Application of conductive hydrogel reinforced by cellulose nanocrystals in monitoring crack process of concrete structure

Abstract

It was a challenge to monitor concrete structure crack under complex environmental action. To obtain conductive hydrogel material with higher stretchability, signal response sensitivity and stability to monitor concrete structure cracking, the conductive hydrogel reinforced by different content of cellulose nanocrystals, which is called polyacrylic acid-cellulose nanocrystals (PAA-CNC), were developed in this paper. The performance improvement of PAA-CNC was studied by scanning electron microscope, resistance, uniaxial tensile and cyclic tensile test. Finally, the concrete crack monitoring accuracy of PAA-CNC was verified by three-point bend loading test. The result showed that combining cellulose nanocrystals with hydroxyl in conductive hydrogels can form uniformly dispersed micelles and three-dimensional network structure, which can increase the ionic conductive path and connection strength between molecules. When cellulose nanocrystals content of hydrogel was 0.12%, the effective strain sensing range and sensitivity within the range reached the maximum. When the content of cellulose nanocrystals was 0.12, the effective strain sensing range and sensitivity of PAA-CNC will reach maximum value. Compared with other contents of cellulose nanocrystals, PAA-CNC_0.12 can produce a stable signal response when tested and quickly recover to the initial resistance after cyclic stretching. The crack width obtained by PAA-CNC_0.12 does not exceed 5% of that obtained by digital image correlation equipment.

Graphical abstract

Keywords

concrete structure / crack monitoring / conductive hydrogel / cellulose nanocrystals

1 Introduction

Concrete has been widely used in building construction and traffic engineering structure such as bridge [1], tunnel [2] and subway station [3]. During service period, concrete may appear different degrees of crack at any time due to the effect of surrounding environmental action, such as temperature change [4], humidity change [5], traffic or soil pressure load changes [6]. When the crack width of concrete exceeds critical value [7], the concrete structure will be damaged [8] or even collapse [9]. Therefore, it is necessary to use appropriate instruments to monitor the crack change of concrete structure, so as to promptly know the security status information of concrete structure [10].

At present, the technology for monitoring the developing of cracks in solid material is mainly divided into contact monitoring method and non-contact monitoring method. The mainstream contact monitoring method include fiber Bragg grating technique [11], optical fibers patch technique [12], optical frequency domain reflectometry technique [13]. The mainstream non-contact monitoring methods currently include digital image correlation (DIC) technology [14] and infrared image technology [15]. The principle of DIC technology is using image processing method to identify and match feature points in the picture [16], then calculate the displacement and strain of structure surface [17]. Due to the advantage of DIC technology in capturing global displacement information, many researchers have used it to monitor the crack development law of different kind of solid material such as anchored rock [18], steel fiber concrete [19] and polyolefin fiber reinforced concrete [20]. However, the monitoring accuracy of DIC image is greatly affected by environmental lighting and camera stability [21]. The principle of infrared image monitoring technology is converting the infrared radiation emitted by material into an image, and extract the temperature distribution of material by color changes in the image to reflect crack increment. Although the infrared image monitoring technology has high imaging efficiency, accurate post-processing analysis of the images is required. So many researchers have combined deep learning with infrared imaging monitoring technology for improvement. Shibano et al. [22] combined random forest algorithm and infrared image monitor technology, studied the distribution pattern of cracks on the surface of concrete bridge pillars. Chen et al. [23] proposed automatic segmentation method based on infrared-visible fusion and deep learning to improve the crack visibility under low light condition.

Although non-contact monitoring method can obtain accurate cracking information of rubber concrete, it is limited by the ability to detect cracks in concealed areas of the structure, and the contact monitoring method overcomes this drawback. Recently the widely used contact crack monitoring methods mainly include fiber optic sensing technology and acoustic emission monitoring technology. The fiber optic sensing technology is widely used in crack monitoring of concrete slab [24], pre-existing concrete pipe [25], reinforced concrete structure [26] and post-tensioned fiber reinforced concrete beams [27]. However, fiber optic sensing instruments are prone to breakage when monitoring large crack in structures and easily affected by temperature change [28]. The acoustic emission monitoring accuracy is high [29–32], but it is easily affected by external sound wave interference [33]. With the development of composite material technology, flexible sensing material has been widely used in the fields of human motion detection such as conductive hydrogel [34]. As conductive hydrogel combined the characteristics of hydrogel and conductive materials, so it can maintain a certain flexibility and conductivity at the same time.

