Electrical resistivity for non-destructive durability assessment of stabilized soils under drying-wetting cycles

1 Introduction

Ground improvement through chemical stabilization has been widely adopted in geotechnical engineering practice to enhance the mechanical properties of weak or problematic soils. Among the various challenges associated with stabilized soils, long-term durability under environmental loading remains a critical concern. In particular, repeated drying-wetting (D-W) cycles driven by seasonal moisture fluctuations can progressively degrade the strength and structural integrity of stabilized ground (Cuisinier and Masrouri, 2020). Ensuring this durability is therefore essential for the safe and sustainable performance of infrastructure built on or with chemically stabilized materials.

Conventional durability assessment of stabilized soils relies predominantly on destructive mechanical testing. The unconfined compressive strength (UCS) test is the most widely used method, applied across a broad range of stabilizer types, including cement, lime, calcium sulfoaluminate (CSA) cement, and eco-friendly alternatives such as biopolymers (Cuisinier and Masrouri, 2020; Nabil et al., 2020; Soldo and Miletic, 2022; Rauf et al., 2025). Beyond UCS, researchers have incorporated additional destructive indicators to capture different aspects of performance degradation: the California Bearing Ratio (CBR) and mass loss have been used to evaluate load-bearing capacity and material erosion under repeated D-W cycles (Abbey et al., 2023), and resilient modulus has been adopted to characterize stiffness degradation under repeated loading (Xu and Huang, 2025). While these destructive tests enable direct durability characterization, they require a separate set of specimens at each measurement interval. This precludes continuous monitoring of the same specimen and limits the ability to capture the full progression of internal damage accumulation. These constraints have motivated increasing interest in non-destructive evaluation (NDE) techniques to track material degradation without damaging specimens.

Among the NDE methods explored in geotechnical and construction materials research, ultrasonic pulse velocity (UPV) has been applied to monitor the mechanical performance and curing progression of fiber-reinforced stabilized soils (Miturski et al., 2021). Shear wave velocity measurements using bender elements have also been employed to track stiffness changes in bio-cemented soils subjected to D-W cycling (Gowthaman et al., 2022). However, these wave-based methods primarily reflect bulk stiffness and respond mainly to changes in elastic modulus. They are therefore less sensitive to progressive microstructural changes—particularly pore structure evolution, pore connectivity, fluid saturation, and microcrack development—that drive strength degradation in chemically stabilized soils. In contrast, electrical resistivity (ER) is governed directly by the ionic conductivity of pore fluid and the connectivity of the pore network, making it inherently more sensitive to the moisture-driven microstructural changes that accumulate under repeated D-W cycling. X-ray computed tomography (CT) has emerged as a promising non-destructive tool capable of directly visualizing and quantifying microstructural changes such as porosity, pore connectivity, and tortuosity under D-W cycles (Pires et al., 2020). In this study, ER and X-ray CT are employed in a complementary manner. ER serves as a practical non-destructive indicator of durability degradation, while X-ray CT provides direct microstructural evidence to support the observed electrical and mechanical responses.

The application of ER in geotechnical and cementitious materials research has evolved considerably. Early studies established foundational correlations between ER and key soil properties such as moisture content, dry density, and plasticity index, demonstrating ER’s potential as a rapid screening tool for site characterization (Sangprasat et al., 2025). Quantitative correlations between ER and mechanical properties—including UCS and shear strength parameters—were then established for cement-treated soils, further demonstrating ER’s utility as a non-destructive strength proxy (Vincent et al., 2017). ER has also been applied to track property changes in unstabilized compacted soils under D-W cycles. Hu et al. (2019) demonstrated that both UCS and ER decrease concurrently with increasing cycle number and amplitude, identifying pore structure evolution as the key governing mechanism. Despite these advances, the application of ER as a non-destructive durability indicator for chemically stabilized soils under repeated D-W cycles remains unexplored, representing a gap in the existing literature.

