Flexural Strength of High-Performance Soil-Cement: A New, Alternative, Sustainable Construction Material

Abstract

Soil-cement is a building material that is considered low-cost and has a low environmental impact. Despite its benefits, performance optimisation studies are scarce compared to other materials such as concrete. The possibility of obtaining soil-cement with improved characteristics, such as flexural strength, would enable the increased use of this product in new applications in construction. The aim of this study is to produce high-performance soil-cement (HPSC) specimens and to evaluate and compare this new material with high-performance concrete (HPC) in terms of flexural strength. A total of 12 specimens were produced with a mixture of 23.5% (by mass) of cement with the application of 10 MPa of pressure for its compaction. The results show that, at 28 days, the specimens reached an average strength of 6.73 MPa and, at 240 days, 12.34 MPa. This means that the HPSC reached a flexural strength resistance equivalent to HPC without the need for mined materials, such as sand and gravel, or the additives adopted in some doses of HPC, such as superplasticisers. Therefore, when using local soil, HPSC can be considered an environmentally preferable alternative to HPC for many construction applications where flexural strength is a requirement.

1. Introduction

Soil-cement is a material composed of a homogeneous, compacted mixture of soil, cement, and water [1]. After the compaction process, it acquires strength over time without the need for the firing process used in common bricks [2,3], although the Portland cement used in the mixture does require firing. The use of soil-cement for construction purposes is similar to the adobe and rammed earth processes [4,5]. The main difference is that in the production of soil-cement, a percentage of binder (cement) is added, enabling the mixture to better withstand degradation due to weathering [6].

The possibility of using soil excavated from the same construction site to produce this material means it can be considered to be of low environmental impact [7]. It uses local soil that would need to be excavated anyway for construction (for the foundations, for example) rather than the aggregates (gravel and sand) used in concrete. It is worth noting that, according to Hammond and Jones [8], transport is a significant contributor to the embedded energy and carbon for aggregates. Therefore, reusing the local soil reduces energy consumption and greenhouse gas emissions in comparison to the use of gravel and sand for concrete in several ways as, with the soil being extracted locally, there is no need to transport gravel and sand or the excavated soil to a landfill since it is reused on site. Reusing the local soil also reduces waste generation. Consequently, soil-cement also has a lower cost than concrete due to this decrease in the use of gravel and sand as aggregates and the reduced need for the transportation of materials to and from the construction site. Another positive factor to consider is the possibility of reusing various types of waste, partially replacing the soil with construction waste [9], steel fibre waste [10], or mining waste [3], among others found in the literature. These factors make soil-cement a sustainable, alternative construction material.

Currently, the most common applications using the soil-cement technique refer to the production of blocks for walls [11], pavement [12,13,14], and foundation piles [15]. However, there are other types of applications, precisely because of the ease of soil-cement production in formats adapted to construction needs. A recent example is the development of soil-cement blocks in a semi-circular shape for erosion control [16].

Despite all the benefits of soil-cement, its development in terms of performance has been modest when compared to other materials such as concrete [17], which has been constantly developing since its first appearance [18,19]. High-performance concrete (HPC) first emerged in the 1970s [19] as a type of concrete with improved performance in aspects of strength, durability, and/or other characteristics that go beyond the values of conventional concrete [20,21]. The difference between concrete and soil-cement is that a similar evolution has not yet been observed.

HPC is a widely searched term in reference to concrete, as it is a widely used material. In a search on Google Scholar for the term “high performance concrete“, 141,000 results were found. However, there is little published work on high-performance soil-cement (HPSC), with only three studies [17,22,23] being found in the literature when searching for ”high performance soil-cement”. In the study by Duong et al. [22], this type of material refers to the soil-cement used to improve the soil conditions for paving and/or improve the settlement of foundations and not for the production of construction pieces. The study by Carvalho et al. [17] defines which production parameters should be improved to obtain a soil-cement with performance characteristics similar to an HPC. The study by Rocha et al. [23] produced HPSC pieces with 4.6 to 5.8 MPa of flexural strength using 30 to 50% cement in the mixture. Currently, no references have been found in the literature for soil-cement with a similar performance to that of HPC, that is, with compressive strengths reaching 50 MPa. The typical values for the compressive strength of soil-cement range from 1 to 10 MPa. When searching for references using the term ”high-strength soil-cement“, the highest values of compressive strength found refer to the production of blocks, with 9.86 MPa with 8% cement [24], 8.34 MPa with 14.28% cement [25], 6.76 MPa with 15% cement [26], and 7.19 MPa with 12% cement [27]. Thus, there is a gap in the literature on the study of soil-cement products with similar performances to HPC. The possibility of obtaining a soil-cement with better characteristics (flexural strength > 2.5 MPa and/or compression > 41 MPa, water absorption less than 8%, and durability) would enable the increased use of this product in new applications in construction [17].

