Strength and freeze-thaw cycle resistance of cement stabilized coastal clay mixed with shell powder

To study the effect of marine shell powder (SP) on the mechanical properties of cement solidified coastal clay (CSC), unconfined compressive strength test, freeze-thaw cycle test and Scanning Electron Microscope test were conducted on the stabilized soil. The results show that SP could improve the mechanical properties of CSC to some extent. When the SP content was 15%, the strength of CSC was the highest. Moreover, under the condition of freeze-thaw cycle, SP can enhance the compactness of coastal cement-clay, which shows that the pore area decreases by 7.2% during 7 freeze-thaw cycles. Finally, a mathematical empirical model of the unconfined compressive strength, SP content and freeze-thaw cycles of coastal cement-clay modified by seashell powder (SPCSC) specimens was established. The model has a good relevance with the investigated data, which can give a theoretical foundation for improving the performance of coastal cement-clay using seashell powder under different environments.


Introduction
Clay is widely distributed in coastal plain, estuarine delta, lake basin and intermountain valley. Most of the coastal clay was formed in the middle and late Quaternary, but some coastal clay was formed before the middle of Quaternary. However, coastal clay generally had large water content and pores, low shear strength and permeability coefficient, and cannot meet the bearing capacity strength indexes required by many projects [1][2][3].
To improve the defects of coastal clay, some cementing materials are often added into coastal clay in engineering. Cement is widely used to enhance coastal clay due to its good improvement effect on mechanical properties [4,5], but cement stabilized soil still has some issues include low tensile strength and high brittleness. Therefore, to further improve the engineering performance of coastal cement-clay, adding new solidified materials to cement has become the research direction of today's scholars [6][7][8][9].
In recent decades, the aquaculture industry has developed rapidly. The global aquatic production of shellfish aquaculture accounts for 42.6%, and many seashells are produced every year. A large number of produced seashells are directly abandoned in the coastal periphery, which has a serious impact on the coastal ecosystem [10][11][12]. Seashells are products with high calcium content. Once they are applied in geotechnical engineering, they will not only effectively dispose of seashells, but also promote the sustainable development of aquaculture. Morris et al [13] believed that shells from aquaculture was a valuable biological material. Barros et al [14] noted that the discarded shells disposal has become one of the most urgent problems in the current industry and studied the method and process of extracting calcium carbonate from shells. The extraction method is relatively mature and feasible. However, seashells with added value have long been regarded as waste discarded on beaches and roadside, which has become an urgent environmental problem to be solved in coastal areas. The utilization of seashells waste resources, but also promote the healthy development of shellfish aquaculture and realize the reciprocity of ecological environment, economic and social benefits. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Dahhou [15], Eziefula et al [16] showed that the main component of shell is calcium carbonate (disposal), which can replace some components of building materials and also the main component of cement clinker. Additionally, studies have shown that the aqueous calcium carbonate can enhance the strength of coastal clay foundation [17][18][19][20][21]. Park et al [22] studied the enhancement effect of calcium carbonate on the strength of sandy soil foundation through uniaxial compression test, showing that after curing for 3, 7, 14, 21 and 28 days, the uniaxial compressive strength of specimens increased to 84.57-378.86 kPa. The calcium carbonate strength created by microbial reaction can greatly improve the soil strength, and the strength increases with an increase in calcium carbonate content in the specimen. At present, some scholars use shell powder as building materials. Wang et al [23] applied shell powder to prepare the concrete used in harsh environment in cold areas and added shell powder into concrete as admixture to improve the frost resistance of concrete. Bamigboye Gideon et al [24] also used shell powder for sustainable utilization and production of concrete binder, and Naqi et al [25] used calcined oyster shell waste as an additive of slag cement and found that it could improve the strength of cement.
To sum up, although some researchers have applied seashell powder to the production of concrete, few studies have applied seashell powder to the strength improvement of coastal cement-clay. In this paper, the seashell powder is used to improve the mechanical properties of coastal cement-clay, and the strength characteristics and pore characteristics of coastal cement-clay enhanced by seashell powder are studied under the condition of freeze-thaw cycle. Moreover, the unconfined compressive strength tests (UCS) and scanning electron microscopy (SEM) tests were carried out on solidified soil with different content of seashell powder at 7, 14 and 28 days to characterize the mechanical properties of seashell powder modified coastal cement-clay. The specimens at 28 days after 0, 1, 3, 5 and 7 freeze-thaw cycles were tested for UCS and SEM tests, which provided theoretical basis for the application of seashell powder modified coastal cement-clay under different environments.

