Study on the mechanical strength and durability of polymer-based underwater non-dispersible cement pastes

2.1. Materials

Ordinary Portland cement (OPC) and calcium sulfoaluminate cement (CSA) are used as the primary materials for underwater non-dispersible paste. The OPC (P.O.42.5R) and CSA comply with the Chinese standards Common Portland Cement (GB175-2020) and Calcium Sulfoaluminate Cement (GB20472-2006), respectively. The chemical compositions of the cements, measured using a Panalytical Axios Fast X-ray fluorescence spectrometer, are listed in Table 1. Particle size distributions were analyzed via dry powder multi-point sampling with a Bettersize 2000 laser particle size analyzer (Fig. 1). For OPC, the frequency distribution diameters D10, D50, D90 were 2.987 μm, 14.50 μm, and 49.18 μm, respectively, with a specific surface area of 367.9 m2/kg. CSA exhibited D10, D50, D90 values of 2.75 μm, 11.10 μm, and 31.60 μm, respectively, and a specific surface area of 923.5 m2/kg. Cement particle morphologies at varying scales were observed using an FEI INSPECT F50 field-emission scanning electron microscope (Fig. 2).

The polyacrylamide (PAM) used has a molecular weight of 10 million Daltons, a density of 1.0±0.1 g/cm³, and appears as white to pale crystalline granules. A PCA®-Ⅰ polycarboxylate superplasticizer, produced by Jiangsu Sobute New Materials Co., Ltd., was employed. The anti-washout paste had a water-to-cement ratio of 0.5, with a superplasticizer dosage of 2.0 % by cement mass. For studies on setting time and compressive strength, PAM content ranged from 0.0 % to 1.4 %. During comparative analyses of OPC and CSA under underwater/land-cast conditions, the optimal PAM dosage was fixed at 0.6 % [7].

2.2. Setting time and compressive strength

The mixing process was carried out using a cement paste mixer. Prior to mixing, the mixing bowl and blades were wiped with a wet cloth. The mixing water was poured into the bowl, followed by the careful addition of cement within 5-10 seconds to minimize splashing. The bowl was then mounted into the mixer and raised to the mixing position. The mixer was started and operated at low speed for 120 seconds, followed by a 15-second pause during which any paste adhering to the blades or bowl wall was scraped into the center. Subsequently, high-speed mixing was conducted for another 120 seconds before stopping. The freshly cast specimens were cured in a controlled environment at 20 °C and 95 % relative humidity. For setting time measurements, parallel tests were performed and the average value was reported. For strength tests, triplicate parallel experiments were conducted, and the results were averaged. Data variability was indicated using error bars.

Fig. 1Particle size distribution of OPC and CSA cement

a) OPC

b) CSA

Fig. 2Particle size distribution of OPC and CSA cement

a) OPC

b) OPC

c) CSA

d) CSA

Setting times of PAM-modified pastes were determined per the Chinese standard Test Methods for Water Requirement of Normal Consistency, Setting Time, and Soundness of Cements (GB/T 1346-2011). Initial setting time was defined as when the Vicat needle penetration depth fell below 25.0 mm, and final setting time as when no visible penetration occurred. Uniaxial compressive strength tests followed Test Method for Cement Mortar Strength (GB/T 17671-2021), using 50.0 mm cubic specimens cured for 1 d, 3 d, 7 d, and 28 d.

2.3. Splitting tensile and bond splitting tensile performance

Splitting tensile and bond splitting tensile strengths were evaluated according to Test Code for Underwater Non-Dispersible Concrete (DL/T 5117-2000). Splitting tensile tests used 70.7 mm cubic specimens loaded at 0.5 N/mm². For bond tests, split specimens (70.7 mm×70.7 mm×35.35 mm) were surface-cleaned and recast under land (25 °C) or underwater conditions, then cured to specified ages.