This study aims to determine the flexural strength of HPSC specimens and is the first to test the specimens produced according to all the appropriate production parameters found in the literature and compare them with those of high-performance concrete (HPC). The specific objectives of the study are:

• To (a) review the literature on the relevant production parameters of soil-cement, their appropriate values, and the flexural strengths obtained when evaluated according to the test specified in ASTM C293 [28]; and (b) perform a limited literature review on the flexural strengths of HPC obtained as per ASTM C293 [28] and according to the cement content.
• To produce soil-cement specimens according to the appropriate parameters found in the literature (cement content, compaction pressure, moisture, and granulometry) and test them for flexural strength at different moments (7, 28, and 240 days), as per ASTM C293 [28], and evaluate the water absorption according to the NBR-13555 standard [29].
• To draw conclusions on the flexural strength of the HPSC produced and compare the results with those of HPC while considering cement consumption. If they are equivalent, HPSC could be a viable alternative to be considered for future new applications in construction.

Figure 1 shows how this research was structured to achieve these objectives. Initially, in Phase 1, a literature review was carried out to understand how soil-cement production parameters affect its performance. Consequently, a gap in the knowledge was revealed as no studies were found in which all the production parameters were simultaneously optimised to achieve the maximum flexural strength in soil-cement. Additionally, a literature review was conducted on the flexural strength of concrete pieces and their dependence on the cement content. In Phase 2, HPSC pieces were produced according to the literature review to achieve the maximum possible flexural strength and were then tested. In Phase 3, the flexural strength of HPC (data obtained from the literature review) was compared with that of the HPSC specimens (data obtained from the tests carried out in this study). Finally, in the conclusions, it is explained that HPSC can be a sustainable solution for some specific solutions in the civil engineering field.

2. Literature Review

2.1. Literature Review on the Production Parameters of Soil-Cement Pieces

According to a preliminary literature review, the production parameters that can be improved to increase the performance of soil-cement are:

• The compaction pressure—An increase in this parameter improves the strength of the soil-cement product [30]. According to the literature, the optimum values for compaction pressure are 4 to 10 MPa [31]. Studies of soil-cement produced with a compaction pressure over 10 Mpa were not found in the literature.
• The moisture of the material—Compacting the mixture with the optimum compaction moisture leads to an increase in product strength [32,33]. The dry density of the soil after compaction depends on the soil type, compaction energy, and moisture content of the mixture [32,34].

Figure 2 was developed by the authors using the data published by Baldovino et al. [32]. It illustrates the compaction curves of two different soil types (S1 and S2) with three different compaction energies (E1, E2, and E3). From this graph, it can be observed that for the same soil type, as the compaction energy increases (E1 < E2 < E3), the dry density of the compacted soil also tends to increase. Additionally, the optimum moisture content for achieving the maximum dry density decreases as the compaction energy increases. When the soil has larger grain sizes, there is a tendency for the optimum compaction moisture content to decrease as well. In the case presented in Figure 2, soil S2 has a coarser particle size distribution compared to soil S1.

3.

The amount of cement—Increasing the percentage of cement increases the strength [35], but it also increases its cost and environmental impact. To make a fair comparison between a soil-cement product and conventional concrete, it is necessary to know if both materials use a similar amount of cement to achieve the same strength. If similar amounts of cement are used on average, concrete would have a greater environmental impact. This is because, in addition to cement, other materials from mining (such as sand and gravel) that require transportation are used in concrete, and in many cases, chemical additives such as superplasticisers are also used [36]. Soil-cement consists of only two materials, soil, for the most part, and cement, and the soil can be extracted and reused from the construction site itself or nearby, thus reducing the need for transport and its consequent environmental impact. For this reason, it is essential to analyse whether the use of soil-cement for the production of construction pieces consumes similar amounts compared to producing concrete with a similar performance.

4.

The granulometry of the soil—According to Venkatarama et al. [37], well-graded soils with percentages of clay lower than 15% should be used. According to Carvalho et al. [17], the ideal soil should allow the mass ratio of clay to cement in the soil-cement mixture to be 30%. Another possibility would be to use a grain size distribution adjusted to the theoretical particle size curve using the packing method [38].

2.2. Literature Review on the Flexural Strength of Soil-Cement Pieces

Typically, compressive strength is a highly analysed value in building materials and is commonly calculated when evaluating the performance aspects of concrete. However, for certain uses and applications, other values are more significant depending on the intended use. For instance, in the production of slabs or pavements, it is more important to assess flexural strength [21]; so, if the slabs are to be made with soil-cement, it would be necessary to obtain flexural strength values. So far, one of the most common products made with soil-cement is blocks [39], but if soil-cement is to be used for producing slabs, it is, therefore, important to evaluate its flexural strength.

A literature review was conducted to assess how soil-cement production parameters can affect flexural strength. More specifically, the objective was to analyse studies on soil-cement that evaluated flexural strength using the test specified in ASTM C293 [28]. These data can help establish a foundation of what has already been produced and understand why the flexural strength of this material has not reached similar values to those of high-performance concrete (HPC).