Materials
The coastal clay for this test was taken from the reconstruction and expansion project site of Shaoxing University in Zhejiang Province, China. Continuous gradation soil specimens with particle size of < 2 mm were obtained after crushing and screening. Its basic physical properties are shown in table 1, grain gradation curve is shown in figure 1(a) and specimen composition is shown in table 2. The used cement is P.O 42.5 ordinary Portland cement (Lanting brand of Shaoxing). Its composition is shown in table 3. and its main technical indicators are shown in table 4. The seashell powder selected for the test is produced by Yaoxin Mineral Products Processing  Plant in Lingshou County, and its basic properties are shown in table 5. Shells have a porous and fibrous double helix structure after modification and calcination [10]. The grain gradation curve of seashell power is shown in figure 1(b). Table 6 shows the test scheme for this paper. In table 6, the cement content is the mass ratio of cement to dried coastal clay, the seashell powder content is the mass ratio of seashell powder to dried coastal clay, and the water content is the mass ratio of water mass to the mixture (dried coastal clay, cement and seashell powder). The specimen number of seashell powder cement is aSP-t, a represents the content of seashell powder, %, t represents the curing time, d. The freeze-thaw cycle test scheme is shown in table 7. The freeze-thaw cycle tests were carried out at 28 days of standard curing. This paper mainly studies the modification effect of SP on the mechanical properties of CSC in different environments, and the experimental results will be applied to engineering. In practical engineering, the optimal dosage of materials is usually used, so we only conduct freeze-thaw cycle tests on the reference group and the optimal dosage group [26]. The specimens were frozen at −20°C for 12 h and melted at 20°C for 12 h, and 24 h was a freeze-thaw cycle. The specimen number of seashell powder modified coastal cement-clay is aSP-28-n, n represents the freeze-thaw times of the specimen in the freeze-thaw box, times, a represents the amount of seashell powder, %. In this paper, we determined that the test should be stopped when the compressive strength loss rate of the specimen reaches 25% or the mass loss rate reaches 5% during the freeze-thaw cycle (GB/T 50082-2009). Because when these two situations occur, the specimen has basically been destroyed [26].

Specimen making
The preparation of UCS specimens is done in accordance with the Standard of Geotechnical Test Method (GB/T 50123-2019). The specific specimen preparation flow chart was shown in figure 2. The specific steps are as follows: (1)Put the coastal clay into the oven for drying; (2) According to the designed mix ratio, weigh the clay, seashell powder, cement, as well as water together and mix it evenly; (3) Weigh mixed materials (205 g per specimen) and pour it into the mold; (4) Compact the mixed materials with a jack to produce column specimen with size of 39.1 mm diameter and 80 mm height, one by one; (5) Cure the specimens at standard environment under 25 ± 2°C temperature, 95% humidity.

Test method
This experiment was conducted using an HBY-40B cement (concrete) constant temperature and humidity standard curing box produced by Zhejiang Geotechnical Instrument Manufacturing Co., Ltd. The freezingthawing cycle process was carried out at a DW-40 low temperature test chamber produced by Zhejiang Luda Mechanical Instrument Co., Ltd. UCS test was conducted using an TKA-WCY-1F automatic multifunctional unconfined compressive strength tester produced by Nanjing Teco Technology Co., Ltd. The microstructure of the specimen is observed using a SEM instrument JSM-6360LV produced by Japan Electronics Co., Ltd.

UCS test
(1)Unconfined compressive strength test: the unconfined compressive strength (UCS) test was carried out at a loading rate of 1 mm min −1 according to the test of Highway Geotechnical Test Regulations (JTG 3430-2020) [27]. When the stress dropped to 0 kPa or peaked, the strain stopping test of 3%-5% was continued. Each test was repeated 5 times and the average value was taken.
(2)Freezing-thawing cycle test: freezing-thawing cycle treatment is carried out on SPCSC specimens prepared according to Highway Geotechnical Test Regulations (JTG 3430-2020). One freeze-thaw cycle was 24 h, and SPCSC specimens were placed in a −20°C low temperature test chamber and frozen for 12 h. After freezing, the SPCSC specimen was thawed in the standard curing room for 12 h. Repeat the above steps for the next freeze-thaw cycle [28]. The freeze-thaw cycle used in this experiment is 0, 1, 3, 5, K times. The test will be stopped when the compressive strength loss rate of the specimen reaches 25% or the mass loss rate reaches 5% during the freeze-thaw cycle. The UCS test is conducted after the set number of freeze-thaw cycles.