2.4. Impermeability

Chloride resistance was assessed via the rapid chloride migration coefficient method (Standard for Test Methods of Long-Term Performance and Durability of Ordinary Concrete, GB/T 50082-2009), using a CABR-RCMP multifunctional tester (China Academy of Building Research). Testing parameters included adjustable DC voltage (0-60 V), current (0-10 A), 0.1 mol/L AgNO3spray, 10 % NaCl catholyte, and 0.3 mol/L NaOH anolyte. Specimens were cured for 28d and tested over 24.0 h.

3.1. Setting time and compressive strength

The setting time and compressive strength of cement paste with different PAM dosages were tested, and the results are shown in Fig. 3(a). The experimental data were fitted using a solid black line and a red dashed line, with correlation coefficients ($R^2$) of 0.993 and 0.991, respectively. Both the initial and final setting times exhibited a linear positive correlation with PAM content. This is attributed to the retarding effect of PAM molecules on early-stage hydration, which delays the solidification process of the paste. This inhibitory effect intensifies with increasing polymer content, leading to a continuous linear increase in setting times. The initial and final setting times increased at rates of 0.4777 and 0.4568, respectively, with increasing PAM content. While this phenomenon has been mentioned in [18], no reasonable explanation was provided. According to the nucleation mechanism of hydration established before (as shown in Fig. 4), as hydration progresses, the later-stage hydration rate of the accelerator gradually increases due to the overcoming of the hydration barrier, resulting in faster particle precipitation and coagulation. Consequently, the growth rate of the final setting time slightly decreases.

The 28-day compressive strength of pure cement paste was 33.2 MPa, whereas the addition of PAM led to a significant reduction, maintaining a range of 21-28 MPa (Fig. 1(b)). Due to PAM’s retardation effect on early hydration, the 1-day and 3-day compressive strengths showed a considerable gap compared to the reference sample. However, after the hydration barrier was overcome, the strength growth rate of the PAM-modified paste between 7 and 28 days exceeded that of the reference sample, gradually narrowing the compressive strength gap with prolonged curing. By the 28-day mark, the hydration degree of all PAM pastes became consistent. Nevertheless, due to their higher macro- and micro-porosity, as well as poor structural integrity and continuity, the compressive strength of the specimens remained at a relatively low level.

Fig. 3Setting time and compressive strength of slurry with different PAM content

a)

b)

3.2. Mechanical properties

Considering the practical engineering application quality of grouts, anti-washout Ordinary Portland Cement and Calcium Sulfoaluminate cement cement pastes with a water-to-cement ratio of 0.5, 2.0 % superplasticizer, and 0.6 % polyacrylamide (PAM) content were cast underwater and on land, respectively. A comprehensive evaluation of the anti-washout grout was conducted by comparatively analyzing hardened properties such as compressive strength, splitting tensile strength, bond splitting tensile performance at different curing ages.

Fig. 4Mechanism model diagram of materials with different PAM content [19, 20]

3.2.1. Compressive strength

The compressive strengths of underwater- and land-cast OPC and CSA cement specimens are shown in Fig. 5(a). Compared to land-cast samples, underwater-cast OPC cement exhibited reduced strength but maintained a low mass loss rate. The uniaxial compressive strengths of underwater-cast specimens at 1 d, 3 d, 8 d, and 28 d were 81.90 %, 90.56 %, 93.32 %, and 93.97 % of their land-cast counterparts, respectively. In contrast, due to its rapid early hydration, CSA cement achieved high underwater strength (16.78 MPa) at 1d, with underwater compressive strengths at 1 d, 3 d, 8 d, and 28 d being 83.94 %, 90.65 %, 99.18 %, and 97.57 % of land-cast specimens, respectively. Additionally, owing to more stable environmental conditions (e.g., temperature, humidity), underwater-cast specimens demonstrated lower strength variability than land-cast ones.

