Abstract
Despite extensive research on unsaturated soil behavior in recent decades, the field of transportation engineering still lacks technically and economically feasible methods for identifying expansive soils along roads, particularly in regions with limited laboratory infrastructure that are distant from developed areas in Brazil. This situation is particularly evident in the state of Acre, located in the southwest of the Brazilian Amazon, where the subgrade predominantly consists of fine, plastic soils and faces challenges of limited availability of crushed stone and high precipitation levels. The volumetric variation of expansive soils in Acre has resulted in significant financial losses and has affected the lives of the local population. This paper proposes a simplified method for the preliminary identification of expansive clays, which is based on geotechnical characterization tests that can be performed at local laboratories. The proposed method utilizes the correlation between the plasticity index (PI) and silt/clay content to determine ranges that express the highest probability of occurrence of soils susceptible to shrink-swell behavior. This method was developed by correlating geotechnical parameters of 321 samples from EMBRAPA's SiSolos database with a geotechnical database of 100 local samples.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Expansive soils affect engineering works across the globe, causing damage to structures and reducing their bearing capacity (Pooni et al. 2019). These soils undergo volumetric variation, swelling as they absorb moisture, usually due to rainfall, and subsequently contracting during evapotranspiration. The swelling is governed by the size and shape of particles, which depend on their mineralogy, degree of aggregation, and chemical interaction with adsorbed water (Terzaghi et al. 1996). Nelson and Miller (1992) proposed that expansion is a more complex phenomenon that essentially depends on three factors: soil properties, environmental conditions, and stress state.
In the Southwest of the Brazilian Amazon, studies on expansive soils for geotechnical purposes are still incipient, despite the recorded damages caused by these soils (Barbosa, 2018; Seixas 1997; Córdova 2011; Oliveira and Ferreira 2006). The problem is further compounded when it is necessary to conduct exploratory surveys for long stretches of roads or rail works. In such cases, direct approaches for identifying expansive soils, such as undisturbed samples and edometric tests, become impracticable and too expensive. Consequently, methodologies that disregard intrinsic aspects of local soils, such as the CBR method for expansion, are often used, which is unsuitable for identifying expansive soils.
Laboratory infrastructure in the region is still very limited and distant from the country's major centers, thus knowledge of subgrade soils is limited, which contributes to numerous problems associated with shrink-swell susceptible soils, as shown in Fig. 1. Another limitation is the high cost of crushed stone due to the absence of rock formations in the state (Barbosa, 2018). Crushed stone is essential for infrastructure works over expansive soil deposits, as it is used to reinforce the subgrade and in road drainage systems. This scenario highlights the need to make better use of scarce local resources, in addition to conducting more detailed studies of the expansive characteristics of soils.
To mitigate geotechnical risks in the Amazon region, it is crucial to establish a method for estimating the shrink-swell susceptibility of the local soils. This method should utilize simple tests commonly employed in geotechnical practice, allowing for easy and low-cost performance in soil laboratories, requiring minimal sample volumes, and enabling broad coverage at various depths. Moreover, the method should deliver reasonably precise outcomes across an extensive geographical range.
The use of Atterberg limits to classify soil shrink-swell susceptibility is a common approach. Researchers such as Seed et al. (1962) and Van der Merwe (1964) have proposed procedures to estimate soil expansion based on Atterberg Limits. They developed charts to classify soils into four expansiveness ranges using equations based on Atterberg limits and soil size.
Mckeen and Hamberg (1981) also presented a method to empirically obtain the suction compression index, γh, without conducting tests that involve the use of filter paper or psychrometers. In this method, γh can be obtained through a graphical procedure that relates different mineralogical groups to Skempton's activity, cation exchange capacity (CEC), and COLE expansion tests. Covar et al. (2001) improved this method using 6,500 soil samples from the United States Department of Agriculture (USDA).
The methods presented in Seed et al. (1962), Van der Merwe (1964), and McKeen and Hamberg (1981) should be used with caution. The direct application of soil classifications developed for different geological formations can lead to significant discrepancies in results. Therefore, it is necessary to develop a classification method based on the tests conducted in the specific region and that is also practical and easy to use due to logistical reasons.
