analysis was performed on several w/p ratios in
ues for NVC, HPSCC and UHSC over time. The age of tested con- order to evaluate the effect of their variation over the COV of the
cretes varied between 2 day and 400 days. rebound index. The following w/p ratios were selected from the
Referring to mean rebound index, UHSC and HPSCC resulted in experimental study: w/p = 0.17–0.29–0.31–0.34–0.5. As it can be
lower COV over the testing period, compared to references mix- observed in Fig. 6, it was experimentally proved by the COV values
tures (NVC). Experiments showed that the COV of the rebound that the rebound index becomes less sensitive with the decrease of
index is very sensitive to the tested concrete type before the age w/p ratio. In the first 28 days, almost all cases showed a rapid
Fig. 5. Effect of concrete types on the coefficient of variation of the rebound index in time.
322 A. El Mir, S.G. Nehme / Construction and Building Materials 131 (2017) 317–326
Fig. 6. Effect the w/p ratio on the coefficient of variation of the rebound index in time.
decrease in the rebound index COV, which can be explained by the mixtures, the latter showed higher rebound index COV
hydration reaction. Yet over 90 days, the rate of hydration becomes values despite that they hold the same maximum aggregate size
lower and it has a noticeable influence on rebound index COV val- (Dmax = 16 mm). This behavior is deduced by the difference
ues. This behavior is explained by the increase of powder content powder content between NVC and HPSCC types. HPSCC contains
in the concrete matrix reaching a more homogeneous material in more powder content, thereby enhancing the concrete
term of particle size and distribution. microstructure.
Also it should be stated out the effect of Dmax on the response of
rebound index COV. As it can be noticed in Figs. 6 and 7, lowest 4.2.3. Binder content
rebound index COV values are observed for UHSC mixtures with One of the most influential parameters on concrete strength is
relative w/p = 0.17 and Dmax = 5 mm. In case of HPSCC and NVC the w/b ratio [29]. The change of the rebound index response is
Fig. 7. Effect the w/b ratio on the coefficient of variation (VR) of the rebound index in time.
A. El Mir, S.G. Nehme / Construction and Building Materials 131 (2017) 317–326 323
Fig. 9. Effect of air entrained admixture on the coefficient of variation of the rebound index in time.
324 A. El Mir, S.G. Nehme / Construction and Building Materials 131 (2017) 317–326
Fig. 10. Coefficient of variation of the rebound index as a function of the mean total porosity provided by different w/p ratios.
Fig. 11. Coefficient of variation of the rebound index as a function of the mean apparent porosity provided by different w/p ratios.
Fig. 12. Coefficient of variation of the rebound index as a function of the mean carbonation depth.
A. El Mir, S.G. Nehme / Construction and Building Materials 131 (2017) 317–326 325
[19] EFNARC, The European Guidelines for Self-Compacting Concrete, Eur. Guidel. [27] BSI, ISO/DIS 17892-3, Geotechnical Investigation and Testing – Laboratory
Self Compact. Concr. http://www.efnarc.org/pdf/SCCGuidelinesMay2005. Testing of Soil. Part 3: Determination of Particle Size Density, 44, 2014.
pdf2005. 63. [28] CEN EN 13791:2007, Assessment of In-situ Compressive Strength in Structures
[20] MSZ-EN12350-5:2009, Testing Fresh Concrete – Part 5: Flow Table Test, Hun. and Precast Concrete Components, Eu. Stand., 2007, p. 30.
Stand., 2009. [29] L. Mlinárik, K. Kopecskó, A. Borosnyói, Properties of cement mortars in fresh
[21] A. El Mir, S. Nehme, Accelerated carbonation effect on rebound index of SCC, and hardened condition influenced by combined application of SCMs, EPA J.
CCC Congr., 2015, pp. 4–7, http://dx.doi.org/10.13140/RG.2.1.1677.3841. Silicon Based Compos. Mater. 68 (2016) 62–66, http://dx.doi.org/10.14382/
[22] A. El Mir, S. Nehme, Application of non-destructive test on self-compacting epitoanyag-jsbcm.2016.11.
concrete, IABSE Symp. Report, Int. Assoc. Bridg. Struct. Eng., vol. 105, 2015, [30] A. Borosnyói, Long term durability performance and mechanical properties of
pp. 1–8. <http://www.ingentaconnect.com/contentone/iabse/report/2015/ high performance concretes with combined use of supplementary cementing
00000105/00000018/art00003>. materials, Constr. Build. Mater. 112 (2016) 307–324, http://dx.doi.org/
[23] S. Zoltán, Nagy és ultra nagy szilárdságú betonok tulajdonságai, 10.1016/j.conbuildmat.2016.02.224.
Diplomamunka. Diplom work n hungarian, 2016. [31] B. Łaźniewska-Piekarczyk, Effect of viscosity type modifying admixture on
[24] BS-EN12504-2:2012, Testing Concrete in Structures – Part 2: Non-destructive porosity, compressive strength and water penetration of high performance
Testing—Determination of Rebound Number, Br. Stand., 2012. self-compacting concrete, Constr. Build. Mater. 48 (2013) 1035–1044, http://
[25] BS-EN12390-3:2009, Testing Hardened Concrete – Part 3: Compressive dx.doi.org/10.1016/j.conbuildmat.2013.07.076.
Strength of Test Specimens, Br. Stand., 2009, pp. 420–457. [32] J.K. Kim, C.Y. Kim, S.T. Yi, Y. Lee, Effect of carbonation on the rebound number
[26] ASTM International, ASTM C 642-97 Standard Test Method for Density, and compressive strength of concrete, Cem. Concr. Compos. 31 (2009) 139–
Absorption, and Voids in Hardened Concrete, Annu. B. ASTM Stand., 1997, pp. 144, http://dx.doi.org/10.1016/j.cemconcomp.2008.10.001.
1–3.