To address this gap, this study aims to evaluate ER as a reliable, non-destructive durability indicator. The specific objectives are: (1) to track the progression of damage in stabilized soils under D-W cycles using concurrent ER and UCS measurements; (2) to develop a novel normalized framework using a Strength Index (SI) and Resistivity Index (RI) to create a unified ER-UCS correlation applicable across different binder types; and (3) to validate the observed deterioration patterns with microstructural evidence from X-ray CT analysis. To meet the objectives, we applied ER measurement with UCS test to two eco-friendly rapid-setting stabilizers: a cementitious-additive-based stabilizer (Mix A) and a calcium sulfoaluminate (CSA) cement-based stabilizer (Mix B). The durability of both stabilized soils and an unstabilized control specimen was evaluated under 12 D-W cycles, with ER and UCS measured at each cycle interval. X-ray CT imaging was also conducted to characterize pore structure evolution and provide microstructural evidence for the observed deterioration patterns.

2 Materials and methods

2.1 Materials and specimen preparation

The soil used in this study was a silty sand with 12.5% passing the No. 200 sieve (0.075 mm), classified as SM under the Unified Soil Classification System (USCS). Standard Proctor compaction tests (ASTM D698-12, 2021) yielded an optimum moisture content (OMC) of 12.5% and a maximum dry unit weight (ρ d − max) of 1.75 g/cm³. The soil characterization results are summarized in Table 1.

TABLE 1. Soil characterization.

Three soil mixtures were prepared for this study: an untreated control (U) and two chemically stabilized mixes (Mix A and Mix B), as shown in Table 2. All mixtures were compacted at the OMC of 12.5% and molded into cylindrical specimens with a diameter of 33 mm and a height of 71 mm. Mix A was prepared with Chemical Stabilizer A, a commercial rapid-setting eco-friendly cementitious stabilizer, and Mix B with Chemical Stabilizer B, a calcium sulfoaluminate (CSA)-based stabilizer, both at a dosage of 4% by dry weight of soil. Following compaction, all specimens were cured for 7 days at 20 °C and 50% relative humidity to allow hydration reactions to proceed prior to the initiation of D-W cycling.

TABLE 2. Soil mix design.

When exposed to repeated D-W cycles, the two stabilizers respond to moisture in distinct ways. For Mix A, repeated wetting can dissolve stabilization products and leach cementitious components from the soil-stabilizer matrix, thereby reducing inter-particle cohesion and compromising mechanical integrity. The subsequent drying phase induces internal shrinkage stresses that promote microcrack initiation and propagation. For Mix B, the rapid-setting nature of the CSA cement accelerates drying and facilitates quicker moisture migration to the surface. This intensifies internal stress development during the drying stage, contributing to a more pronounced reduction in stiffness under repeated moisture cycling (Bazarbekova et al., 2025).

2.2 Drying-wetting (D-W) cycle procedure

After 7 days curing, specimens were subjected to repeated D-W cycles to simulate seasonal moisture variations representative of field conditions. During the wetting stage, specimens were placed on a porous stone submerged in water and allowed to absorb moisture through capillary suction rather than full submersion. This approach better replicates field conditions in which moisture typically enters through infiltrating rainfall or a rising groundwater table. Untreated specimens were turned every 15 min over a total wetting duration of 1 h, while stabilized specimens were wetted for 5 h with a single turning at approximately 2.5 h. The wetting durations were determined based on preliminary measurements. The degree of saturation was monitored by measuring the mass of each specimen before wetting (dry state) and at designated time intervals during the wetting process. The degree of saturation (S r) was calculated using:where M w is the specimen mass at a given wetting time, M d is the dry mass prior to wetting, ρ w is the density of water (1.0 g/cm³), and V v is the total void volume of the specimen, determined from the known dry density and specific gravity of the soil. Under both protocols, the degree of saturation exceeded 90%, ensuring comparable moisture conditions across all mixtures. During the drying stage, all specimens were dried in a curing room at 40 °C for 24 h.

S r = M w − M d / ρ w × V v × 100

The untreated specimens sustained severe structural damage after three D-W cycles and were therefore limited to three cycles. Stabilized specimens (Mix A and Mix B) were subjected to up to twelve D-W cycles. To ensure consistent measurement conditions, UCS tests were conducted immediately after the drying stage of each cycle, while ER measurements were performed immediately after the wetting stage. Measuring ER at the end of the wetting stage minimized variability due to differences in moisture content and allowed the influence of microstructural damage on resistivity to be more clearly isolated.