The literature review was carried out using the Google Scholar platform searching for the keywords “soil-cement” and “ASTM C293” to obtain studies on soil-cement materials that could provide performance values related to flexural strength. A total of 41 studies were found, of which those related to soil-cement blocks or specimens that described the values of their flexural strengths based on cement percentages were selected. Even though one of the searched keywords was “soil-cement”, some of the studies analysed different aspects. Therefore, the studies related to the following themes were excluded from the subsequent analysis: (1) concrete elements; (2) walls made with soil-cement blocks where the flexural strength evaluation refers to the entire wall, including blocks and mortar; (3) soil-cement products for pavement applications where flexural analyses are based on cyclic loads; and (4) studies where specific flexural strength test values could not be found. This process resulted in 13 studies out of the 41 to continue with the analysis.

From the literature review, it could be concluded that the most important parameters influencing soil-cement flexural strength are compaction pressure, compaction moisture, and cement percentage, as summarised in Table 1. It should be noted that, in some studies, many samples with different values for these parameters were tested. Table 1 only includes the proportions from each study that produced the specimens with the highest flexural strengths. As observed in Table 1, the production parameters considered were:

• Compaction pressure: The usual values in the literature, when found, ranged from 1 to 2.5 MPa. In several studies, it was difficult to find this data, and imprecise values were often reported. For example, in the studies by Jose et al. [40], it was mentioned that manual compaction was performed without specifying the compaction pressure. As explained earlier, this compaction pressure could reach up to 10 MPa to achieve the highest possible strengths [31].
• Compaction moisture: It was described in two ways depending on the study, as the water-to-cement mass ratio (w/c) or as a percentage (%) of the weight of the dried piece. The w/c form is commonly used when producing concrete components. Lower w/c ratios are usually associated with higher strengths in concrete [21]. However, this is not always the case with soil-cement. As explained earlier, there is an optimum compaction moisture for a specific energy level and soil type (Figure 2). Typically, to facilitate the production process of soil-cement, the compaction moisture values used are below the optimum value. In this case, an increase in moisture (higher water content) would increase the compaction of the specimen and, consequently, its strength. Three studies directly described the values of compaction moisture as a percentage of the weight of the dried piece: Rocha et al. [23], with values of 13% moisture when using 30, 40, and 50% cement; Sitton et al. [41], with values of 11.4% moisture when using 10.9% cement; and Donkor et al. [42], with values of 9% moisture when using 8% cement. The specific optimum compaction moisture cannot be determined as it varies depending on the compaction energy used. However, when the compaction pressures are below 2.5 MPa, it is estimated that the moisture should be above 12% as in the study by Souza et al. [43].
• Cement percentage used: As previously explained, higher cement percentages result in higher strength. However, for environmental and cost reasons, this parameter should ideally be kept as low as possible. By examining some examples of concrete products [44,45], it was estimated that flexural strengths close to or greater than 6 MPa can be achieved with cement percentages ranging from 17 to 23%. In most of the soil-cement studies found, the cement percentage values are below 13%. Only in the studies by Rocha et al. [23], Namboonruang [46,47], and Ademati et al. [48] were higher cement percentages used.

Table 1. Summary of the literature review of the flexural strength in soil-cement specimens.

Reference	Dimensions of the Piece	Compaction Pressure (MPa)	Compaction Moisture	Cement Percentage (by Weight)	Flexural Strength (MPa) (at 28 Days of Age Unless Indicated Otherwise)
Jose, A. and Kasthurba, A. K. (2021a) [49]	305 × 143 × 100 mm	2.5	w/c = 0.3	2.5 to 12.5%	<2 MPa
Namboonruang, W., Rawangkul, R., Yodsudjai, W., Suphadon, N., Boongurd, A., and Yong-Amnuai, P. (2013) [50]	125 × 250 × 100 mm	-	w/c = 0.26 to 0.45	10%	<1.6 MPa
Rocha, H. J. A., Roa, J. P. B., de Carvalho, F. A., Junior, M. F. D. S. S., Gonçalves, H. H. A., Junior, A. C. L. A., and Prat, B. V. (2023) [23]	140 × 60 × 40 mm	1.2 MPa	Moisture 13%	30%	4.6 MPa at 120 days
				40%	5.6 MPa at 120 days
				50%	5.8 MPa at 120 days
Namboonruang, W., Rawangkul, R., Yodsudjai, W., and Suphadon, N. (2012a) [46]	125 × 100 × 250 mm	-	w/c = 0.165 for 8% of cement w/c = 0.35 for 35% cement w/c = 0.45 for 50% cement	8%	<1.5 at 28 days <2.0 at 360 days
				35%	4.5 at 28 days 6.4 at 360 days
				50%	<5.5 at 28 days <8.5 at 360 days
Namboonruang, W., Rawangkul, R., and Yodsudjai, W. (2012b) [47]	125 × 100 × 250 mm	-	w/c = 0.45	50%	<5.5 MPa at 28 days <7.5 MPa at 180 days
Curto, A., Lanzoni, L., Tarantino, A. M., and Viviani, M. (2020) [51]	150 × 150 × 150 mm	-	-	12.5%	< 2.28 MPa
Jose, A., and Kasthurba, A. K. (2021b) [40]	305 × 143 × 100 mm	Manual Press	w/c = 0.3	2.5 to 12.5%	<3.5 MPa for 12.5% cement
Sitton, J. D., Zeinali, Y., Heidarian, W. H., and Story, B. A. (2018) [41]	355 × 110 × 180 mm	-	11.4% for 10.9% of cement	3.6 to 10.9%	<2.01 MPa for 10.9% cement
Donkor, P., Obonyo, E., Matta, F., and Erdogmus, E. (2014) [42]	413 × 102 × 102 mm	1.6 MPa	9%	8%	< 1.03 MPa
Namboonruang, W., and Yongam-Nuai, P. (2016) [52]	125 × 250 × 100 mm	-	w/c = 0.15 to 0.39	10%	<1.7 MPa at 200 days
Ademati, A. O., Akinwande, A. A., Balogun, O. A., and Romanovski, V. (2022) [48]	Clay Bricks	-	-	24.82%	1.98 MPa
Banker-Hix, W. A. (2014) [53]	-	-	-	8 to 12%	<1 MPa
Dormohamadi, M. And Rahimnia, R. (2020) [54]	-	-	-	0%	<2 MPa