The main objective of this research is to propose a method that mitigates geotechnical risks associated with expansive soils in pavements, by increasing their design life and reducing costs. The proposed method relies solely on conventional engineering tests related to particle size and Atterberg Limits.
2 Geotechnical Characterization
2.1 Localization
The studied region is in the state of Acre (152,581 km2), which is the westernmost region of the Brazilian Amazon and shares borders with Bolivia and Peru (Fig. 2). According to the Köppen Classification, the climate in Acre is of the AW type, which is a humid tropical climate. The average annual temperature is around 25 ºC, and the average annual precipitation varies from 1700 mm to 2000 mm (IBGE, 2005; Embrapa, 2017).
The subsoil in Acre is part of a sedimentary basin with a wide variety of soils (Fig. 2 and Table 1), whose genesis is directly related to the uplift of the Andes. About 80% of the state is located within the Solimões Formation, which has a high occurrence of silt–clay sediments composed of 2:1 clay minerals, resulting in poor drainage and a low leaching rate.
"Less weathered soils and a drier pedoclimate are predominant, which is an unexpected characteristic considering the local hot and humid climate, and differs from other soils in the Amazon basin (Amaral, 2007; SEMA, Acre, 2010). Since Brazilian deposits are mainly composed of typical tropical soils with low silt content, it might not be possible to simply extrapolate these parameters to the high silt content Acre soils (Kotlar et al. 2020).
From the Pedological Map of Acre, a predominance of argisols and cambisols can be observed. Despite the map's small scale, it is estimated that at least 70% of the area has a high potential for expansion, with a greater concentration in the central region, as evidenced by the frequent reconstruction of the BR-364 Highway.
2.2 Geotechnical Database
EMBRAPA is a federal government corporation that conducts agricultural research and maintains a database of thousands of soil profile samples from across Brazil. These samples were obtained through technical-scientific projects conducted by the institution's researchers and provide a detailed description of the morphological, physical, chemical, and mineralogical characteristics of the soil profiles. The resulting data are compiled in the Brazilian Soil Information System (SISolos), which is maintained by EMBRAPA (Embrapa, 2020).
For this study, we compiled all data related to the Solimões Formation in the state of Acre from 1975 to 2010. The samples were taken from soils of different pedological classifications, with a predominance of investigations conducted along the BR-364 road, the only east–west road segment in the state of Acre, as shown in Fig. 2.
Among the parameters conventionally employed by EMBRAPA researchers in the SISolos database, physical and chemical tests were used to identify the mineralogical composition of the clay fraction. Chemical data was utilized to analyze the weathering indice Ki and the CEC. Initially, 635 samples were selected for the study, but some of them lacked chemical attributes for correlation analysis. After conducting the data analysis, only 321 samples from Solimões Formation were retained, and approximately 99% of these samples had particles with an equivalent diameter < 2.0 mm, as demonstrated in Table 2.
2.3 Physical Properties
Physical indexes commonly used in geotechnical characterization are not typically utilized by agricultural researchers, resulting in the absence of Atterberg Limits in EMBRAPA's soil investigation routine. These indexes are fundamental to understanding expansive soil behavior; therefore, the physical indexes corresponding to the 321 samples compiled in Table 2 were inferred through correlation.
The physical indexes were obtained through linear regressions applied to a dataset of 100 soil tests conducted in Rio Branco by Borges (2019). Consequently, linear equations with high determination coefficients were obtained for the liquidity limit (LL), plasticity limit (PL), and plasticity index (PI), based on the clay fraction (%), as illustrated in Fig. 3. As expected, the linear regression analysis revealed a strong correlation between the < 2 µm fraction of soil particles and their plasticity, with a determination coefficient R2 = 0.94. Figure 3 displays the data of the linear regressions.
