1 Introduction

The essential characteristic for reinforced concrete existence is the perfect adhesion between the concrete (C) and the reinforcement rebars (RR). However, the mechanism of adhesion between these two different materials results from association between the chemical, mechanical and friction adhesion [1]. According to Daoud et al. [2] the bond existent at reinforcement bars and the surrounding concrete is indicate as the fundamental factor for reinforced concrete performance, and comprises an important aspect that should be taken into account during the structural assessment. Nonetheless, the bond between the rebars and the concrete can be affected by a large number of factor, as the water/cement relation, aggregates characteristics and cure conditions.

A recent study on crack emergence in the zone of adhesion steel–concrete was presented by Ožbolt et al. [3] considering the effect of the different levels of corrosion using the 3D chemo-hygro-thermo-mechanical model. Basically this model considered the physical, electrochemical process with the concrete mechanical performance and also considered radial concrete resistance as linear elastic, and the adhesion zone as an isotopic area. In this analysis, finite element method was employed, and the numerical model was calibrated based on simplified experimental test. The results showed that chemo-hygro-thermo-mechanical properties presents a large influence in the crack emergence and consequently for bond resistance between the RR-C [2].

Studies related with the porosity influence on cement matrix and its relationship with the adhesion mechanisms between the cement paste, aggregates and reinforced rebars (in case of reinforced concrete) has been carried out from reinforced concrete advance. However, only in the 70 decade, the studies on RR-C bond characteristics were oriented aiming the obtainment of relations between the mechanical properties of the concrete and the cement. Now, the bond RR-C has been performed in terms of numerical modeling and the zone characteristics implemented aiming a most real model about the behavior of rebar-concrete surrounding zone [4–6].

According to NBR 6118 [7] the capacity of resistance, performance in service and durability are some of the minimum quality requirements that concrete structures must to attend. Additionally, if this parameter was measured in the early ages dramatic situations can be analyzed and decision could be adopted for safety maintenance in a short time.

One of the most important parameters considered during safety assessment is the compressive strength of the concrete. However, for compressive strength assessment a destructive method based on axial rupture of cubic or cylindrical samples is commonly employed (see [14, 15]). In fact, for this destructive method the concrete specimens should be casted with concrete and after a period determined the compressive strength should be determinate in laboratories through specific pressure machine. But sometimes the laboratories are not of easy accessed to owners and this can make difficult the quality control, and especially in emerging economies this fact contributes to owners treat this important step as a subjective factor for the building safety. This way, advances on the methods for quality control of buildings material are essentials for maintenance of the quality control and safety of the buildings, especially during it application in field.

According to Silva et al. [8] the use of compressive strength test is motivated by: (1) low complexity of test procedures; (2) easy interpretation of results; (3) satisfactory correlation of this parameter with others resistances, tensile strength, shear strength, bond strength armor-concrete. Nonetheless, some limitations in this experimental test may be identified, as: (1) the test is characterized by a displacement test imposed at support sides which does not create a uniform stress field imposed on the sides for a heterogeneous material such as concrete; (2) the influence of technique and materials used in the regularization or leveling the top at test results; and (3) not consideration of the bond RR-C.

In truth, the bond RR-C characterization is a relevant parameter for structural safety assessment, and should not be excluded from technological control. This way, the employment of pull out test should be considerate in order to improve more information about the reinforced concrete behavior [4] and, in addition can be used for provide forecast on compressive strength of concrete.

The pull out method is standardized by RILEM/CEB/FIP RC6 [9]. This proceeding consists in pull out one steel bar centered in a specimen placed on support plates of a testing machine. The traction force applied to one end of bar and enough to slip inside the concrete block the other end is measured. The test results can be expressed as the variation of slip measured at the free end of the bar versus pull out strength.

Recently, low cost assessment methods has been developed on the field of the liquid measurement and thermal monitoring [10–12]. Nevertheless, exist a gab on the techno-scientific field regarding to a low cost pull out test for concrete samples, with easy applicability and high results accuracy.