Table 1. Summary of the literature review of the flexural strength in soil-cement specimens.

Reference	Dimensions of the Piece	Compaction Pressure (MPa)	Compaction Moisture	Cement Percentage (by Weight)	Flexural Strength (MPa) (at 28 Days of Age Unless Indicated Otherwise)
Jose, A. and Kasthurba, A. K. (2021a) [49]	305 × 143 × 100 mm	2.5	w/c = 0.3	2.5 to 12.5%	<2 MPa
Namboonruang, W., Rawangkul, R., Yodsudjai, W., Suphadon, N., Boongurd, A., and Yong-Amnuai, P. (2013) [50]	125 × 250 × 100 mm	-	w/c = 0.26 to 0.45	10%	<1.6 MPa
Rocha, H. J. A., Roa, J. P. B., de Carvalho, F. A., Junior, M. F. D. S. S., Gonçalves, H. H. A., Junior, A. C. L. A., and Prat, B. V. (2023) [23]	140 × 60 × 40 mm	1.2 MPa	Moisture 13%	30%	4.6 MPa at 120 days
				40%	5.6 MPa at 120 days
				50%	5.8 MPa at 120 days
Namboonruang, W., Rawangkul, R., Yodsudjai, W., and Suphadon, N. (2012a) [46]	125 × 100 × 250 mm	-	w/c = 0.165 for 8% of cement w/c = 0.35 for 35% cement w/c = 0.45 for 50% cement	8%	<1.5 at 28 days <2.0 at 360 days
				35%	4.5 at 28 days 6.4 at 360 days
				50%	<5.5 at 28 days <8.5 at 360 days
Namboonruang, W., Rawangkul, R., and Yodsudjai, W. (2012b) [47]	125 × 100 × 250 mm	-	w/c = 0.45	50%	<5.5 MPa at 28 days <7.5 MPa at 180 days
Curto, A., Lanzoni, L., Tarantino, A. M., and Viviani, M. (2020) [51]	150 × 150 × 150 mm	-	-	12.5%	< 2.28 MPa
Jose, A., and Kasthurba, A. K. (2021b) [40]	305 × 143 × 100 mm	Manual Press	w/c = 0.3	2.5 to 12.5%	<3.5 MPa for 12.5% cement
Sitton, J. D., Zeinali, Y., Heidarian, W. H., and Story, B. A. (2018) [41]	355 × 110 × 180 mm	-	11.4% for 10.9% of cement	3.6 to 10.9%	<2.01 MPa for 10.9% cement
Donkor, P., Obonyo, E., Matta, F., and Erdogmus, E. (2014) [42]	413 × 102 × 102 mm	1.6 MPa	9%	8%	< 1.03 MPa
Namboonruang, W., and Yongam-Nuai, P. (2016) [52]	125 × 250 × 100 mm	-	w/c = 0.15 to 0.39	10%	<1.7 MPa at 200 days
Ademati, A. O., Akinwande, A. A., Balogun, O. A., and Romanovski, V. (2022) [48]	Clay Bricks	-	-	24.82%	1.98 MPa
Banker-Hix, W. A. (2014) [53]	-	-	-	8 to 12%	<1 MPa
Dormohamadi, M. And Rahimnia, R. (2020) [54]	-	-	-	0%	<2 MPa

It can be observed that, in general, the flexural strength values obtained at 28 days were always below 2.5 MPa. The only flexural strengths that exceeded 2.5 MPa were reported in the study by Rocha et al. [23], achieving 5.8 MPa at 120 days, and those of Namboonruang et al. [46,47], who obtained values <5.5 MPa at 28 days, albeit using a very high cement percentage (up to 50%). Therefore, even when using a high percentage of cement, the flexural strength values remained below 6 MPa at 28 days.

None of the analysed studies produced a soil-cement piece optimising the three essential production parameters at the same time to achieve maximum strength: (1) compaction pressure with values close to 10 MPa; (2) optimum compaction moisture for that pressure; and (3) cement percentages above 20%. Thus, in the present study, the soil-cement specimens were produced with these appropriate parameters to determine the maximum flexural strengths that can be achieved in soil-cement pieces.

