Hydro-mechanical degradation of strength in saline soils under moisture infiltration and leaching

2.1. Sampling strategy

Soil material was collected from active construction zones within the Dostlik district where saline sediments and shallow mineralized groundwater are widespread. Sampling depth corresponded to typical foundation levels (approximately 1.5-2.5 m). Two specimen categories were prepared: undisturbed monoliths for deformation and shear testing, and disturbed bulk samples for physical and chemical analyses. Immediately after extraction, specimens were sealed to preserve natural structure, moisture, and salt composition.

2.2. Initial laboratory characterization

Before water exposure, baseline parameters were determined using standard geotechnical procedures. The testing program included natural moisture content ($w$), bulk and dry density $\left(\rho ,{\rho }_d\right)$, particle density (${\rho }_s$), porosity ($n$) and void ratio ($e$), Atterberg limits $\left(w_L,w_P\right)$, plasticity index ($I_p$), shear strength parameters $\left(c,\phi \right)$, deformation modulus ($E$), relative settlement under pressures of 0.1-0.3 MPa, and initial yield pressure. The measured reference parameters are summarized in Table 1.

To improve reproducibility, the laboratory equipment and typical measurement accuracy are listed:

– Moisture content ($w$): drying oven + electronic balance (temperature ±2 °C; mass ±0.01 g).

– Density $\left(\rho ,{\rho }_d\right)$: density rings + electronic balance (volume ±1 cm³; mass ±0.01 g).

– Particle density (${\rho }_s$): pycnometer + electronic balance (mass ±0.01 g).

– Atterberg limits: Casagrande/cone device (water content ±0.5 %).

– Direct shear $\left(c,\phi \right)$: shear box apparatus (normal load ±1 %; displacement ±0.01 mm).

– Oedometer / compressibility ($E$): oedometer frame + dial gauge (vertical deformation ±0.01 mm).

– Filtration gradients ($J$): sealed filtration cell + head control (head ±1 mm; filtrate volume ±1 mL).

– Dry residue / TDS: gravimetric method (temperature ±2 °C; mass ±0.01 g).

– Ion composition: titration (burette reading ±0.05 mL).

– pH: pH meter (±0.01 pH).

Note: if specific instrument models were not recorded, standard laboratory equipment was used; the listed accuracies represent typical values for the applied methods.

2.3. Chemical analysis of soluble salts

Water extracts were prepared using controlled solid–liquid ratios. Total dissolved salts (dry residue) and ionic composition were determined by titration and gravimetric methods. Dry residue and ion contents are reported per mass of dry soil (mg/kg) based on the water-extract procedure. The quantitative chemical characteristics are presented in Table 2 and were considered the initial salinity state prior to hydro-chemical impact.

These values were considered the initial salinity state prior to hydraulic influence.

2.4. Controlled filtration experiment

To reproduce groundwater seepage, a vertical one-dimensional filtration scheme was applied [2]. Undisturbed samples were installed in a sealed filtration cell, and the side surfaces were isolated to prevent lateral flow. Distilled water was introduced gradually under hydraulic gradients $J=$10, 30, 60, and 100, where $J$ is the hydraulic gradient. Effluent was collected at each stage, and the salt concentration in the filtrate was determined analytically. Leaching continued until the difference in salt concentration between successive stages became negligible. For each soil type, $n=$3 replicate specimens were tested at each hydraulic gradient stage, and the reported values represent the mean values of the replicates. The reduction in soluble salt content was calculated as:

where $\mathrm{\Delta }S$ is the decrease in soluble salt content (mg/kg), $S_{initial}$ is the initial soluble salt content before leaching (mg/kg), and $S_{residual}$ is the residual soluble salt content after a given filtration stage (mg/kg).

2.5. Diffusion leaching

For low-permeability samples, salt migration under concentration-gradient conditions was examined without external hydraulic pressure. The specimens were immersed in distilled water, which was periodically renewed. The leaching degree was evaluated as:

where $\beta$ is the leaching degree (%), $S_{initial}$ is the initial soluble salt content (mg/kg), $S_t$ is the soluble salt content after exposure time $t$ (mg/kg), and $t$ is the exposure duration.