To obtain conductive hydrogel material with higher stretchability, conductivity sensitivity and stability to monitor concrete cracking, the conductive hydrogel reinforced by different content of cellulose nanocrystals (PAA-CNC) were developed. The performance improvement of PAA-CNC was studied by electronic microscope scanning, resistance, uniaxial tensile and cyclic tensile test. Finally, the concrete crack monitoring accuracy of PAA-CNC was verified by three-point bend loading test.

2 Material composition of polyacrylic acid-cellulose nanocrystals

2.1 Raw material composition

The free radical polymerization reaction primarily utilizes radicals to continuously grow chains, ultimately forming polymers. It can be activated from monomer molecules under specific conditions using initiators and is characterized by fast reaction rates and minimal by-products. This study is mainly based on the free radical polymerization reaction of cellulose nanocrystals (CNC) to prepare PAA-CNC hydrogel, with corresponding parameters shown in Table 1. The raw materials used in the experiment mainly include CNC, acrylic acid (AA), stearyl methacrylate, sodium chloride (NaCl), ammonium persulfate (APS), sodium lauryl sulfonate (SDS), glycerol (GL) and deionized water.

2.2 Production steps

In this experiment, a certain amount of NaCl and the surfactant SDS are first added to a dispersion of CNC, followed by constant stirring at 60 °C for 30 min in a water bath to ensure thorough integration of the mixture. Subsequently, hydrophobic monomer octadecyl methacrylate (C₁₈M) is added, and the temperature is reduced from 60 to 35 °C to allow the C₁₈M to melt into a waxy state and disperse evenly in the mixture. Stirring continues for 3 h to ensure uniformity of components. Next, AA monomer and glycerol are added to the above mixture and stirring continues for another 2 h at the same temperature until a visually clear, homogeneous solution forms. Afterwards, a 0.1 g/mL solution of APS and GL is prepared as the initiator and added to the mixture, stirred for 10 min to ensure the initiator is completely dissolved and evenly distributed. Finally, the homogeneous precursor mixture is transferred to a mold and subjected to a 6 h constant temperature polymerization reaction at 60 °C in a water bath to form a smooth, uniformly textured PAA-CNC, as shown in Fig. 1.

2.3 Number of test groups

At the beginning of the test, the dispersion containing 12 wt.% cellulose nanocrystals are thoroughly stirred to ensure an even distribution of solid components. Subsequently, depending on the test condition, this dispersion is diluted to a specific concentration. The prepared hydrogels are named PAA-CNC_x, where “x” denotes the concentration of CNC in the dispersion. The hydrogel mix ratios for each condition are detailed in Table 2.

2.4 Microstructure of polyacrylic acid-cellulose nanocrystals

The micelle structure in PAA-CNC can also be observed under a scanning electron microscope, as shown in Fig. 2. The hydrogel is mainly characterized by a relatively uniform network pore structure, and its pores are approximately oval, and are relatively evenly dispersed inside the hydrogel. Under the influence of SDS, the hydrophobic monomer C₁₈M enters the micelle particles, forming large-diameter micelles that are uniformly dispersed, and then copolymerizes with AA at 60 °C through a free radical polymerization reaction. The resulting micelle structure serves as stable physical cross-linking points connecting the gel network, further enhancing the mechanical strength and stability of the hydrogel.

As known from Fig. 2, the network structure of PAA-CNC hydrogel allows ions like Na⁺ and Cl^- to move freely, thus exhibiting excellent electrical conductivity. The resistance of the hydrogel specimens under each condition is measured using a digital bridge (the instrument model is TH28330), applying a voltage of 1V until the measured values stabilize, at which point the resistance value is recorded as the specimen’s resistance size. Figure 3 is the energy dispersive spectroscopy of PAA-CNC, there were C, Na, Cl, O, S major elements in it. Energy spectrum analysis shows that most of the elements in prepared PAA-CNC hydrogel are uniformly distributed. This result indicate that the material has excellent compositional uniformity at the micrometer scale, which is typically a key sign of good process control and attainment of stable and reliable performance.