2.3. Literature Review on the Flexural Strength of High-Performance Concrete

In the last phase of the literature review, the flexural strength of HPC was analysed. The objective was to determine whether, with the use of similar amounts of cement, the produced HPSC achieved similar performance values to HPC. To perform this comparison, studies with high or ultra-high-performance concrete dosages with flexural strength values obtained according to ASTM C293 [28], as in the present study, were searched for on the Google Scholar platform. The key search terms were ”high performance concrete” and ”ASTM C293”. The intention was not to conduct an exhaustive review of the literature, but to find the relevant publications on these terms to compare whether the flexural-strength values of the HPSC produced in the present study are similar or not to those of HPC.

Subsequently, articles were selected where concrete mixtures with a percentage ranging from 20 to 45% of cement without metallic fibres were indicated, as the use of such fibres increases flexural strength [55]. Thus, for a fair comparison between soil-cement and concrete, it was decided to compare these two materials without the use of fibres. The selected articles and the relevant parameters are presented in Table 2.

Table 2. Summary of the literature review of the dosage and flexural strength of high-performance concrete specimens.

Reference	Cement (%)	Silica or Metakaolin (%)	Use of Superplasticizer (%)	Age (Days)	Flexural Strength by ASTM C293 Method (MPa)
Fares et al., 2020 [56]	21.4%	0	0.24%	7	4.25
	21.4%	0	0.24%	28	6
	21.4%	0	0.24%	90	7.8
Manigandan et al., 2021 [57]	21.4%	0	0	7	4.7
	21.4%	0	0	28	6
	21.4%	0	0	56	6.5
Hilles and Ziara, 2019 [58]	28.57%	0	2%	7	4.84
	28.57%	0	2%	28	6.35
Wang et al., 2022 [59]	34%	10% of silica fume	3.1%	14	17.46
	34%	10% of silica fume	3.1%	28	17.89
Aghayari et al., 2019 [60]	42.8%	6.4% of silica fume	3%	28	14.28

Table 2. Summary of the literature review of the dosage and flexural strength of high-performance concrete specimens.

Reference	Cement (%)	Silica or Metakaolin (%)	Use of Superplasticizer (%)	Age (Days)	Flexural Strength by ASTM C293 Method (MPa)
Fares et al., 2020 [56]	21.4%	0	0.24%	7	4.25
	21.4%	0	0.24%	28	6
	21.4%	0	0.24%	90	7.8
Manigandan et al., 2021 [57]	21.4%	0	0	7	4.7
	21.4%	0	0	28	6
	21.4%	0	0	56	6.5
Hilles and Ziara, 2019 [58]	28.57%	0	2%	7	4.84
	28.57%	0	2%	28	6.35
Wang et al., 2022 [59]	34%	10% of silica fume	3.1%	14	17.46
	34%	10% of silica fume	3.1%	28	17.89
Aghayari et al., 2019 [60]	42.8%	6.4% of silica fume	3%	28	14.28

3. Materials and Methods

HPSC is a product-process invention that was developed at the Universidade Federal dos Vales do Jequitinhonha e Mucuri (Federal University of the Jequitinhonha and Mucuri Valleys) (UFVJM), for which a patent has been applied for at the Instituto Nacional da Propriedade Industrial (Brazilian National Institute of Industrial Property) (INPI), reference number BR102023006072-2. This patent is currently under evaluation [61].

3.1. Phase 1–Production of the Specimens

Specimen production was carried out at UFVJM on the campus located in the city of Diamantina-MG (Brazil). The materials and utensils used in this phase were the following: Portland cement type CP III 32 from the company CSN (Companhia Siderúrgica Nacional—Sao Paulo, Brazil); a rectangular mould for pressing 1 kg masses; an oven, model AL-102-250, for the water absorption analysis; and a container for the curing process and water absorption analysis.

The material dosage recommendations of Carvalho et al. [17] used for HPSC production were mixtures with a cement content above 20% by weight and clay content of approximately 30% relative to the amount of cement. Following these guidelines, 23.5% (by weight) of cement and 76.5% of soil were used. The soil used was made up of 10.5% clay, 42.5% fine sand with a grain size of 0.075 to 0.3 mm, and 48.5% coarse sand with a grain size of 0.6 to 2.38 mm. With this dosage, the particle size distribution of the soil-cement mixture is shown in

Table 3. Granulometric analysis of the soil-cement mixture.

	% Material	% of Cumulative Retained Material	Sieve Opening (mm)	% of Actual Passing Material	% of Theoretical Passing Material *
Clay (grain size < 0.002 mm)	6.9	100	0.00005	0	0
Cement (grain size 0.002 to 0.06 mm) with uniform distribution	23.5	93.1	0.002	6.9	7.2
Fine sand (grain size 0.075 to 0.3 mm) with uniform distribution	32.5	69.6	0.075	30.4	30.4
Coarse sand (grain size 0.6 to 2.38 mm) with uniform distribution	37.1	37.1	0.6	62.9	62.8

* Equation used to calculate the percentage of passing material based on grain size. CPFT = cumulative percentage of particles smaller than diameter D. CPFT (%) = 100 $\frac{D^q-D_s^q}{D_L^q-D_s^q}$ (Funk and Dinger, 1994) [63], D = diameter of soil particles; $D_s$ = diameter of the smallest particle in the distribution (0.00005 mm); $D_L$ = diameter of the largest particle in the distribution (in this case, 2.38 mm); and q = the distribution coefficient or modulus. For this case study, q = 0.325 was used (Lopes et al., 2020) [38].