2.4 Weathering Index Ki
Concentration of oxides, or their interrelation, provides important indications for soil characteristics such as weathering and probable mineralogical composition. Smectites are formed in conditions where silica (SiO2) is abundant due to the 2:1 structure (silica:alumina), like the flocculated structures formed between silica and alumina. These structures are formed in soils with high pH, high electrolyte concentration, and higher concentrations of Mg2+ and Ca2+ than Na+ and K+. Additionally, climatic conditions with higher evapotranspiration than precipitation, and poorly drained sites with low leaching also contribute to smectite formation, which is the source of swelling in expansive soils (Mitchell, 2005).
To evaluate the soil's expansive behavior, it is necessary to distinguish highly weathered clayey soils from soils with more expansive clay minerals. Therefore, soils with a predominance of smectites in the clay fraction usually have a higher molecular ratio of SiO2 to Al2O3 greater than 2. On the other hand, values below 2 indicate that highly weathered soils, formed by kaolinite and iron and aluminum sesquioxides, are not related to expansive behavior (Mitchell, 2005; IBGE, 2005).
In this study, the weathering index Ki utilized was determined using Eq. 01 (IBGE, 2005), with the oxide content determined through the solubilization of soil samples with a 1:1 H2SO4 (sulfuric acid attack), as described by Embrapa (2017). Soils with a high degree of weathering have Ki values ≤ 2.0, while soils with a low degree of weathering have Ki values > 2.0 (Mitchell, 2005).
2.4.1 Cation Exchange Capacity (CEC) of the Clay Fraction
In general, the expansion potential increases with the total number of exchangeable cations required to balance negative charges on the surface of clay minerals, also known as cation exchange capacity (CEC). Higher CEC values indicate the presence of more reactive clay minerals, such as smectites, which are associated with higher soil activity. Conversely, lower CEC values indicate the presence of more stable clay minerals, such as kaolinite, which are associated with lower soil activity (Nelson et al. 2015).
The clay fraction cation exchange capacity (CEC) is defined as the CEC of soil particles divided by the clay fraction of the soil. CEC values ≥ 27 cmolc / kg of clay indicate high activity, while < 27 indicate low activity, according to the classification adopted by the USDA (1999) and EMBRAPA (2018). The CEC values used in this study correspond to the T-value, which is determined by the sum of exchangeable cations and the potential acidity (H+ + Al3+) (Embrapa, 2017).
3 Method for Assessing the Shrink-Swell Susceptibility of Studied Soils
3.1 Physical and Chemical Attributes
The main physical and chemical attributes of the Solimões Formation soils in the state of Acre state were selected and analyzed on box plots, using the procedures presented in item 2. Figure 4a shows a predominance of the silt and clay fractions, with fractions varying between 20 and 50%, respectively, from the first to the third quartile. Granulometry is one of the factors that most influence the physical–chemical factors related to volumetric variations in soils (Mitchell, 2005), thus it is the first to be analyzed. From these results, it can be deduced that the occurrence of expansive soils in the region is possible since expansive behavior is associated with decreasing particle sizes.
The Atterberg Limits were analyzed in Fig. 4b, where LL varied between 28 and 48%, from the 1st to the 3rd quartile; PL had a smaller range, with a median value of 19%; and high plasticity, with PI varying between 11 and 27%, and a median of 19%. Soils with higher plasticity tend to have more expansive clay minerals (Nelson et al. 2015).
The parameters related to soil mineralogy are presented in Fig. 5. The median Ki value is 2.7 and the median CEC value is 47 cmolc / kg, both of which are associated with soils containing expansive clay minerals.
The results presented in Figs. 4 and 5 indicate a high shrink-swell susceptibility of the Solimões Formation soils samples from Acre, which is consistent with field observations. Correlating these parameters is necessary to determine the range that indicates expansive soils.
3.2 Correlations Analysis of Data
After identifying the predominance of silt and clay fractions in the soils, the granulometry was correlated with Ki and CEC parameters. Subsequently, the mineralogical relationships were expressed in terms of parameters routinely used in transportation engineering works, such as the Plasticity Index. Figure 6 illustrates the relationship between silt/clay ratio and plasticity index, with data sorted according to the mineralogical divisions presented in Sects. 2.4 and 2.5.