2.3 Experimental measurements

UCS tests were performed to evaluate the extent of mechanical degradation in untreated and stabilized soils subjected to D-W cycling. Tests were conducted in accordance with the standard procedure (ASTM D2166/D2166M-24, 2024) at a constant strain rate of 0.5%/min using an Instron 6,800 Universal Testing Machine (Instron, MA, USA). A minimum of three replicate specimens were tested for each mixture at each cycle interval, and the average value was used for analysis.

ER measurements were conducted using a two-electrode method. This configuration was selected over the four-electrode method because it provides a simpler setup suited for small cylindrical specimens and captures the overall bulk electrical response over a wider measurement range. The four-electrode method offers higher sensitivity but is more appropriate for localized measurements in larger samples. Although a standard test method (ASTM G187-23, 2023) exists for two-electrode ER measurement, a customized setup was adopted in this study, as illustrated in Figure 1. This enabled ER measurements on the same Harvard miniature-sized cylindrical specimens used for UCS testing, ensuring identical material conditions between the two measurements and eliminating the influence of specimen shape and size. The ASTM G187-23 standard is designed for a soil box configuration that differs geometrically from the cylindrical specimens used here, making direct application of the standard setup impractical. Copper tape electrodes were attached to the top and bottom faces of each specimen, with an inter-electrode distance of 50 mm, and connected to an LCR meter. An alternating current (AC) signal was applied, and the resulting electrical resistance (R) was recorded. The frequency and voltage of the AC signal were determined from preliminary measurements across all mixtures; a frequency of 5 kHz and a voltage of 0.5 V were selected because these conditions produced stable measurements free from polarization effects. The electrical resistivity (ρ) was calculated from the measured resistance using:where A is the cross-sectional area of the specimen and L is the electrode distance (50 mm). For each mixture and D-W cycle, three specimens were tested and three measurements were taken per specimen, yielding nine measurements per condition; these were averaged and used.

ρ = R A / L

To complement the ER and UCS measurements, X-ray computed tomography (CT) imaging was conducted to characterize microstructural changes—particularly void evolution—in the stabilized specimens under specific D-W cycles. Scanning was performed using a Bruker Skyscan 1,275 system (Bruker, Belgium) at 100 kV and 100 mA with a copper (Cu) filter applied, at a resolution of 20 μ m. CT scans were acquired at cycles 3, 4, 7, and 8 for the stabilized specimens. The acquired images were processed and analyzed using Bruker’s CTAn software to quantify the total void volume within each specimen.

3 Results and discussions

3.1 Degradation of UCS and ER under D-W cycles

Figure 2 presents the evolution of UCS and ER as a function of D-W cycle number for all mixtures. Both properties exhibit a general decreasing trend with increasing cycle number across all mixtures. This confirms that repeated D-W cycling induces progressive deterioration in both the mechanical load-bearing capacity and the pore-fluid conduction pathways of the tested soils. The untreated soil data is presented only up to cycle 3 in Figure 2, as the specimens experienced severe macroscopic cracking and complete structural disintegration after the third cycle, rendering further testing infeasible. Notably, the stabilized specimens (Mix A and Mix B) showed little change or even a slight increase in both UCS and ER during the first three cycles. This early-cycle behavior seems to be attributed to the continued hydration of unreacted cementitious clinker particles during wetting. As hydration proceeds, the formation of additional hydration products fills the pore spaces and progressively densifies the matrix, temporarily enhancing both stiffness and electrical resistivity. This is consistent with observations that repeated wetting supplies moisture for ongoing hydration reactions, resulting in continued pore refinement and microstructural densification in cement-stabilized soils (Tawiah and Shi, 2025). The pronounced drop in both UCS and ER at cycle 4 might be due to the transition from hydration-dominated behavior to damage-dominated behavior. As repeated wetting and drying accumulate tensile and shrinkage stresses within the cemented matrix, the bonding capacity of the stabilization products is eventually exceeded (Zeng et al., 2024). This triggers irreversible microcrack initiation and propagation and pore enlargement, dramatically increasing electrical conduction pathways and resulting in a sharp decrease in both UCS and ER.

