**1. Introduction**

All engineering projects encounter a range of challenges associated with the most widely used building material, concrete. Being a major problem in current concrete construction, concrete cracking or damage requires a continuous search for new methods and improvement of the existing concrete assessment techniques. This is especially important for fresh concrete, which a ffects the behavior of concrete under load.

Due to the multilevel nature of concrete, with qualitatively distinct mechanisms taking place during the formation of the concrete, the interaction of various parameters must be considered and the ways to study these relationships and e ffects have to be found to detect damage. Concrete deterioration occurs primarily through technological cracks (microcracks) and di fferent interfacial properties (cracks) formed at various structural levels, which propagate and initiate operational cracks affecting the usability and strength of concrete elements. The composition of the concrete largely affects its properties. Concrete has high compressive strength and is durable. It can be formed into virtually any shape. Weak points of this material include low tensile strength, shrinkage during the hardening process, and susceptibility to external influences, such as moisture [1], temperature, chemical influences, etc. [2,3].

Particularly important for concrete elements is the early period accompanied by a number of phenomena related to cement hydration [4–6].

Chemical reactions occurring in the cement paste during the hydration process, drying out (water evaporation), the cement paste properties themselves (e.g., bleeding in fresh concrete and temperature changes), as well as volumetric changes due to external factors (temperature and air humidity) cause swelling and chemical, plastic, autogenous, and drying shrinkage. These volume changes of hardening concrete generate natural stress, including "micro" stress. Stresses occur most often in the interfacial transition zones (ITZ) between the grains of aggregate and cement paste and decide on the mechanical properties of these zones and their microcracking. At stress concentrations exceeding the tensile strength of concrete, the microcracks may propagate into the deeper layer of the cement paste or to the surface of the element. Examples of damages in the concrete elements shown in Figure 1.

**Figure 1.** Examples of damages in the concrete microstructure, adapted from [6,7].

To mitigate microcracking in concrete, an addition of fly ash, an application of the blast-furnace slag cement or low density aggregate is good practice, as demonstrated in [8]. However, in the case of normal-weight concrete, under the influence of destructive external factors, such as high temperature, frost, and loading, these microcracks can develop into cracks, thereby reducing structural durability and serviceability and in rare cases lead to failures, e.g., walls in tanks [9], concrete slabs [10,11], precast elements [12], or other structural elements [13].

Objective assessment of damage formation and development in concrete, which is independent of the components, additives, and external impacts is essential.

Various non-destructive methods have been used for this purpose [14–16]. Acoustic emission (AE) is the technique capable of detecting, classifying [17], and locating [18,19] damage in concrete. Traditional use of acoustic emission methods in the building industry includes the monitoring of damage [20,21] and crack development under load [22–26], cement setting and curing [27–32], or an assessment of the ASR (alkali–silica reaction) in concrete [33,34].

The research analyzing stress in concrete is especially related to the acoustic emission phenomenon. The acoustic emission method enables the determination of basic parameters of fracture mechanics necessary to analyze the course of stress affecting concrete destruction. Depending on the grade of concrete tested, the criteria for the estimation of the level of stress were established [35,36]. Another approach aimed at estimating the correlation between acoustic emission and stress in compressed concrete is the technique that relies on the Gutenberg–Richter (GBR) law [37]. It was observed that the event frequency in concrete samples during compression corresponds to about 70% of the maximum stress.

Progressive damage of structural elements (Reinforced Concrete beams—RC beams) under bending is assessed using the Keiser effect by monitoring AE activity during cyclic loading [38]. The Keiser e ffect is used to estimate the stress to which the structural element was previously exposed. To estimate the Kaiser e ffect (according to NDIS-2421 by JSNDI—the Japanese Society for Non-Destructive Inspection) two ratios are calculated: the load ratio and calm ratio. Their values are the basis for damage qualification as intermediate, minor, or heavy. Digital image correlation (DIC) techniques supporting the AE method are applied for providing information about the level of damage in the RC beams [39]. The relaxation ratio may also be a good indicator of damage status. During initial stages of loading the deflections increase slightly (loading phase). The element is in a serviceable state up to 50% of the deflection limit. In a higher deflection range (50–85%) the structural element is no longer serviceable. Deflection higher than 85% represents the failure of the element. The acoustic emission method is also used for the observation of the crack mouth opening displacement (CMOD). The di fferent nature of dissipated and emitted energy rates was observed in [38] during the loading process.

The methods performed on concrete subjected to compression and bending do not consider an influence of internal stress on concrete strength. At the initial stage of concrete setting, the cement paste shrinks and meets the resistance of aggregate grains that do not shrink. A self-balancing state of compressive and tensile stress arises. If the tensile stresses in the cement paste exceed the tensile strength, microdefects occur. These defects may form in the matrix and in the interfacial transition zone (ITZ) around the aggregate [4] (first destructive process). Internal microcracks interact with each other; they can join together in a damage network. This happens when the structure surrounding the internal microcracks in the cement paste is not able to transfer accumulated stresses. This is when the second destructive process arises. Furthermore, the heterogeneous increase in temperature in the cross-section of the element, as well as water evaporation from the surface layers causes the stretching in the outer zones and compression of the inner zones of the element. These non-stationary and non-linear temperature and humidity areas generate macrostress in the cross-section [36] that can lead to microcracks on the concrete surface (third destructive process) and then their propagation (fourth destructive process). These destructive processes result in discontinuities in the structure. Local structural defects initiate future destruction of the concrete and may reduce the strength of elements, causing their linear deformation and a ffecting serviceability functions [4,36,40].