SEM test
During the specimen production process, we have thoroughly stirred various materials. Breaking the specimen into powder can make the specimen particles themselves more visible in SEM, which effectively displays the microstructure relationships within the specimen particles. The damaged specimens from UCS test were placed in the oven for 24 h of drying, and the specimens were ground into powder after drying [29], and then placed on slides respectively, as shown in figure 3(a). The ion sputtering instrument was used to spray gold (figure 3(b)), and then placed on the specimen table [30]. Finally, the SEM images were shot under different magnification ratios using vacuum scanning electron microscopy (SEM) instruments as shown in figure 3(c). The magnification ratios in this study were ×2000 and ×5000.

Results and discussions
3.1. Standard curing condition 3.1.1. Stress-strain curve The stress-strain diagrams of SPCSC specimens with different seashell powder contents at the curing time of 7 days and 28 days are shown in figure 4, from which it can be found that the stress-strain curves of SPCSC are all softening curves. Under the same curing time, when the seashell powder content is within 20%, with the seashell powder content increasing, the strength of SPCSC increases and increases first and then decreases. At the curing time of 7 days, when the content of seashell powder increases from 5% to 10%, the strength increases the most. At the curing time of 28 days, the strength increases the most when the seashell powder content increases from 10% to 15%. When the content of seashell powder exceeds 15%, the specimen is easy to fracture during demolding, so optimal seashell powder content is 15%. Excessive seashell powder led to a decrease in the strength increase of SPCSC specimens, mainly because the seashell powder could not disperse freely, so the hydration reaction of cement was not sufficient, and there was clumping accumulation, thus connection structure between particles was not cemented, which affected the strength.  Coastal clay is characterized by high natural water content, large natural pore ratio, and high compressibility. Seashell powder and limestone powder have similar mechanisms and a certain filling effect in coastal clay [31]. The gap between the coastal cement-clay particles is large and the filling is poor. The seashell powder can be filled between the coastal clay particles to improve the grain distribution of the soil material and enhance the density of soil.
Seashell powder has a certain morphology (dilution) effect, it can improve the fluidity of coastal clay, show a certain effect of water reduction, and improve the comprehensive performance of coastal clay. Coastal clay is rich in mineral compositions. Except for quartz, feldspar and mica in silty grains, minerals in coastal clay grains are mainly illite, followed by kaolinite. Additionally, coastal clay contains a certain amount of organic matter, which can be as high as 8% ∼ 9%. The main component of seashell powder CaCO 3 can form hydration products such as carboaluminate complex with mineral components in coastal clay. Such hydration products bond with other hydration products to increase the compactness of coastal clay, thus improve the strength and durability of coastal clay [32].

Stress strain curve
The optimal seashell powder content is 15%. So, we only conduct freeze-thaw cycle tests on 0%SPCSC and 15% SPCSC specimens. The stress-strain curves of 0%SPCSC and 15%SPCSC specimens at 28 days of curing time under different freeze-thaw times are shown in figure 6, it can be found that the stress-strain curves of SPCSC specimens are all softening curves. The stress-strain curve shows that the unconfined compressive strength of SPCSC decreased with the increase of freeze-thaw times. Under freeze-thaw cycle, when the temperature is lower than 0°C, the frozen volume of water in the pore of the specimen expanded. When the temperature is higher than 0°C, the ice in the soil melts and shrinks, and the internal structure of the specimen is damaged by repeated frost heave and melt contraction. Moreover, the continuous extension and cracks development of soils lead to the increase of soil particle gap and the specimen strength decrease. From the figure 6, the strength of 15% SPCSC is significantly higher than 0%SPCSC under the same freeze-thaw cycle, indicating the seashell powder addition can play a role in filling pores, which is conducive to improving the damage caused by frost swelling and thawing contraction of specimens [33].