To better understand the cycle-to-cycle degradation behavior within each stabilized mixture, the results were examined individually, as shown in Figure 3. For the stabilized mixtures, the largest decremental decreases occurred at two specific cycle intervals: cycles 3 to 4 and 7 to 8. For Mix A, UCS and ER decreased from cycle 3 to cycle 4 by approximately 31% and 28%, respectively, and from cycle 7 to cycle 8 by approximately 20% and 22% (Figures 3A,B). For Mix B, the corresponding reductions from cycle 3 to cycle 4 were approximately 61% and 62% for UCS and ER, respectively, and from cycle 7 to cycle 8 were approximately 26% and 19% (Figures 3C,D). Although the exact magnitudes differ between mixtures, UCS and ER exhibited consistent degradation tendencies within each mixture. The greatest incremental decreases consistently occurred at the same D-W cycle intervals.

These results indicate that D-W cycling induces stages of significant deterioration rather than continuous degradation. Certain cycle intervals are disproportionately damaging to both the soil structure and electrical conduction pathways. The consistent co-occurrence of pronounced drops in UCS and ER at the same cycle intervals supports the hypothesis that resistivity is sensitive to the same damage events that govern mechanical deterioration. However, the differences in magnitude between ER and UCS responses suggest that material-specific calibration is necessary when using ER as a quantitative durability indicator.

3.2 Normalized strength and resistivity indices

To facilitate direct comparison of degradation behavior across mixtures with different initial property magnitudes, the measured UCS and ER values were normalized to dimensionless indices. The Strength Index (SI) and Resistivity Index (RI) were defined as the ratio of UCS or ER at a given cycle to the corresponding value at cycle 1, such that an index of 1.0 represents the condition at the first cycle and values below 1.0 indicate progressive performance loss. The evolution of SI and RI with D-W cycle number is presented in Figure 4. As shown in the figure, both SI and RI exhibit a progressive decreasing trend with increasing D-W cycle number, consistent with the overall degradation patterns observed in the absolute UCS and ER results.

The relationship between RI and SI was subsequently investigated by plotting SI against RI for all mixtures, as shown in Figure 5. Linear regressions were fitted individually to the untreated soil, Mix A, and Mix B, all yielding positive correlations. This confirms that decreases in ER are consistently accompanied by proportional decreases in UCS within each mixture. Although the slopes and intercepts differ slightly among mixtures, combining all data points reveals a single unified linear trend with a coefficient of determination R ² of approximately 0.91. This indicates that the proposed linear model accounts for the variability in the normalized strength response across different mixtures. A strong linear correlation between the normalized RI and SI across all mixtures demonstrates that the RI-SI relationship is statistically robust and applicable across different binder types, confirming ER as a reliable surrogate for strength.

3.3 Microstructural evidence from X-ray CT imaging

X-ray CT imaging was conducted to provide microstructural evidence for the stepwise deterioration patterns identified in the electrical resistivity results. Total pore volume (%) was quantified at cycles 3, 4, 7, and 8 for both stabilized mixtures, corresponding to the two cycle intervals at which the most pronounced drops in UCS and ER were observed. The results are presented in Figure 6.

For Mix A (Figure 6A), total pore volume increased from 1.560% at cycle three to 1.728% at cycle 4, and from 1.760% at cycle seven to 1.952% at cycle 8. For Mix B (Figure 6B), the corresponding increases were from 2.655% at cycle three to 3.011% at cycle 4, and from 3.072% at cycle seven% to 3.397% at cycle 8. In both mixtures, the incremental increases in pore volume were more pronounced at the transitions from cycle 3 to 4 and from cycle 7 to 8, directly corresponding to the cycle intervals at which the largest drops in UCS and ER were recorded. In both mixtures, consistent with the UCS and ER results, the change in pore volume between cycles 4 and 7 was substantially smaller than that observed at the cycle 3 to 4 and cycle 7 to 8 transitions.