3 Stretchability, signal response sensitivity and stability of polyacrylic acid-cellulose nanocrystals

3.1 Stretchability analysis by uniaxial tensile test

As a strain sensor, the PAA-CNC hydrogel undergoes changes in resistance corresponding to changes in its length, thus allowing the determination of strain from the resistance changes. If the stretching length cannot cover large deformation and cracks of material, effective monitoring is not possible. Therefore, it is necessary to assess its tensile mechanical property.

The tensile properties of the PAA-CNC hydrogel are tested using a micro-controlled electronic universal testing machine. The initial spacing is set at 3 cm, and the hydrogel is naturally stretched without tension, then stretched at a rate of 100 mm/min until it breaks. The elongation at break for each condition is recorded, as shown in Fig. 4.

Figure 5 shows the relation curve between stress and strain under uniaxial tension, it can be seen that with the increase of cellulose nanocrystals content, the tensile strength of PAA-CNC increases nonlinearly, compared with materials polyvinyl alcohol/sodium alginate-chelation performance and cellulose nanocrystals/ carboxymethyl cellulose sodium/ polyvinyl alcohol [35–37]. The PAA-CNC has better advantage in simultaneously keep good strength and elongation. The reason why the stiffness of the blue line increases is that during uniaxial tension test, the distribution of cellulose nanocrystals inside some conductive hydrogel is not uniform, and the polymer chains inside the conductive hydrogel may undergo local dense accumulation, resulting in an increase in material stiffness during the loading phase. Figure 6 shows the elongation at break of PAA-CNC with different cellulose nanocrystals contents, when the cellulose nanocrystals increase to 0.12, the value of elongation at break increases by 40%, and the tensile strength increases by 40%. When the critical content exceeds 0.12, the value of elongation at break begins to decrease.

3.2 Signal response sensitivity analysis by conductivity test

When monitoring cracks of concrete, high-quality flexible strain sensors need to have a wide strain sensing range in order to capture the entire process from small cracks to significant cracking. At the same time, high sensitivity is also an essential characteristic. The initial cracking of rubber concrete is often below the millimeter level, and high sensitivity represents the excellent ability to monitor such small strains. In addition, in order to accurately reflect the size of cracks in actual monitoring, sensors must maintain good linear response throughout the entire strain range. This means that the resistance change of the sensor should be proportional to the strain change, so that the crack size can be characterized by the resistance change in the actual monitoring process. If the sensitivity of the sensor shows nonlinear changes at different strain levels, it may lead to misjudgment of the extent of crack propagation. The strain sensing range, linear response capability under different tensile strains are important indicators for evaluating the Signal response sensitivity of PAA-CNC, as shown in Fig. 7.

The relative resistance change during the testing process is calculated as follow:

(1)RRC=\mathrm{\Delta }R/R_0=\frac{R-R_0}{R_0}\times 100\mathrm{\%},

where R and R₀ represent the initial resistance of the hydrogel flexible sensor and the real-time resistance during the test, respectively, and are the resistance changes. The sensitivity of the sensor is generally represented by the gauge factor (GF), which is the sensitivity factor of a piezoresistive strain sensor.

The GF is generally used to represent the sensitivity of the sensor, ε is the strain, the sensitivity factor of the piezoresistive strain sensor is calculated as follow:

(2)GF=\frac{\mathrm{\Delta }R/R_0}{\epsilon }.

Before testing, ensure that the distance between the upper and lower clamps is consistent with the original length of the flexible sensor specimen. Then, set the program to stretch the clamp of the universal testing machine upwards at a constant rate of 100 mm/min until the elongation distance is three times the original length, that is, control the total strain to 300%. Record the real-time resistance of the test piece during the testing process.

Figure 8 shows the strain sensing range and linear response capability of the specimens under various operating conditions. As shown in the figure, with the continuous increase of strain, the overall resistance of each specimen shows a monotonic increasing trend. But when the strain increases to a certain extent, there is a situation where the signal response is unstable. When the strain increases to 192% and 244%, respectively, the resistance signals of specimens with CNCs content of 0.12 wt.% and 0.18 wt.% exhibit rapid shaking phenomenon. If the sensing range of each specimen is defined solely by the stability of signal transmission, the upper limits of the sensing range for the four operating conditions are 300%, 192%, 244%, and 300%, respectively.