Unconfined compressive strength
The unconfined compressive strength of SPCSC specimens under different freeze-thawing times at 28 days is shown in figure 7 below. As can be seen from figure 7, under the curing time of 28 days, the unconfined compressive strength of SPCSC specimens shows a decreasing trend with the increase of the number of freezethaw cycles. The freeze-thaw cycle can significantly reduce the unconfined compressive strength of SPCSC. From the microscopic perspective, the change of freeze-thaw temperature and the rise in the number of freezethaw cycles will lead to the gradual decomposition of large particles into small particles in coastal clay, and the arrangement of particles will change, leading to the corresponding reduction of compressive strength. The strengths of 0%SPCSC freeze-thaw cycles 0, 1, 3, 5 and 7 were 2035, 1515, 1373, 1149 and 900 kPa, respectively. The strengths of specimens after 7 freeze-thaw cycles decreased by 55.7% compared with those without freezethaw cycles. The strength of 15%SPCSC freeze-thaw cycles 0, 1, 3, 5 and 7 were 2301, 1908, 1854, 1775 and 1669 kPa, respectively. The strength of specimens after 7 freeze-thaw cycles decreased by 27.5% compared with that of  specimens without freeze-thaw cycles, indicating that the seashell powder addition, the ability of specimen to resist the damage caused by freeze-thaw cycle has been significantly improved. After 7 freeze-thaw cycles, the strength of both groups of specimens decreased by more than 25%, so we only tested 0, 1, 3, 5 and 7 freeze-thaw cycles. The seashell powder addition can make the hydration reactants generated by the hydration reaction of cement fill the pore between soil particles together, the pore size of the specimen becomes smaller, the soil compactness increases, and the strength and permeability of soil-cement improve. The saturated moisture content of soil in the specimens mixed with seashell powder is reduced, thus the freeze-thaw damage is weakened under the condition of freeze-thaw cycle. Therefore, the strength of soil subjected to the freeze-thaw cycles can be improved by adding seashell powder.

Mathematical model
To better characterize the relationship between the unconfined compressive strength of SPCSC specimens and the seashell powder content and the number of freeze-thaw cycles, the fitting was carried out in a 3D coordinate with seashell powder as the X-axis, the number of freeze-thaw cycles as the Y-axis, and the strength as the Z-axis. The fitting curve at 28 days was shown in figure 7, and the fitting equation obtained was shown in equation (1) According to equation (1), the unconfined compressive strength of SPCSC specimens can be calculated at any dosage of seashell powder and any number of freeze-thaw cycles. The measured values are plotted in the fitting curve function diagram shown in figure 8, the fitting results are in good agreement with the measured data, and the measured values are evenly distributed above and below the fitting function. It can provide a good mathematical description of the strength variation law of SPCSC specimen. This fitted equation can provide a theoretical basis for the actual engineering of seashell powder reinforcement cement coastal clay, which has important engineering significance.

Microscopic mechanism analysis
5.1. XRD analyses Figure 9 shows the XRD patterns of SP and 0%SPCSC and 15%SPCSC specimens at different curing times. Some hydration products usually include calcium silicate hydrate gel (C-S-H), calcium hydroxide (Ca(OH) 2 ) and a certain amount of ettringite are found. Each XRD patterns in the figure has obviously similar diffraction peaks for hydration products. The diffraction peak structure of the specimen with seashell powder changed because seashell powder mainly plays a filling role in cement and does not react with Portland cement [34].  Figure 10 shows SEM images of SPCSC specimens at × 2000 and × 5000 magnification. The image clearly shows that the cement hydration products from 0%SPCSC specimen are relatively loose and have relatively obvious pores. When the seashell powder content is 15%, the SEM image of SPCSC specimen can clearly see that there are only small cracks, and the pores are obviously reduced. The seashell powder of 15%SPCSC small and medium-sized particle size filled the micro pores in the specimen, making the structure denser. Compared with 0%SPCSC specimen, its overall structure was significantly improved, so its strength increased. The particle size of seashell powder is smaller than that of cement, so it can fill in the gap between the cement slurry matrix and the interfacial transition zone, making the coastal cement-clay structure and the interfacial structure denser. As a result, the porosity and pore size are reduced, the pore structure is improved, and the specimen strength is increased. Figure 11 shows the SEM (×2000 and × 5000) images of SPCSC-28-7 specimens. As shown in this figure, specimens have many large pores. The freeze-thaw cycle also weakens the cementation between particles, and the freeze-thaw cycle destroyed the structure between soil particles. When the seashell powder content increased to 15%, there is no large area fracture and porosity in the SEM image, indicating that the seashell powder addition significantly improves the resistance to the damage caused by the freeze-thaw cycle. After the freezethaw cycle, the cementation between soil particles was weakened.