There is no information about the assessment procedure of fresh concrete quality by acoustic emission before loading. In most cases analysis of non-loaded concrete is based on ring-down counting, which involves counting how many times the amplitude passes the fixed threshold or event-counting corresponding to number of AE waves recorded by a single sensor [41]. In these cases, acoustic emission signals the damage (crack formation) without being able to identify the underlying processes. Only some of the AE techniques, such as the methods described in [19,23,24], allow for effective identification and location of the destructive process.

The non-invasive acoustic emission method (modified IADP method—Identification of Active Destructive Processes method) presented in [42–44] has been shown to be suitable for investigating defect formation process at the early stage of hardening of young concrete.

The study presented in this paper demonstrates that this method is of a general nature and allows observation and identification of destruction processes regardless of the aggregate used, cement types, admixtures added, hardening conditions, temperature, or the presence of reinforcement. It also enables quantitative assessment of destructive processes, which can be important when assessing the strength properties of concrete.

The method can thus be applied to diagnosing elements made of reinforced concrete, controlling the concrete hardening stage, and supporting decision making (e.g., related to demolding), thereby ensuring the reliability of the structure.

### **2. Materials and Methods**

A total of 30 samples (ten concrete series, W2, W3, W4, W5, W6, W7, W8, B2, B3, and B4, of three samples each—A, B, and C) were tested.

Twenty-one samples (W2–W8) were made of C30/37 concrete, six samples (B2, B3) were made of C40/50 concrete and three (W2) of C25/30 concrete. Except for sample W2 (100 mm × 100 mm × 500 mm), all samples had square cross-sections with 150 mm on each side and the length of 600 mm. Samples B2 were made with chemical admixtures (plasticizer and air entraining agent), other samples without admixtures. Samples denoted by "W" were made with limestone aggregate from the Trzuskawica quarry, while these marked with "B" with basalt aggregate from Góra ˙zd ˙ze quarry. All samples were made with cement CEMI 42,5N—MSR/NA from the Warta cement plant (Cementownia Warta S.A., Tr˛ebaczew, Poland) (except B4—CEMIII/A 42,5N—LH/HSR/NA from the cement plant in Małogoszcz, (Cementownia Lafarge Małogoszcz, Poland). The chemical compositions of the cements are compiled in Table 1. Mixture proportions of samples W2–W8 and B2–B4 are listed in Table 2.





limestone aggregate, 2 basalt aggregate.

Three samples B4 were made with basalt aggregate, blast furnace slag, and cement CEM III without any additions.

The W3 and W4 samples after fabrication were cured in water for 10 days and then tested for 58 days under cyclic temperature variations (Figure 2). Additionally, steel reinforcement was embedded in the W4 samples (Figure 3a).

**Figure 2.** Temperature conditions during the first 14-days of test.

The samples in series W7 were tested for 58 days without water curing at a constant temperature (+22 ± 2 ◦C).

Before the test, the AE (Acoustic Emission) sensors were attached to one side of each sample (Figure 3b,c).

To provide appropriate conditions, the test stand was developed, comprising of a thermally and acoustically insulated chamber. A list of samples examined is shown in Table 3.

AE signals were recorded for 58 days in 12-h stages.

The proposed identification of active damage processes (IADP) method was presented in [19,21,24] and applied for damage identification and location in reinforced concrete beams under loading [22]. It relies on the study of AE signals produced by the process causing the deterioration of strength properties in structural elements. The results recorded in samples (AE signals) were compared with the reference signals obtained in the laboratory.

Then the modified version of this method was applied to detect damage in young concrete [23,42].

The IADP method outline is shown in Figure 4. This concept is based on the comparative analysis of waves generated by defects in concrete (detected by sensors) with a database of reference signals created earlier. Preamplifiers with a gain of 35 dB were used to amplify signals generated by defects. Then the signals were detected, transformed into electric signals, measured, recorded, analyzed, and assigned to the reference signals in the database using Noesis software and unsupervised learning methods.

**Figure 3.** (**a**) Schematic diagram of reinforcement of W4 samples; (**b**) sample during the test; and (**c**) acoustic emission (AE) sensor arrangemen<sup>t</sup> (unit: mm).

Figure 3 shows the AE sensors arrangemen<sup>t</sup> on the test sample. Two piezoelectric sensors with a gain of 25–80 kHz allow not only detection of destructive processes (AE source) but also finding their linear location.


**Table 3.** Sample parameters and testing conditions.

The preliminary reference signal database was developed based on 12 parameters of the AE signal: counts, counts to the peak, amplitude signal duration, signal rise time, signal amplitude, signal energy, signal strength average, effective voltage, absolute energy, average frequency, reverberation frequency, and initiation frequency. There are four destructive processes, described in [2–4,7], which may be a source of AE in freshly made concrete before loading. In [42,43,45] damage processes were ascribed to four signal classes recorded in non-loaded concrete (Table 4).

**Figure 4.** The concept of the method—IADP (Identification of Active Destructive Processes).