Obviously, this definition method is not comprehensive, and its sensitivity and linear response capability should also be considered comprehensively. From Fig. 8 it can be seen that there are obvious inflection points in the monotonically increasing resistance signal, and after passing through these inflection points, the slope of the curve changes significantly. The curve can be divided into multiple approximately linearly increasing segments by the bending points, and each segment can be regarded as a linear response of the signal. The sensitivity of each segment can be calculated using Eq. (2).

In the actual process of crack monitoring, it is necessary to capture the entire process from the generation to significant expansion of cracks. Therefore, when evaluating the effective sensing range of each specimen, the maximum strain that can be achieved in the first linear response should be the effective strain sensing range of the specimen, starting from when the strain is 0. At the same time, the sensitivity of the sensing specimens within this range should also be considered to ensure the ability to monitor small cracks in the actual process. The results indicate that the upper limits of the effective strain sensing range for the four operating conditions are 80%, 119%, 154%, and 70%, respectively, with corresponding maximum sensitivities of 1.15, 1.94, 1.09, and 1.10, respectively. It can be seen that when the CNCs content is 0.12 wt.%, the effective strain sensing range and sensitivity within the range reach their maximum, which are 154% and 1.94%, respectively.

3.3 Signal response stability analysis by cyclic tensile test

Within the effective sensing range that can be achieved under various working conditions, applying different strains and cycling a small number of times for each strain can further evaluate whether the signal response of the sensor under different strains has discrimination, and study the changes in resistance during different degrees of stretching and release processes.

During the testing process, the program is set to uniformly stretch the clamp on the universal testing machine at a speed of 100 mm/min, causing the sensor specimens to release when they produce strains of 10%, 40%, and 70%, respectively. Each operating condition is cycled five times [38–40].

Figure 9 shows the relative resistance variation curves of sensors under different tensile strains for four different operating conditions. It can be seen that there are significant differences in the relative resistance changes of the sensor under different tensile strains, mainly manifested as an increase with the increase of strain. At the same time, after 10% and 40% cyclic stretching, the sensor resistance can almost recover to its initial resistance, and the resistance is basically the same every time it reaches the maximum stretching. The increase in resistance during this process is mainly due to the reversible changes in the pores and network structure of the hydrogel. The ion migration behavior is blocked during the stretching process, but it recovers during the release process. The final performance is that the overall resistance of the hydrogel is basically the same as before stretching. When subjected to 70% cyclic stretching, the resistance at each maximum stretch increased compared to the previous one, and after five stretches, the resistance did not return to the initial resistance, but increased to a certain extent. The reason is that the overall structure of the water gel is irreversibly destroyed during the process, reducing the ion migration channels, and cannot be recovered after the tensile release, so the resistance increases with each tensile performance. In addition, the hydrogel sensor test pieces were observed during the test, and it was found that the length of the other test pieces increased significantly after 15 times of tensile recovery exercise, except for the test pieces with 0.12 wt.% content.

To study the sensitivity of sensors under different strain rates, during the testing process, the program was set to allow the universal testing machine to perform tensile recovery movements on the clamp heads at 25, 50, 100, 200, and 400 mm/min, respectively. The tensile strain was set to 10% to simulate a small strain scenario, and each rate was cycled 5 times continuously in sequence.

Figure 10 shows the relative resistance change curve of hydrogel under four working conditions at different stretching rates. As shown in the figure, the time required to complete 5 cycles at different rates is inevitably different. As the stretching rate increases, the time required to complete all cycles decreases accordingly, resulting in a denser relative resistance change curve. But at different stretching rates, the resistance response of each working condition sensor is almost consistent with the overall cyclic motion process, and there is no signal lag phenomenon. However, the increasing tensile rate will produce gradually greater stress in the hydrogel sensor, which will lead to a certain increase in its relative resistance.