The presented data in Fig. 6 demonstrates a noticeable trend, particularly in the range indicating the prevalence of expansive clay minerals. According to Table 3, the exponential regression model exhibits the best fit for the given data.
3.3 Chart for the Assessment of the Shrink-Swell Susceptibility of Soils in the Solimões Formation in the State of Acre
Using the models presented in Table 3, a 95% confidence interval was applied to determine the lower and upper limits of the range for shrink-swell susceptibility of soils in the Solimões Formation, considering the clay fraction Ki and CEC attributes (Fig. 7). The two limits were then combined to obtain a region with a higher probability of occurrence of expansive soils, and the resulting chart is presented in Fig. 8. All analyses were performed using MATLAB and its Curve Fitting Toolbox.
3.4 Evaluation of the Shrink-Swell Susceptibility of Samples From Brazil.
After developing the graph depicted in Fig. 8, the subsequent step involved comparing the proposed method with established procedures documented in the literature, utilizing additional tropical Brazilian soils associated with expansive behavior.
Figures 9, 10, and 11 illustrate the soil swelling susceptibility of 37 Brazilian samples using different methods. Table 4 summarizes the data of the 37 samples utilized, comparing the swelling susceptibility obtained in this study with that of the Van der Merwe (1964) and Seed et al. (1962) methods. The first 27 samples investigated in Seixas (1997) and Barbosa (2022) relate to the subgrade of the Rio Branco International Airport and its surroundings, which exhibit an expansive nature that causes the frequent occurrence of ripples and cracks in nearby pavements.
Van der Merwe (1964) classification chart for the swelling susceptibility of 37 Brazilian soil samples with a history of failures related to expansive soils
Seed et al. (1962) classification chart for the swelling susceptibility of 37 Brazilian soil samples with a history of failures related to expansive soils
There is a correspondence between the proposed method and the charts widely used in the literature to assess the susceptibility of soil samples to expansion, particularly for samples from the Solimões Formation, ranging from 01 to 22. Additionally, it is worth noting that the graph proposed in this research better identifies the expansive behavior in soils with high silt/clay ratios, as evidenced by samples 28, 35, 36, and 37. This finding suggests the high cation exchange capacity in these soils, despite their lower clay fraction when compared to other deposits.
3.5 Flowchart of the Shrink-Swell Susceptibility Identification Method
The flowchart of the procedures required to determine the degree of soil sample's swelling susceptibility is illustrated in Fig. 12, which employs the graph depicted in Fig. 8.
4 Conclusions
A broad database was utilized in this study to devise a graphical method for the initial assessment of soil shrink-swell susceptibility in the Solimões formation in Acre. The implementation of agricultural tests has been demonstrated to be technically suitable, which is critical for transportation engineering in the region, where limited geotechnical data is accessible.
The chart provides the benefit of using low-cost and simple geotechnical tests commonly employed in transportation geotechnics, such as the Atterberg Limits and complete granulometric analysis, including sedimentation. Moreover, it is applicable to a broad spectrum of soils found in Brazil.
Due to the high silt content in soils of Acre, correlations based on granulometric attributes have demonstrated greater statistical significance with the silt/clay ratio content compared to the percentage of clay. This methodology could potentially be extended to forecast expansive behavior in other regions of Brazil, particularly those possessing similar physical–chemical characteristics.
It is important to note that although the chart provides an indication of the swelling susceptibility of soils in the state of Acre, its use does not provide a definitive identification of expansive soils. Therefore, it is recommended for preliminary studies and to plan more detailed tests for infrastructure projects. It could be a useful tool for construction managers to identify and investigate the occurrence of expansive soils at specific locations.
Data Availability
The authors declare that the data supporting the findings of this study are available within the paper.
