1. Introduction
Self-compacting concrete (SCC) is a new category of high performance concrete characterized by its ability to spread into place under its own weight without vibration, and to self-compact without bleeding and segregation [
1]. Currently, high-volume fly ash, as much as 50%, is used to replace the cement in the self-compacting concrete. The high-volume fly ash helps to achieve good slump flow of the SCC. Using fly ash in SCC is beneficial to improve workability and flowability, as the incorporation of continuously graded cementitious materials and fillers reduces interparticle friction [
2,
3]. Also, the application of fly ash improves rheological properties and reduces cracks in concrete due to the lower heat of hydration with lower amounts of Portland cement in the SCC mixing [
4]. The significant use of fly ash as a cement replacement for SCC production opens the possibility for more sustainable concrete. The interests of developing more environment-friendly concretes where cement consumption is reduced using fly ash as cement replacement materials justifies the further exploration of the role this mineral material plays in concrete performance. The mechanical properties, durability and fresh states of SCC with high-volume fly ash have been investigated by many researchers [
5,
6,
7]. It has been revealed experimentally by the co-authors that the strengths of SCC with high-volume fly ash were lower than that of concrete with pure Portland cement, especially at early age, as shown in
Figure 1 [
8]. This could be because of the dilution effect and low pozzolanic reaction [
6,
9,
10]. It has been well reported in the literatures that the strength of hardened concrete is mainly determined by the curing during the SCC hydration process [
7]. By using isothermal calorimetry, thermogravimetry (TGA), X-ray diffraction (XRD), scanning electron microscopy (SEM) techniques, and pore solution analysis, Weerdt et al. [
11] studied the effect of minor additions of limestone powder on the properties of fly ash blended cements. Weerdt et al. also investigated the interaction between limestone powder and fly ash in ternary composite cement [
12]. Berry et al. described the investigations of high-volume fly ash (HVFA)-Portland cement (PC) binders, the physical and chemical properties of which have been characterized up to 365 days of curing [
13]. Although there are plenty of studies about the mechanisms of hydration reactions. The monitoring of the effect of fly ash on hydration characterization in concrete at very early ages is still rather limited. Therefore, it is necessary to investigate the effect approach to monitor the characterization of mineral material on the hydration process.
Until now, many researchers have carried out research on early-age concrete hydration monitoring to reveal this fundamental process of concrete settling and hardening. The temperature measurement is a traditional method to monitor early-age concrete hydration [
14]. Optic fiber sensors, infrared thermography technology, and thermal couples were used to measure the temperature variations during the hydration process and the concrete early-age strength development can be estimated based on the measured temperature information [
15,
16]. Apart from the temperature measurement, wave propagation-based concrete hydration monitoring has also received attention [
17,
18]. Ultrasonic wave measurement is a popular method to reflect the change of concrete properties due to the high sensitivity of the ultrasonic testing method to microstructural changes in materials [
19]. Lead Zirconate Titanate (PZT) is widely used [
20,
21] to generate and detect the ultrasonic waves because of its strong piezoelectric effect [
22,
23]. Based on PZT, two major approaches, the electromechanical (E/M) impedance [
24] and the active sensing method [
25], were often used.
Impedance of PZT is very sensitive for the changes of concrete properties [
26,
27]. A re-usable PZT transducer setup for monitoring the initial hydration of concrete and structural health was developed in Reference [
28]. The influence of setting and initial hardening of concrete were investigated to detect the gradual bonding between a steel reinforcing bar and fresh concrete [
29]. With the help of fuzzy logic [
30], the strength development process of a high-strength concrete was monitored. An artificial neural network algorithm to estimate the early-age strength of concrete based on the EMI dynamic response of PZT sensor was proposed [
31]. A wireless monitoring system which combined with the EMI approach was proposed by Kim et al. [
32].
On the other hand, an active-sensing approach using a pair of piezoceramic smart aggregates (SAs) was first proposed by Song et al. [
33] for concrete structural health monitoring. Smart aggregates (SA) were also utilized to perform structural health monitoring of a reinforced concrete (RC) bridge column subjected to pseudo-dynamic loading by Kong et al. [
34]. Kong et al. also proposed a comparative study of the very early age cement hydration monitoring using compressive and shear mode smart aggregates using the active sensing method [
35]. A support vector machine (SVM) was used to classify the concrete strength in the research by Kim el al [
36]. Based on the velocity change, early-age concrete strength can also be evaluated [
37,
38,
39]. A swept sine wave and several constant frequency sine waves were produced in Reference [
40] to completely understand the hydration condition of the inhomogeneous, over-cluttering, and high-scattering characteristics of concrete.
However, the influence of high-volume fly ash on the very early-age hydration performance of SCC has not received enough attention in the existing literature. It is well known that concrete hydration at a very early age (0–24 h) plays a significantly important role in the entire hydration process and the concrete experiences a complicated chemical reaction from the liquid stage to the hardened stage after concrete casting. In the study by Zhu et al., piezoceramic bender elements were utilized to measure the velocities of the P-wave and S-wave to determine the hydration of concrete paste successfully during the first 6 h after casting [
41]. Until now, little research has been conducted on the use of continuous testing with piezoceramic transducers on hydration performance of SCC with high-volume fly ash.