Figure 11 is relative resistance after cyclic tensile at different stage, the result shows that in the range of 10% tensile strain, the hydrogel flexible sensors under four working conditions all exhibit excellent extensibility and resistance recovery ability. Within the 40% tensile strain range, sensors with CNCs content of 0.12 wt.% and 0.18 wt.% exhibit good stability and repeatability. Within the 70% tensile strain range, the sensor performance is optimal with a CNCs content of 0.12 wt.%. Figure 12 is the relative resistance after cyclic tensile at different rate, the result shows that after cyclic tensile at different rates, the relative resistance changes compared with the initial resistance values under each working condition were 2.75%, 0.196%, 2.46%, and 12.7%, respectively, indicating that only the hydrogel sensor with 0.12 wt.% CNCs content in the four working conditions could basically recover to the initial resistance values.

4 Concrete cracking monitoring test by polyacrylic acid-cellulose nanocrystals

4.1 Raw materials and ratios for polyacrylic acid-cellulose nanocrystals

The size of the concrete specimens poured in this research institute is 100 mm × 100 mm × 400 mm, there are 4 groups of component, which are used to verify the accuracy of testing at different material under same load. The concrete mix proportions, as shown in Table 3, the raw materials for each part are the cement, natural river and sand used, the stone material is continuously graded limestone crushed stone with a particle size distribution of 5–31.5 mm, as shown in Fig. 13. The three-point bend loading test was according with the standard for test method of mechanical properties on ordinary concrete [41].

4.2 Test measurement point layout plan

At first, hydrogel specimens with a size of 10 cm × 1 cm were cut out from the sample containing 12 wt.% CNCs, as shown in Fig. 14, and fixed two strips of 5 mm wide copper tape as electrodes on both ends of the hydrogel test piece. Subsequently, it was directly glued onto the corresponding position on the surface of the specimen using 705 silicone rubber. After curing, it became transparent rubber and still maintained good elasticity and flexibility, this ensures that the stretching of the sensor is not affected during the testing process. Measure the resistance of sensors S1-S8 separately during the holding period. Then repeat the loading process until the rubber concrete specimen completely fractures. To obtain information on crack changes during the loading process of rubber concrete, DIC monitoring technology was used as a comparison of the measured data during the experiment. During the experiment, the camera position was fixed, the light intensity and temperature were maintained.

4.3 Comparative analysis of monitoring results

With the continuous accumulation of load, the crack width gradually increases and gradually expand upwards. When the load reached 6–7 kN, the cracks penetrate the entire rubber concrete. Figures 15 and 16 show the DIC displacement cloud map and stress cloud map of the first group of rubber concrete loaded to 4 and 7 mm, respectively, it can be seen that the crack direction is not vertically upward, and the turning point is located in the middle of the structure.

Figure 17 show the crack width variation curve of points S5 and S6 obtained through PAA-CNC and DIC monitoring technology under different loads, it can be seen that the variation trend of test curve obtained through PAA-CNC is basically consistent with the curve obtained through DIC, the variation of load does not affect the monitor accuracy of PAA-CNC. The crack width of concrete beam under three point bending is calculated by Extended finite element method (XFEM), as shown in Fig. 18, the calculated value is basically consistent with the test value in the early stage of loading, but there is a significant deviation in the softening stage. The reason is that adopting XFEM to simulation crack is not considered the random distribution characteristics of aggregates inside the concrete, resulting in differences between the cracking path and the test result. Figure 19 is the test error statistics between PAA-CNC and DIC under different cellulose nanocrystals content, the result shows the error of the PAA/CNC_0.12 is the minimum, which does not exceed 5%. In summary, it is reasonable and feasible to monitor the cracking process of concrete by PAA/CNC_0.12.

5 Conclusions

In this study, a new material called PAA-CNC is proposed to monitor cracking process of concrete structure, and the material is mainly made of conductive hydrogel and cellulose nanocrystals. The application characteristic of PAA-CNC is that it can maintain higher stretchability, signal response sensitivity and stability to monitor concrete cracking. The novelty of the current research is that we have selected the optimal ratio of conductive hydrogel reinforced by cellulose nanocrystals through resistance and mechanical tests, and successfully applied this component to monitoring concrete cracking. The relevant conclusions are as follows.