2.6. Mechanical testing after hydro-chemical impact

After each filtration or diffusion-leaching stage, the mechanical properties were re-determined by direct shear and compression testing. Strength degradation was quantified in terms of cohesion loss and friction-angle change as $\mathrm{\Delta }c=c_{initial}-c_{after}$ and $\mathrm{\Delta }\phi ={\phi }_{initial}-{\phi }_{after}$, where $c_{initial}$ and ${\phi }_{initial}$ are the initial strength parameters before hydro-chemical exposure, and $c_{after}$ and ${\phi }_{after}$ are the corresponding values after leaching.

2.7. Data processing and modeling

Experimental results were analyzed to identify relationships between salinity decrease, moisture increase, and variation of mechanical parameters. The obtained empirical relationships support the engineering interpretation of deformation growth and bearing-capacity reduction of saline loam and sandy loam soils under prolonged groundwater exposure.

3.1. Changes in soluble salt content during leaching

In this section, $D_{sol}$ denotes the soluble salt content expressed as a percentage of dry soil mass, $c$ denotes cohesion (kPa), $\phi$ denotes the internal friction angle (degrees), and $\beta$ denotes the leaching degree (%). Figs. 1-3 and Tables 3-8 summarize the variation of $D_{sol}$, $c$, and $\phi$ with increasing leaching degree.

The experimental results demonstrate a consistent reduction in soluble salt content ($D_{sol}$, %) with increasing leaching degree $\beta$ for all tested soil states (Tables 3 and 6). The decrease is monotonic and becomes more pronounced at higher $\beta$, indicating active salt removal under prolonged water exposure and filtration/diffusion mechanisms. This confirms that groundwater rise and recurrent wetting can substantially reduce the salinity-related cementing effects in foundation soils. Figs. 1-3 visualize the dependence of $D_{sol}$, cohesion $c$, and friction angle $\phi$ on the leaching degree $\beta$.

Fig. 1Soluble salt content Dsol (%) versus leaching degree β (%) for gypsum-bearing loam (Table 6)

Fig. 2Cohesion c (kPa) versus leaching degree β (%) for gypsum-bearing loam (Table 7)

3.2. Cohesion response to leaching

Cohesion ($c$, kPa) decreases with increasing $\beta$ for both datasets (Tables 4 and 7). For the gypsum-bearing loam series (Table 7), cohesion degradation is particularly strong at low initial plasticity: for example, at preliminary plasticity 8 %, c decreases from 25 kPa ($\beta =$ 0) to 1.5 kPa ($\beta =$ 100). This suggests that initial bonding and apparent cementation contributed by soluble salts/gypsum are progressively lost during leaching, leading to significant reduction in shear resistance. Similar leaching-driven strength degradation has been reported for saline and gypsum-enriched soils under soaking/leaching exposure [7-10].

3.3. Internal friction angle response is soil-dependent

The internal friction angle $\phi$ shows a soil-dependent trend. In the loam-series dataset (Table 5), $\phi$ increases moderately with $\beta$, which can be associated with particle rearrangement and increased interlocking after partial salt removal. In contrast, for gypsum-bearing loam under extended leaching (Table 8), $\phi$ decreases with increasing $\beta$ (e.g., at preliminary plasticity 8 %, $\phi$ decreases from 28° to 24° for $\beta =$ 0 → 100). This indicates that, in gypsum-bearing soils, progressive loss of cementation and fabric disturbance may dominate over interlocking effects, resulting in overall weakening of frictional resistance at high leaching degrees [7], [9], [10].

Fig. 3Variation of internal friction angle φ with leaching degree β: comparison of the loam-series dataset (Table 4) and the gypsum-bearing loam dataset (Table 7)

Fig. 3 highlights a soil-dependent $\phi$ response: $\phi$ increases moderately in the loam-series dataset (Table 5), while it decreases in gypsum-bearing loam under extended leaching (Table 8). This indicates competing mechanisms of particle interlocking and cementation loss during prolonged wetting.

Tables 6-8 present the salinity levels and property variations of gypsum-bearing loam at different leaching degrees. These tables summarize changes in soluble salt content ($D_{sol}$, %), cohesion ($c$, kPa), and friction angle ($\phi$, degrees) as functions of leaching degree ($\beta$, %) and preliminary plasticity. The results provide a basis for evaluating the influence of leaching on both mineralization and the mechanical behavior of gypsum-bearing loam. Tables 7-9 provide the full numerical dataset used to build Figs. 1-3 and to interpret the long-term leaching trend.