Fig. 5Compressive strength and Split tensile strength of OPC and CSA casted underwater and on land

a)

b)

3.2.3. Splitting tensile performance

The bond splitting tensile performance of underwater- and land-cast Ordinary Portland Cement (OPC) and Calcium Sulfoaluminate (CSA) cement specimens is illustrated in Figure 7. For OPC cement, the land-cast specimens exhibited a bond strength of 0.96 MPa at 1d. As hydration progressed, the bonding performance gradually improved, reaching 3.05 MPa at 28 d – comparable to the splitting tensile strength of pristine cement paste. In contrast, the underwater-cast specimens demonstrated bond splitting tensile strengths of only 21.88 %, 35.71 %, 33.91 %, and 40.30 % of their land-cast counterparts at 1 d, 3 d, 8 d, and 28 d, respectively. Notably, the growth rate of underwater bond strength significantly decelerated over time. CSA cement, benefiting from its rapid early hydration, achieved a land-cast bond strength of 2.5 MPa at 1 d – 70.42 % of its 28 d splitting tensile strength. However, underwater-cast specimens showed reduced performance, with bond strengths consistently measuring 26.0 %-29.0 % of land-cast values across all tested ages (1 d, 3 d, 8 d, and 28 d).

Fig. 6Bond strength of OPC and CSA cement casted underwater and on land

3.3. Impermeability strength

At a curing age of 28 days, the 1-day permeability test results of Ordinary Portland Cement (OPC) and Calcium Sulfoaluminate (CSA) cement are shown in Fig. 9, where (a), (b), (c), and (d) represent land-cast and underwater-cast specimens of OPC and CSA cement, respectively. The white coloration indicates the chromogenic reaction between nitrate solution and chloride ions, representing the actual penetration depth of chloride ions. The impermeability of the specimens was characterized by both penetration depth and the non-steady-state chloride migration coefficient, with results summarized in Table 2. The penetration depth was calculated using the bisection method, while the non-steady-state chloride migration coefficient was determined by the following formula:

where, $D_{RCM}$ refer to the non-steady-state chloride migration coefficient (×10^-12 m²/s); $U$ refer to absolute value of the applied voltage (V); $T$ refer to average temperature of the anolyte at the beginning and end of the test (°C); $L$ refer to thickness of the specimen (mm); $X_d$ refer to average chloride ion penetration depth (mm); $t$ refer to test duration (h).

Fig. 7Chloride permeation test results of underwater non - dispersible slurry

a)

b)

c)

d)

The data reveal that underwater-cast specimens exhibit greater chloride ion penetration depths compared to land-cast counterparts. For OPC, the penetration depth increased from 28.15 mm to 38.35 mm (36.23 % rise), while CSA showed an increase from 38.05 mm to 50.3 mm (32.19 % rise). Correspondingly, the non-steady-state chloride migration coefficients were significantly higher in underwater conditions: OPC increased from 4.279×10^-12 m²/s to 6.278×10^-12 m²/s (46.78 % rise), and CSA rose from 6.509×10^-12 m²/s to 9.410×10^-12 m²/s (38.75 % rise). The reduced chloride resistance of underwater-cast specimens is attributed to higher porosity and weakened structural connectivity caused by PAM, with OPC and CSA demonstrating comparable degradation rates. Notably, CSA cement inherently exhibits lower chloride penetration resistance than OPC. This phenomenon occurs because sulfate depletion in CSA systems triggers ettringite phase transformation into monosulfoaluminate, which reacts with chlorides to form Friedel’s salt – a process that intrinsically compromises chloride barrier performance.

According to engineering standards [16], underwater-cast cementitious materials need only achieve 80.0 % compressive strength and 60.0 % splitting tensile strength of their land-cast equivalents for practical applications. The developed OPC and CSA formulations in this study surpassed these thresholds, attaining 93.97 %/97.57 % (compressive) and 95.64 %/94.84 % (splitting tensile) of land-cast performance at 28 days. However, splitting tensile strength was reduced to approximately one-third of land-cast values due to interfacial water film formation and OPC alkali leaching, accompanied by a 45.0 % increase in chloride migration coefficients. CSA cement's rapid early-strength development makes it particularly valuable for emergency underwater repairs. Conversely, OPC’s superior chloride resistance positions it as the preferred choice for durable underwater concrete structures requiring long-term serviceability.