These findings confirm that the substantial deterioration observed in the mechanical and electrical responses is accompanied by discrete expansions in void space within the specimen microstructure. The concurrent increases in pore volume, UCS reduction, and ER reduction at the same cycle intervals suggest that void expansion is a critical microstructure driver of both mechanical weakening and increased electrical conductivity pathways in the stabilized soils. Furthermore, Mix B consistently exhibited higher absolute pore volumes than Mix A across all measured cycles. This is consistent with the generally larger magnitude of UCS and ER reductions observed for Mix B, and may reflect differences in the pore structure developed by the CSA-based binder relative to the cementitious-additive-based stabilizer.

Therefore, the results from UCS testing, ER measurement, and X-ray CT imaging provide a mutually consistent and complementary picture of durability degradation in chemically stabilized soils under repeated D-W cycles. The integrated mechanical, electrical, and microstructural responses point to the general conclusion: deterioration under D-W cycling is not a smooth, continuous process, but rather occurs in discrete steps concentrated at specific cycle intervals. ER and UCS degradation are strongly coupled, supported by direct microstructural evidence of pore volume expansion. This demonstrates that ER measurement offers a reliable and practical non-destructive means of tracking durability loss in chemically stabilized soils. The unified SI–RI correlation established in this study, with R ² ≈ 0.91, further confirms that this relationship holds across different binder types, suggesting that the proposed framework has the potential for broader application in the non-destructive durability assessment of stabilized soils.

4 Concluding remarks

This study proposed an integrated non-destructive durability assessment framework for chemically stabilized soils under repeated drying-wetting (D-W) cycles. Electrical resistivity (ER) measurement was applied alongside unconfined compressive strength (UCS) testing and X-ray computed tomography (CT) imaging to two eco-friendly rapid-setting stabilizers—a cementitious-additive-based stabilizer (Mix A) and a calcium sulfoaluminate (CSA) cement-based stabilizer (Mix B).

• Both UCS and ER decreased progressively with increasing D-W cycle number across all mixtures, confirming that repeated moisture cycling induces deterioration in both the mechanical load-bearing capacity and the pore-fluid conduction pathways of chemically stabilized soils. The stabilized specimens exhibited little change or a slight increase during the first three cycles, attributed to the continued hydration of unreacted cementitious particles.
• Deterioration under D-W cycling occurred in discrete stages rather than continuously. The most pronounced drops in both UCS and ER were concentrated at cycles 3 to 4 and cycles 7 to 8 for both stabilized mixtures. The consistent co-occurrence of these drops at the same cycle intervals indicates that ER is sensitive to the same damage events that govern mechanical deterioration.
• X-ray CT imaging supported the observations from UCS and ER by revealing the discrete expansions in total pore volume at the same critical cycle intervals. This provides direct microstructural evidence that void expansion is a critical driver of both mechanical weakening and increased electrical conductivity pathways in the stabilized soils.
• Normalization of UCS and ER to dimensionless Strength Index (SI) and Resistivity Index (RI) eliminated the variability among different mixtures, enabling a more general comparison of degradation behavior across all mixtures. A strong positive linear correlation between SI and RI was established. The RI–SI relationship is statistically robust and applicable across different binder types, confirming ER as a reliable surrogate for strength.
• The results demonstrate that ER measurement offers a reliable, practical, and non-destructive means of tracking durability loss in chemically stabilized soils under D-W cycling. It has the potential to serve as a surrogate indicator for strength degradation, eliminating the need for destructive testing at every cycle interval.

Future studies should extend the present framework to a broader range of chemical stabilizers and soil types to evaluate and further refine the generality and accuracy of the normalized ER–UCS correlation. In addition, future research should explore microstructural variables beyond porosity that significantly influence electrical resistivity—such as pore connectivity, microcrack geometry, and ionic composition of pore fluid. Their correlations with variations in SI and RI should also be investigated. Such efforts would not only deepen the understanding of the combined effects of damage mechanisms on strength and resistivity but also contribute to the development of more accurate models for ER-based non-destructive monitoring of stabilized ground structures.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Fei Wang, Mississippi State University, United States

Reviewed by: Panagiotis G. Asteris, School of Pedagogical & Technological Education, Greece

Shihao Huang, Tarleton State University, United States