SEM analysis of solidified soil under freeze-thaw cycle.
When the specimen was exposed to the freeze-thaw cycle, the free water in the soil body generates ice crystals, and compresses the surrounding soil particles, forcing the relative displacement of the soil particles. As a result, the cementation between the soil particles because weakened, and the microstructure of the soil particles also changed. The internal structure of the specimen was damaged by repeated frost heave and thawing contraction, and the continuous extension and development of cracks between soil particles led to the increase of soil particle gap and the weakening of cementation between soil particles, which was mainly represented by the development of micropores into medium pores. When the seashell powder is used, the pore between soil particles was filled, which makes the pores become smaller, increases the soil compactness, and weakens the freeze-thaw damage.

SEM image processing
The image obtained by scanning electron microscope is a gray image, which can clearly distinguish the distribution of pores and soil particles between soils [35]. The magnification ratios in this study were 2000× and 5000×. The pores mentioned in the article are not the ones we see in the traditional sense, but rather the pores in specimen particles magnified thousands of times by scanning electron microscopy, which are micrometer sized. In previous studies by scholars, the method of crushing specimens into powder was also used in SEM experiments [29] and calculating particle pore area using SEM images [36,37]. Meanwhile, to quantitatively analyze the images obtained by SEM, the image-Pro Plus 6.0 software was used to binary process the gray SEM Image (figure 11): the threshold segmentation method is used to define the boundaries of particles, with a particle grayscale threshold of (0, 120) [36], and the image is adjusted to a black and white binary image. Some SEM images after processing are shown in figure 12, in which the particle is denoted as black and the pores as white. Then, the area and area ratio of measurement indicators are selected to calculate the pore area parameters of each specimen and output the microstructure information needed for the research [38].

Pore area ratio of soil
Due to the fact that SEM images are usually very random, the pore area parameters of each specimen are calculated by averaging the data extracted from three images of different parts of the same specimen (figure 13), as shown in table 8. As the seashell powder content increased, the pore area ratio of the specimen gradually decreased. When the seashell powder content was 20%, the pore area was at the minimum (1.94%). The pore area ratio of 0%SPCSC and 15%SPCSC specimens increased with the increase of freeze-thaw times, and the pore area of specimens increased significantly during the first freeze-thaw cycle. When the number of freeze-thaw cycles continued to increase, the pore area increases of 0%SPCSC specimen gradually became larger. In the 7th freeze-thaw cycle, the pore area ratio increased to 15.63%, which was 67.23% higher than that of the first freezethaw cycle. The pore area of 15%SPCSC specimen increases slowly. In the 7th freeze-thaw cycle, the pore area ratio increases to 8.43%, which is 30.05% higher than that of the 1st freeze-thaw cycle. Combined with the unconfined compressive strength results, it is found that the pore ratio of specimen is negatively correlated with the compressive strength. The larger the compressive strength of the specimen is, the smaller the pore area ratio of the specimen is. Meantime, it also reflects that when the pore area ratio of the specimen is smaller, the soil particles became denser, the specimen integrity became better, and the compressive strength of the specimen was greater in the macro performance.

Conclusions
(1)Seashell powder can further improve the strength and anti-freeze-thaw cycle of coastal cement-clay. The unconfined compressive strength of solidified soil increased with an increase in seashell powder content. When the seashell powder content is 15%, the stabilized coastal clay has a maximum strength, and seashell powder can improve the early strength of coastal cement-clay.
(2)Seashell powder can enhance the freeze-thaw cycle resistance of coastal cement-clay. When the seashell powder content is 15%, the strength loss of the specimen exposed to 7 freeze-thaw cycles is 28%, which is less than that of the specimen without seashell powder.
(3)Based on the mathematical analysis method, the mathematical model of the unconfined compressive strength of SPCSC specimens, the amount of seashell powder and the number of freeze-thaw cycles was established, and the empirical formula of the unconfined compressive strength of SPCSC specimens was obtained. The correlation coefficient R 2 = 0.92, indicating a good fitting result.
(4)The seashell powder addition can play a filling role, forming a good overall structure with free soil particles, reducing pores, improving soil compactness, and thus improving strength.
(5)With the seashell powder content increasing, the pore area decreases gradually. The pore area of the specimen increases with the increase of freeze-thaw times, and the pore area of the specimen increases slowly after adding seashell powder. When the compressive strength became larger, the pore area ratio became smaller, and the pore area ratio was negatively correlated with the strength.

Data availability statement
The data cannot be made publicly available upon publication because they are not available in a format that is sufficiently accessible or reusable by other researchers. The data that support the findings of this study are available upon reasonable request from the authors.