Therefore, the authors applied an active-sensing approach using SAs based on piezoceramic transducers to monitor very early age hydration characteristics of SCC with different volumes of fly ash. A pair of SAs were embedded in the test specimens before concrete casting. In the hydration process, one SA which works as an actuator to emit the sweep sine signal (100–100k Hz), and the other one works as a sensor to detect the propagated wave.
The evolution of hydration development gives a complete view on the complicated chemical reaction between the binder (cement and active mineral material) and water, and allows the determination of the effects of fly ash on hydration kinetics. Three states during SCC hydration were found: (1) liquid; (2) transition; and (3) solid. The monitoring results were compared with the mechanical test of penetration resistance of SCC and mortar. By comparing the results from SAs and penetration resistance, the meaningfulness and feasibility of this test method based on SAs were assessed in the case of hydration characteristics of SCC with high-volume fly ash.
2. Smart Aggregate Based Active Sensing Approach
The design schematic of smart aggregate (SA) is shown in
Figure 2a. The core of the SA is a pair of bonded PZT patches that share common electrodes. The poling directions of the two PZT patches are against each other. In between the two PZT patches is a shared thin copper film electrode that is bonded to both patches using conductive epoxy. Wrapped outside of the two PZT patches is a thin copper film that serves as an electrode and as an electromagnetic shield. This thin copy film is bonded to the PZT patches using conductive epoxy. Lastly, these two PZT patches with electrodes are applied with a layer of nonconductive epoxy and are sandwiched by two mating marble blocks to provide the protection. The outer copper thin film provides an electromagnetic shielding effect to the PZT patches, and inner nonconductive epoxy provides waterproof. Finally, the marble blocks provide mechanical protection to the fragile PZT patches. The dimension of each PZT patch is 15 mm × 15 mm × 1 mm.
Figure 2b shows the photo of the smart aggregate that was used in the experiment. The height and diameter of the smart aggregates is 20 mm and 25 mm, respectively.
Owing to the piezoelectric properties of the PZT, smart aggregate can work both as an actuator and a sensor. In this research, the active sensing approach was utilized. The schematic of the active sensing approach is shown in
Figure 3.
There are two SAs in each SCC specimen. SA1 works as an actuator and SA2 works as a sensor. A sweep sine signal (100–100k Hz) is emitted by SA1. The period of the sweep sine signal in this research is 1 s. When the emitted sweep signal propagates to the SA2, the signal is detected by the sensor. As the hydration progresses, the received signal changes in both time and frequency domains. By analyzing the received signals, the hydration characterization of SCC can be achieved.
3. Wavelet Packet Energy Analysis
The hydration of SCC significantly affects the stress wave propagation in both time and frequency domains [
40]. The hydration characterization in the early age should be quantitatively evaluated. Wavelet packet energy analysis is extensively used in structural health monitoring [
42,
43]. By integrating the wavelet-based signal processing with an active sensing system, an online monitoring system for composite structures was established [
44]. Du et al. [
45] developed a wavelet packet-based damage index matrix to evaluate crack damages on a pipeline structure. The wavelet packet analysis was applied by Zhang et al. [
46] to monitor the looseness of the cuplock scaffolds connections.
In the process of early-age hydration monitoring, the received signal was recorded every half hour during the first 24 h. The received signal at
kth (
k = 0, 1, 2, …, 48) half hour can be expressed as
Hk. It should be notes that
k = 0 represents the beginning of hydration. By an
n-level wavelet packet energy decomposition, the received signal
Hk is decomposed to
n + 1 signal sets in different frequency ranges {
S1k}, {
S2k}, …, {
Snk}, {
Sn+1k}. Each signal sets can be expressed as
Xij k,
where
i = 1, 2, …,
n + 1,
j = 1, 2, …,
m, and
m is the number of samples in each set [
47]. The energy of each set of {
S1k}, {
S2k}, …, {
Snk}, {
Sn+1k} can be calculated by,
the energy of received signal at
kth half hour
Hk is defined as:
where
represents the energy of
Hk.
Based on the
, a hydration completion index (HCI) was proposed as following:
where
(
l =
k + 1 = 1, 2, 3, …, 48) is the hydration completion index at different hydration periods. Attentions should be paid to
k = 0, which means that it is the completion at the beginning of hydration. In order to show the result more clearly, hydration completion index (HCI) is normalized,
where
and
are the calculated hydration completion indices at the first (0.5 h) and the last (the 24th hour) received signal, respectively.
presents the normalized hydration completion index (NHCI) at
lth half hour (
l = 1, 2, 3, …, 48). Based on the proposed novel normalized hydration completion index (NHCI), the hydration characteristic during different hydration periods can be quantitatively and clearly revealed.