Based in the above mentioned reasons, the present works intents to provide a novel pull out methodology focused to obtainment of the compressive strength of concrete. In the following section the experimental campaign and the proposed methodology will be described, as well the results and the conclusion.

2 Experimental

2.1 Novel pull out methodology

For the proposed methodology application, a pull out test machine was built, as showed Figs. 1 and 2. The pull out machine consist in a metallic structure with a rectangular base (100 cm × 20 cm) linked with a hydraulic jack with maximum load capacity of 10 tones, analogical pressure gauge and manual loading. For the pull out test the concrete samples should be initially centralized and the loads applied (manual override of hydraulic jack) considering the intervals of 10 s aiming most accuracy of the data measurements.

Fig. 1
figure 1

Overview of machine of pull out test modified

Full size image
Fig. 2
figure 2

Overview of machine of pull out test modified model

Full size image

The work mechanism of the proposed pull out machine consists that the force applied at the bar should be measured although a manometer accomplished to pull out machine till failure of the adhesion RR-C.

Sequentially, the measured values obtained by pull out test should be converted to adhesive strength through Eq. 1, where \( f_{bd} \) = adhesive strength of calculation; \( R_{s} \) = force applied along the rebar length; \( \emptyset \) = steel bar diameter; \( l_{b} \) = anchorage length. Then, the characteristic tensile resistance can be obtained by Eq. 2, where \( \eta_{1} = 1.0 \) (without rib bar); \( \eta_{1} = 1.4 \) (carve bar); \( \eta_{1} = 2.25 \) (rib bar); \( \eta_{2} = 1.0 \) (good adherence); \( \eta_{2} = 0.7 \) (poor adherence); \( \eta_{3} = 1.0 \) (diameter bar lower than 3.2 mm); \( \eta_{3} = {{\left( {132 - \emptyset } \right)} \mathord{\left/ {\vphantom {{\left( {132 - \emptyset } \right)} {100}}} \right. \kern-0pt} {100}} \) (diameter bar greater than 3.2 mm), and finally the compressive strength of the concrete can be estimated by Eq. 3, where: \( f_{ctd} \) = characteristic tensile resistance; \( f_{ck} \) = compressive strength of concrete, and \( \gamma_{c} \) = specific mass of concrete [7].

$$ f_{bd} = {{R_{s} } \mathord{\left/ {\vphantom {{R_{s} } {\left( {\pi .\emptyset .l_{b} } \right)}}} \right. \kern-0pt} {\left( {\pi .\emptyset .l_{b} } \right)}} $$
(1)
$$ f_{bd} = \eta_{1} .\eta_{2} .\eta_{3} .f_{ctd} $$
(2)
$$ f_{ctd} = 0.21 \times \left( {f_{ck}^{2/3} /\gamma_{c} } \right) $$
(3)

2.2 Materials and concrete samples preparation

For the concrete mixtures, the following materials were used: a pozzolan cement, namely Portland cement (CPII-Z) with specific mass of 3.02 g/cm3 and specific surface area (Blaine) of 3200 cm2/g was employed. A fine sand was also used part of the aggregates, characterized by 1.510 kg/dm3 of unitary mass, specific mass of 2.67 g/cm3, fineness modulus of 2.40 and maximum nominal size of 2.40 mm. The coarse aggregate presented 1.651 kg/dm3 of unitary mass, specific mass of 2.63 g/cm3, fineness modulus of 0.85 and maximum nominal size of 19.00 mm.

For the concrete mixtures design the American Concrete Institute’s method (ACI) [13] was employed. Three different concrete mixtures were designed for each compressive strength class, C20 (20 MPa), C25 (25 MPa) and C40 (40 MPa) and then, 18 concrete samples were casted, 9 with reinforced rebars diameter of 10 mm (here titled as E10) and 9 with diameter of 16 mm (assigned as E16). Additionally, nine reference concrete samples (titled as E0) were casted according to [14]. In the total, 27 concrete samples were tested after 28 days submitted to cure by immersion.

The concrete mixtures presented a mortar index around 54 %, maximum slump of 75 mm and minimum slump of 25 mm, index of incorporated air around 2 %. The water/cement index changed between 0.54 (C20) to 0.40 (C40).