1) The tensile test show that when the content of CNC was increased to 0.12 wt.%, the critical tensile fracture length of PAA-CNC reached maximum. It also illustrates that CNC can form stable micelle structure and physical cross-linking points with the original conductive hydrogel, and effectively enhance the mechanical strength of the conductive hydrogel.

2) The addition of cellulose nanocrystals can greatly improve the conductivity of existing hydrogel, the reason is that mixing cellulose nanocrystals with hydrogel can form uniformly dispersed micelles and network structure, which can increase the ionic conductive path. The conductivity sensitivity test show that when the content of CNC increased to 0.12 wt.%, the effective strain sensing range and sensitivity within the range will reach maximum, which are 154% and 1.94%, respectively.

3) When the cellulose nanocrystals content of hydrogel is 0.12 wt.%, it can generate a stable signal response within multiple cyclic stretching. The resistance can quickly recover to the initial resistance after cyclic stretching at different rates, indicating that the recoverability and the reusability of the PAA-CNC_0.12 is excellent.

4) Test result shows that adopt PAA-CNC to monitor concrete structure crack can maintain higher accuracy and stability during load process when cellulose nanocrystals content is 0.12, the maximum error does not exceed 5%.

5) The PAA-CNC can be widely used in concrete crack monitoring projects for subway stations and subway tunnels. The reason is that the PAA-CNC can not only monitor the concrete cracking information between structure and soil contact surface (concealed side), but also maintain the accuracy and stability of monitoring data under the comprehensive influence of soil load changes and train vibration load.

6) Research has shown that PAA-CNC can effectively monitor cracks in concrete, and this material can be further improved in the future to apply in various fields [42,43] such as human motion and physiologic signal monitoring, medical diagnosis, multidimensional motion perception and control, electronic skin, and flexible interfaces.

References

[1] Ma Y , Yang Z , Liu Y , Liu J , Zhao W , Qiu R , Chen X , Qiao Y , Deng Y . Flexural behavior of rectangular concrete filled steel tubular (RCFST) composite truss bridge. Structures, 2025, 72: 108260

[2] Yu S , Jin H , Cao M . Study on corrosion characteristic of semi-ring steel plate for strengthening shield tunnel under DC stray current. Construction and Building Materials, 2022, 347: 128631

[3] Chen J , Xu C , EI Naggar H , Du X . Quantification of seismic performance index limits and evaluation of seismic fragility for a new rectangular prefabricated subway station structure. Tunnelling and Underground Space Technology, 2023, 138: 105183

[4] Shen D , Huang Q , Huang Y , Liu C , Kang J . Early-age cracking failure behavior of high strength concrete with temperature rising inhibitor under different temperature histories. Construction and Building Materials, 2024, 450: 138581

[5] Miao Y , Lu Z , Wang F , Wang H , Li Y , Lin J , Jiang J . Shrinkage cracking evolvement in concrete cured under low relative humidity and its relationship with mechanical development. Journal of Building Engineering, 2023, 72: 106670

[6] Li B , Wang X , Yang Y , Fang H , Du X , Wang N , Zhai K , Di D , Shi M . Long-term performance of concrete pipes under fatigue traffic loads. Engineering Failure Analysis, 2024, 165: 108778

[7] Wang H , Jia M , Yang X , Wang Y , Yu R C , Wu Z . Determination of fracture toughness of concrete based on actual critical crack length: Theoretical model and experimental validation. Engineering Fracture Mechanics, 2025, 318: 110966

[8] Gao C , Li Z , Li M , Zhang L , Ma H . Dynamic time-history analysis of reinforced concrete structures considering damage in the non-principal axis direction. Magazine of Concrete Research, 2025, 77(11–12): 646–666

[9] Nguyen V , Lim N , Tan K . Effect of torsion on the mobilisation of alternate load paths in double-span reinforced concrete edge beams subjected to progressive collapse. Structures, 2025, 77: 109106

[10] Zhang H , Hu W , Zheng X , Sun X D . Practical evaluation method for structural safety of steel-concrete composite beam cable-stayed bridges under bridge deck fire. KSCE Journal of Civil Engineering, 2025, 29(12): 100273