. The table presents the grain size range utilised for the cement (from 0.002 to 0.06 mm) [21] and for the clay component (< 0.002 mm) [62]. This grain size distribution fits the theoretical granulometric curve using the packing method [38], with a value of q = 0.325, and considering a maximum grain size of 2.38 mm and a minimum size of 0.00005 mm. This granulometric distribution was selected to achieve the maximum possible compaction of the grains in the mixture.

Based on the following concepts, the optimum compaction moisture content was determined for producing the specimens in this study. As seen in Table 1, the usual values of compaction moisture of soil-cement range from 9 to 13%. Therefore, to find out the optimal compaction moisture for the case study in this paper, samples were produced using the same production process and parameters as the study specimens and then compacted with 9.0, 10.5, 12, and 13.5% water. The aim was to find out which one leads to the highest dry density of the sample. It was observed that the highest dry density of the material was obtained at a compaction moisture content of 10.5%; therefore, this was the moisture content used in the study specimens. This moisture content is lower than the optimal moisture content used in other studies for the production of soil-cement blocks. In the study by Souza et al. (2008) [43], the optimal compaction moisture content was 12.0% for soil with 80% sand. The reason why the optimal compaction moisture in this study is lower is because the material used has a larger grain size and the compaction energy (application of a pressure of 10 MPa for 60 s) is higher.

The production process for the soil-cement specimens, shown in Figure 3, was as follows:

(1)

The soil and cement were mixed in a dry state until a homogeneous mass was obtained.

(2)

Water was added to obtain the optimal compaction moisture content (10.5%).

(3)

The material was mixed further with the application of shear forces to ensure that the clay contained in the soil was homogeneously mixed. The clay fraction has a high specific surface area which increases soil cohesion due to the attraction between particles through electrostatic forces [64]. The application of shear forces during the mixing process promotes the rupture of soil aggregates and, consequently, the individualisation of clay particles, obtaining a more homogeneous soil-cement mixture.

(4)

The moulds were filled with the homogeneous soil-cement mixture. Figure 3 shows the dimensions of the specimens produced according to the ASTMC293 standard (60 mm × 40 mm × 140 mm).

(5)

Vibration application and subsequent compaction were performed using a hydraulic press. The applied pressure was 10 MPa, as recommended by Fay et al. [31], for a minimum time of 60 s.

(6)

The specimens were placed on a bench for 48 h in a room with a controlled environment with an average temperature of 20 °C and humidity above 50%. This is the observed necessary time for the specimen to not undergo changes when submerged. After the 48 h, the specimens were subjected to the curing process by being submerged in water for 28 days. A longer curing period was chosen instead of the usual 7 days to ensure the correct hydration of the cement.

(7)

Subsequently, the specimens were placed in the laboratory at room temperature before being tested for strength and water absorption. Only the specimens that were tested at 7 days of age were submerged in water for fewer days in order to test their breakage.

3.2. Phase 2–Analysis of Flexural Strength and Water Absorption

A total of 12 HPSC specimens were tested for flexural strength, these being divided into three groups (4 specimens each) according to their age: 7, 28, and 240 days. In soil-cement, it has been observed that the strength is 79% higher on average when comparing specimens 240 days of age with those aged 28 days [43]. This increase is less significant for concrete, being around 20% [21]. This may be due to the aggregates in soil-cement being finer than in concrete and the material being compacted, in addition to having a lower hydration speed. Due to this phenomenon, it was decided to evaluate the strength of specimens at an age higher than 28 days in order to account for this gain in strength in the longer term.

The tests were performed with an electromechanical, micro-processing TIME Group Universal Testing Machine, Model WDW100EB, with a capacity of 100 kN (Figure 4 and Figure 5), in accordance with ASTM C293/C293M-16. The placement of the specimen before the start of the test is shown in Figure 6. As shown in Figure 7, the bending rupture of the specimen was located in the centre of the piece, the position for the theoretical calculation of flexural strength.

The test was performed such that an axial tensile force was applied on a standardised specimen, as shown in Figure 8. The specimen was placed on two supports and the force was applied in the centre of the specimen. The distance between the supports (l) was defined according to the ASTM C293 standard which recommends a distance of three times the height (h) of the specimen. Thus, the distance was stipulated as 120 mm, which corresponds to three times the height of the specimen. The applied load was measured instantaneously by a load cell in the test equipment.

According to ASTM C293, the flexural strength was obtained using Equation (1).

$${ \sigma }_f=\frac{3P_{f}l }{2b{h}^2}$$

(1)

where:

${\sigma }_f$ is the bending flexural strength (MPa);

$P_f$ is the maximum applied load indicated by the test machine (N);

$l$ is the span length (mm);

$b$ is the average width of the specimen, at fracture (mm);

$h$ is the average depth of the specimen at fracture (mm).