The concrete samples dedicated to the pull out test (E10 and E16) were casted by a 1.5 L plastic bottle, where reinforced rebars with 40 cm of length were passed by the top hole till crossing the plastic bottle symmetrically. Following, the plastic bottle was fill by concrete and after 24 h submitted to cure by immersion during 28 days. For the pull out test, the plastic bottle was not removed because does not present influence to test measures.

The E0 concrete samples were casted in metallic cylindrical shapes, with diameter of 10 cm and height of 20 cm. The recommendations described in [14] were adopted. After 24 h the metallic cylindrical shapes were removed and the E0 concrete samples were also submitted to cure by immersion till the experimental campaign.

2.3 Methodology

Initially, the E0 samples were submitted to compressive strength test according with NBR 5739 [15]. For this, a hydraulic pressure machine with load limit capacity of 1000 kN was used. Now, for E10 and E16 tests was necessary to fill with cotton the bottle cap to create a non-adherence zone in order to prevent the reaction of the backing plate that compresses the concrete, restricts the transverse deformation and causes the overestimation of the bond stress in concrete, mainly because the concrete remains confined in its plastic mold during the test. The samples E10 and E16 were carefully centralized in the pull out machine in order to avoid measure distortions. Following, the loading was manually applied till the maximum pull out stress be achieved and the tension measured in the manometer was noted. All the details of these procedures are shown in the Fig. 3.

Fig. 3
figure 3

Pull out testing with the concrete samples E10 and E16, and the details of the plastic bottle used as molds

Full size image

It is necessary to state that the proposal methodology is limited to reinforced rebars diameters equal or less than 16 mm, due to top hole of the plastic bottle support diameters till 16 mm.

3 Results

The compressive strength values obtained for the E0-C20 concrete samples testing were of 17.39, 21.03 and 22.41 MPa. Now, the compressive strength values were of 23.45, 26.01 and 26.72 MPa for E0-C25 and of 42.35, 45.68 and 47.32 MPa for E0-C40. The compressive strength average values for the E0 samples were of 20.27 MPa (C20), 25.39 MPa (C25) and 45.11 MPa (C40). The E0 results were plotted and used for comparison with the E10 and E16 samples results, as showed by Figs. 4 and 5. In general, the E0 samples presented compressive resistance values really near to the compressive strength design, also presented a low dispersion between each samples, considering the concrete classes C20, C25 and C40.

Fig. 4
figure 4

Compressive strength results of the E0 and E10

Full size image
Fig. 5
figure 5

Compressive strength results of the E0 and E16

Full size image

Concerning to pull out tests results, taking into account the R s values collected during the experimental campaign the \( f_{bd} \), \( f_{ctd} \) and \( f_{ck} \) values were obtained through expressions 1, 2 and 3, as previously mentioned. All the experimental results range are showed in the Table 1.

Table 1 Experimental results of the E10 and E16
Full size table

The compressive strength results of E0 and E10 samples are compared in the Fig. 4, where is possible to observe that the results present similar values really closed each other. Taking into account the C20 class, the standard variation of the compressive strength between E0 and E10 was of 1.80 MPa, and to classes C25 and C40 the standard variation values remained low, 1.52 and 2.50 MPa respectively. However, for 77.7 % of the concrete samples the compressive strength values of the E0 were superiors to E10 values, this same characteristic is also observed in the E16 results.

Concerning to E16 results, a comparison between them and the E0 compressive strength values also confirmed low standard variation values, namely 1.81 MPa for C20, 1.52 MPa for C25 and 2.50 MPa for C40.

4 Conclusions

In this work a novel methodology for obtainment of the compressive strength of concrete based on pull out test was presented, namely a low cost method. The results found showed that the compressive strength can be easily determinate by employment of the RILEM expressions and that the compressive strength values measured during the experimental campaign presents a linear relation with the values from the regular compressive strength test, standardized by NBR 5738 [15].

In fact, this method can be employed as an alternative test for determination of the compressive strength of concrete in field applications, and can be used as a support tool for concrete quality control.