[11] Jo H , Park D , Kim K , Ahn T J . Real-time monitoring of concrete curing using fiber Bragg grating sensors: Strain and temperature measurement. Sensors and Actuators A: Physical, 2023, 362: 114650

[12] Liu S , Li Y , Zhang Z , Ma M , Qiu W , She L , Zhou H , Li K , Sun K . Monitoring the circumferential cracking of inner concrete core in prestressed concrete cylinder pipe during wire-wrapping using distributed optical fiber sensors. Tunnelling and Underground Space Technology, 2025, 157: 106330

[13] Zhao L , Tang F , You R , Li H N . A silicone rubber packaged distributed optical fiber sensing tape for strain-crack monitoring based on OFDR technique. Sensors and Actuators A: Physical, 2024, 372: 115383

[14] Wang S , Hu S , Yang J , Hu Y . The influence of crack on the fracture parameters of concrete based on digital image correlation technology. Magazine of Concrete Research, 2021, 73(22): 1167–1179

[15] Singla R , Sharma S , Siddique R . Monitoring of chemical-assisted healing of fine cracks in concrete using ultrasonics and infrared thermographic techniques. Journal of Building Engineering, 2025, 104: 112279

[16] Ren H , Qian Z , Huang W , Bo W , Chen T , Cao H . Evaluation of fracture behavior in asphalt concrete through the combination of semi-circular bending test and digital image correlation technology. Construction and Building Materials, 2024, 451: 138854

[17] Zhong Z , Wang F , Kong F , Chen Y . Study of fatigue damage behavior in off-axis CFRP composites using digital image correlation technology. Heliyon, 2024, 10(3): e25577

[18] Su H , Zhang L , Yu L , Jiang Y , Feng X , Huang X . Analysis of the fracture response of anchored rock under three-point bending from experimental and numerical perspectives. Theoretical and Applied Fracture Mechanics, 2024, 134: 104692

[19] Yu L , Liu Y , Liu Y G . Advanced monitoring of damage behavior and 3D visualization of fiber distribution in assessing crack resistance mechanisms in steel fiber-reinforced concrete. Journal of Building Engineering, 2024, 97: 110980

[20] Azandariani M , Vajdian M , Asghari K , Mehrabi S . Mechanical properties of polyolefin and polypropylene fibers-reinforced concrete—An experimental study. Composites Part C: Open Access, 2023, 12: 100410

[21] Zhang C , Lei M , Chen Y , Kuang B , Liu S , Ding Y , Fang Q , Li X , He W , Xie H . Automatic crack tip localization in enormous DIC images to in-situ characterize high-temperature fatigue crack growth behavior. International Journal of Fatigue, 2024, 185: 108364

[22] Shibano K , Morozova N , Shimamoto Y , Alver N , Suzuki T . Improvement of crack detectivity for noisy concrete surface by machine learning methods and infrared images. Case Studies in Construction Materials, 2024, 20: e02984

[23] Chen H , Zhang Z , Yin W , Zhou G , Wang L , Li Y , Zhao C , Wang C . Shape characterization and depth recognition of metal cracks based on laser infrared thermography and machine learning. Expert Systems with Applications, 2024, 238: 122083

[24] Buda-Ożóg L , Zięba J , Sieńkowska K , Nykiel D , Zuziak K , Sieńko R , Bednarski Ł . Distributed fibre optic sensing: Reinforcement yielding strains and crack detection in concrete slab during column failure simulation. Measurement, 2022, 195: 111192

[25] Wu P , Wu J , Dong Y , Wu L , Wu D . Pre-existing concrete pipe disconnection detection based on fiber-optic distributed acoustic sensing. Measurement, 2025, 254: 117965

[26] Gomes Y , Ple O , Lushnikova A , Saidi M . Cracks monitoring in reinforced concrete structure with optical fibers patch. Optics and Laser Technology, 2024, 179: 111333

[27] Tan X , Abu-Obeidah A , Bao Y , Nassif H , Nasreddine W . Measurement and visualization of strains and cracks in CFRP post-tensioned fiber reinforced concrete beams using distributed fiber optic sensors. Automation in Construction, 2021, 124: 103604