The flexural strength was evaluated from beam bending tests, the results of which were determined in terms of mean and standard deviation.

Subsequently, water absorption for specimens ruptured at 240 days was evaluated according to the NBR-13555 standard [29]. The specimens were submerged in water until saturation and then weighed to obtain their wet weight. They were then placed in an oven at 100 °C for 24 h to obtain their dry weight. The water absorption was calculated as the ratio measured in percentage between the weight of the water (weight of the wet sample minus weight of the dry sample) and the dry weight of the specimen. The evolution of the water absorption of the specimens as a function of age was not evaluated for this study because the objective of the analysis was only to verify if, in the long run, this type of material could perform better than a conventional soil-cement in terms of water absorption. The parameters of absorption and porosity are directly related, as the greater the water absorption, the higher the porosity of the concrete and/or soil-cement. Greater porosity allows the entry of more chemical agents that can affect the strength of the concrete and/or soil-cement in the long term.

4. Results and Discussion

The failures of the tested specimens had similar behaviour to that shown in Figure 9. The specimens failed quickly after initiating a deep crack oriented perpendicular to the direction of the tensile stress. This initial fissure was located on the lower face of the specimen. As the specimens did not have steel reinforcement, the rupture was of the brittle type, the specimen being divided into two halves of similar proportions, with the rupture plane parallel to the direction of load application.

The flexural strength values obtained according to ASTM C293 [28], and the water absorption values obtained according to the NBR-13555 standard are presented in

Table 4. Results of the flexural strength and water absorption tests of the specimens.

	Flexural Strength (MPa)			Water Absorption (%)
	7 days	28 days	240 days	240 days
	4.042	5.601	10.74	3.13
	4.281	6.577	12.23	3.33
	4.960	7.610	13.87	3.29
	4.750	7.141	12.52	3.22
$\mathrm{Average}: \overline{x}=\frac{\sum _{\mathrm{i}=1}^4{\mathrm{x}}_{\mathrm{i}}}{4}$	4.51	6.73	12.34	3.30
$\mathrm{Deviation}: \sigma =\sqrt{\frac{{\sum _{\mathrm{i}=1}^4{(\mathrm{x}}_{\mathrm{i}}-\overline{x })}^2}{3} }$	0.421	0.864	1.284	0.09

.

The results obtained in Table 4 show the following:

(1)

The flexural strengths at only 7 days of age already exceeded values of 2.5 MPa for the four specimens tested. As shown in the previous data from the literature review on the flexural strength of soil-cement (Table 1), the values at 28 days are typically below 2.5 MPa. These values can therefore be considered high-performance values for a soil-cement mixture [17].

(2)

On average, when the age of the specimens is 240 days, the flexural strength values increased by 83% compared to the results at 28 days of age. These values are similar to the compressive strength gains obtained in soil-cement blocks in other studies [43].

(3)

The water absorption values were below 8%, which enabled this soil-cement to be considered HPSC [17]. The deviation of the water absorption parameter was low when compared to its average, having a deviation of less than 5% of the mean value. This is due to the use of optimal compaction moisture (10.5%) and constant compaction pressure (10 MPa).

(4)

It has been observed in other studies [3] that a lower water absorption leads to a higher flexural strength. However, in this study, the variation in water absorption was minimal due to the consistent mixture type, compaction energy, and moisture content. Therefore, it can be considered that the variation in flexural strength may be attributed to other parameters such as dimensional error and/or imperfect sample homogeneity.

(5)

The flexural strength values of these specimens were compared with those of the studies found in the literature review (Section 2 of the present study). Looking at the values from the literature review of soil-cement specimens with specific flexural tests according to the ASTM C293 standard [28], only three studies yielded similar strengths [23,46,47]. However, even in these cases, strengths above 6 MPa at 28 days were not achieved, and the amount of cement used in many cases was much higher (up to 50% cement). It is worth noting that in this study, a strength of 6.73 MPa was achieved at 28 days with a cement content of 23.5%. When comparing these findings with other soil-cement studies, the strengths in the present study were also significantly higher, as in Otoko’s study [65], a strength of 1.21 MPa was obtained with a 12% cement mixture; Gutiérrez et al. [66] achieved 0.67 MPa using 9% cement, and Venkatarama et al. [27] obtained 1.22 MPa with a 12% cement content. This seems to be due to the fact that, in the present study, the specimens were produced according to all the appropriate parameters found in the literature mentioned in Section 2.2.

Comparative Analysis of the Flexural Strength of the High-Performance Soil-Cement and the High-Performance Concrete

The results of the present study (Table 4) can be compared with other studies on HPC found in the literature (Table 2). As shown in Figure 10, the HPC flexural strengths at 28 days are higher than those of all the HPSCs produced in this study, but only for cement percentages over 40%. When 28.57% cement was used in the study by Hilles and Ziara [58], this resistance was actually slightly lower than that obtained in the HPSC at 28 days with a 23.5% cement content, being 6.35 MPa versus an average of 6.73.