[28] Zhang X , Hua Z , Picklesimer C A , Chuirazzi W C , Sun C , McMurtrey M D , Rufner J . Integrating fiber optic sensors into metallic components for sensing in harsh environments. Optics and Laser Technology, 2024, 170: 110188

[29] Strantza M , Van Hemelrijck D , Guillaume P , Aggelis D G . Acoustic emission monitoring of crack propagation in additively manufactured and conventional titanium components. Mechanics Research Communications, 2017, 84: 8–13

[30] Diakhate M , Bastidas-Arteaga E , Moutou Pitti R , Schoefs F . Cluster analysis of acoustic emission activity within wood material: Towards a real-time monitoring of crack tip propagation. Engineering Fracture Mechanics, 2017, 180: 254–267

[31] Michalcová L , Ruzek R . Fatigue test of an integrally stiffened panel: Prediction and crack growth monitoring using acoustic emission. Procedia Structural Integrity, 2016, 2: 3049–3056

[32] Morscher G , Maxwell R . Monitoring tensile fatigue crack growth and fiber failure around a notch in laminate SIC/SIC composites utilizing acoustic emission, electrical resistance, and digital image correlation. Journal of the European Ceramic Society, 2019, 39(2–3): 229–239

[33] Liu S , Li J , Xu W . Active acoustic measurement of sound speed for quantifying underwater gaseous bubble emissions. Applied Acoustics, 2024, 224: 110132

[34] Wu W , Chang J , He Y , Guo Z , Wang S , Mao J . Phytic acid-based super antifreeze multifunctional conductive hydrogel for human motion monitoring and energy harvesting devices. Sustainable Materials and Technologies, 2024, 42: e01126

[35] Li S , Wu B , Wang S , Jiang M , Pan C , Dong Y , Xu W , Yu H , Tam K C . Multi-level high entropy-dissipative structure enables efficient self-decoupling of triple signals. Advanced Materials, 2025, 37(3): 2406054

[36] Dong Y , Gao Z , Mi Q , Tian Y , Zou F , Pan C , Tang D , Yu H Y . Highly sensitive and structure stable polyvinyl alcohol hydrogel sensor with tailored free water fraction and multiple networks by reinforcement of conductive nanocellulose. International Journal of Biological Macromolecules, 2024, 281: 136128

[37] Wang H , Li Z , Zuo M , Zeng X , Tang X , Sun Y , Lin L . Stretchable, freezing-tolerant conductive hydrogel for wearable electronics reinforced by cellulose nanocrystals toward multiple hydrogen bonding. Carbohydrate Polymers, 2022, 280: 119018

[38] Wang Z , Ma Z , Wang S , Pi M , Wang X , Li M , Lu H , Cui W , Ran R . Cellulose nanocrystal/phytic acid reinforced conductive hydrogels for antifreezing and antibacterial wearable sensors. Carbohydrate Polymers, 2022, 298: 120128

[39] Wang Y , Yao A , Dou B , Huang C , Yang L , Liang J , Lan J , Lin S . Self-healing, environmentally stable and adhesive hydrogel sensor with conductive cellulose nanocrystals for motion monitoring and character recognition. Carbohydrate Polymers, 2024, 332: 121932

[40] Zhang Y , Li J , Yu X , Han D , Xu Y . Micro-corrugated chiral nematic cellulose nanocrystal films integrated with ionic conductive hydrogels leads to flexible materials for multidirectional strain sensing applications. International Journal of Biological Macromolecules, 2025, 295: 139569

[41] GB/T50081. Standard for Test Method of Mechanical Properties on Ordinary Concrete. Beijing: China Architecture and Building Press, 2019

[42] Ge D , Mi Q , Gong R , Li S , Qin C , Dong Y , Yu H Y , Tam K C . Mass-producible 3D hair structure-editable silk-based electronic skin for multiscenario signal monitoring and emergency alarming system. Advanced Functional Materials, 2023, 33(46): 2305328

[43] Gao Z , Dong Y , Huang C , Hussain Abdalkarim S Y , Yu H Y , Tam K C . On-Demand plus and minus strategy to design conductive Nanocellulose: From low-dimensional structural materials to multi-dimensional smart sensors. Chemical Engineering Journal, 2025, 510: 161552

RIGHTS & PERMISSIONS

Higher Education Press