For this type of HPC, besides having a higher consumption in terms of mineral exploration (aggregates), it is usually also necessary to use chemical components such as superplasticisers and other binders such as silica. For HPC, flexural strength values above those of the HPSC in the present study are only achieved with high dosages of binders. In the study by Wang et al. [59], a mass of 44% binder was used (34% cement and 10% silica) to achieve a flexural strength of 17.46 MPa. In the production of HPSC for this study, the flexural strength of soil-cement with cement contents above 23.5% was not evaluated. Such research could be conducted in the future.

The results obtained show that HPSC can compare to HPC in terms of flexural strength when the following production parameters are adopted:

(1)

Compaction pressure of 10 MPa maintained for at least 60 s;

(2)

Use of 23.5% cement in a mixture with sandy soil and 10.5% clay;

(3)

Granulometry of the soil-cement mixture based on the packing method with a coefficient q = 0.325 [38];

(4)

Optimal compaction moisture content of 10.5%. This optimal moisture has been obtained for and is applicable to this type of mixture (sandy soil with 10.5% clay and for the compaction energy exerted under a pressure of 10 MPa maintained for a minimum time of 60 s).

Therefore, the HPSC could be a viable alternative to HPC in terms of flexural strength for consideration in future applications in construction.

5. Conclusions

In this study, a literature review on the relevant parameters for the production of soil-cement was performed, and suggested (1) a compaction pressure close to 10 MPa; (2) the optimum compaction moisture for that pressure (10.5%); (3) cement percentages above 20%; and (4) a clay content of approximately 30% relative to the amount of cement. The conducted literature review also revealed a profound lack of knowledge on HPSC, whereby only 3 studies were found, compared to 141,000 on HPC. No single study was found in which specimens were produced and tested according to all these parameter values for soil-cement.

Therefore, to determine the flexural strength that can be achieved with soil-cement, specimens were produced according to the appropriate parameters found in the literature and, more specifically, a cement content of 23.5% and a sandy soil with 10.5% clay. The specimens were tested at 7, 28, and 240 days for their flexural strengths, as per the ASTM C293 standard [28], and for water absorption following the NBR-13555 standard [29].

It was verified that the production of soil-cement according to these parameters enabled the achievement of an average flexural strength of 4.51 MPa at 7 days, 6.73 MPa at 28 days, and 12.34 MPa at 240 days. The typical values of flexural strength for soil-cement found in the literature are lower than 2.5 MPa at 28 days and in only two studies did they reach a value < 5.5 MPa, but with a cement content of 50%, which is more than double the content used in the present study (23.5%). The average water absorption at 240 days was 4.94%. According to these results, this soil-cement can be considered an HPSC, as defined in Carvalho et al. [17].

The mean flexural strength of the tested soil-cement specimens at 28 days of 6.73 MPa compares with and is even slightly higher than that of the HPC found in the literature (6.00 MPa for a 21.4% cement content and 6.35 MPa for a 28.57% cement content). This demonstrates the potential of the HPSC developed in this study to be used as a replacement for precast concrete production components such as floor slabs, paving blocks, and other construction elements that do not require steel reinforcement.

The produced HPSC has a flexural strength equivalent to HPC with lower amounts of cement and without the need for materials from mining, such as sand and gravel, or additives such as superplasticisers, which are adopted in some HPC mixtures. Therefore, HPSC can be considered more environmentally desirable than HPC as it can use part of the soil from on-site earthwork or foundations, which would otherwise be discarded as waste in a landfill, thus avoiding waste and transportation to another location and reducing the environmental impact. These reductions in the use of sand and gravel and its transport and the reduction in the excavated soil that goes to waste would, in turn, reduce costs. Furthermore, the idea of seeking new construction methods and techniques to mitigate environmental and economic impacts is always advisable.

6. Limitations and Recommendations for Future Research

One limitation of this study is the lack of microscopic and chemical evaluation to further understand the reasons behind the increase in flexural strength when optimising different production parameters. A microscopic and mineralogical analysis of the product would provide more information regarding the strength enhancement, which is likely due to improved adherence among the different particles that compose the soil-cement matrix.

As a future research recommendation, it is advisable to separately assess the influence of each production parameter, (1) compaction pressure; (2) compaction moisture content; (3) cement percentage; and (4) soil particle size distribution, on the flexural strength of soil-cement for more precise optimisation beyond the literature search already performed.

It would also be advisable to study the long-term performance of the HPSC.

Another area of research that could further expand the applications of soil-cement is the study of steel behaviour within the soil-cement matrix. It should be evaluated whether steel’s durability within the matrix is equal to or better than in concrete. If steel could be incorporated into soil-cement products, larger soil-cement construction elements could be produced.

It would also be of interest to conduct a deeper, quantitative analysis of the environmental impacts and costs of HPSC compared to HPC under different hypotheses of construction elements and transport distances. Notwithstanding that the HPSC could be used in other circumstances, it could be especially useful in remote areas or areas not easily accessible due to a lack of transport infrastructures as local soil can be used instead of transporting sand and gravel.