References
Amaral EF. Stratification of environments for environmental management and the transfer of the knowledge in the State of Acre, Western Amazonia (2007) 185 p. Tese de Doutorado, UFV, Viçosa, MG
Barbosa VHR (2022) Proposal for assessment of expansive soils susceptibility: applications in pavements in the State of Acre. Western Amazônia, Tese de Doutorado, IME, Rio de Janeiro, RJ, Brazil, p 197
Barbosa VHR, Marques MES, Guimarães ACR (2018) Mineralogical characterization of an Acre soil in order to produce calcined clay aggregates. Revista Matéria, Rio de Janeiro
Borges MV (2019) Geotechnical Cartography of the city of Rio Branco - Acre: use of descriptive statistics, geostatistics and multivariate statistics. Universidade Federal de Brasília, Brasília, Tese de Doutorado (in Portuguese)
Córdova HG (2011) Study and Evaluation of the Runway Pavement Structure of the airport of Rio Branco, State of Acre. Brazil, Dissertação de Mestrado, IME, Rio de Janeiro, RJ, Brazil (in Portuguese)
Covar AP, Lytton RL. Estimating Soil Swelling Behavior Using Soil Classification Properties. Expansive Clay Soils and Vegetative Influence on Shallow Foundations. Proceedings of Geo-Institute Shallow Foundation and Soil Properties Committee Sessions at the ASCE 2001 Civil Engineering Conference. Geotechnical Special Publication 115. ASCE, Houston, Tex., 2001, pp. 44–63.
EMBRAPA – EMPRESA BRASILEIRA DE PESQUISA AGROPECUÁRIA. Brazilian Soil Information System, SISolos. Disponível em. Acesso em: 01/03/2020 (in Portuguese)
EMBRAPA – EMPRESA BRASILEIRA DE PESQUISA AGROPECUÁRIA. Brazilian Soil Classification System. 5ª Ed., Brasília, DF, 2018 in Portuguese.
EMBRAPA – EMPRESA BRASILEIRA DE PESQUISA AGROPECUÁRIA. 2017 Manual of Methods of soils analysis. Centro Nacional de Pesquisa de Solos, Brasília, DF, pp 577
Fredlund DG, Ahmad N, Mermut A. Geotechnical problems associated with swelling clays. Vertisols and technologies for their management., c. 14, p. 499–524, 1996.
IBGE – INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 2005, Technical Manual on Geosciences - Pedology, Manuais Técnicos em Geociências, Rio de Janeiro, 2 ed., n.4, 316 p. Brazil in Portuguese.
Kotlar AM, Van Lier QJ, Brito ES. Pedotransfer functions for water contents at specific pressure heads of silty soils from Amazon rainforest. Geoderma, 361 (2020).
McKeen RG, Hamberg DJ (1981) Characterization of expansive soils. Trans Res Rec 790:73–78
Merwe DHV. The Prediction of Heave from the Plasticity index and percentage clay fraction of soils. Transactions of the South African Institution of Civil Engineers, 6, June 1964.
Mitchell JK, Soga K (2005) Fundamentals of Soil Behavior, 3rd edn. John Wiley & Sons Inc, Hoboken, New Jersey
Morais JJO. Expansion evaluation of a clayey soil of the municipality of Paulista-PE. Dissertação de Mestrado, UFPE, Recife, PE, Brazil, 2017 (in Portuguese).
Nelson JD, Miller DJ (1992) Expansive Soils, Problems and Practice in Foundation and Pavement Engineering. John Wiley & Sons Inc, New York
Nelson JD, Chao KC, Overton DD, Nelson EJ (2015) Foundation Engineering for Expansive Soils, Problems and Practice in Foundation and Pavement Engineering. John Wiley & Sons Inc, Hoboken, New Jersey
Oliveira MA, Ferreira AL. 2006 Geological and Geotechnical Assessment Report - City of Rio Branco, state of Acre, Brazil. In: Relatório Final Ministério de Minas e Energia, Manaus, AM
Oliveira AGS, Jesus AC, Miranda SB. Geological - Geotechnical Study of Expansive Soils in northeastern Brazil (Bahia state). II Simpósio Brasileiro de Jovens Geotécnicos - II Geojovem, UERJ, Nova Friburgo, RJ, Brazil, 2006 (in Portuguese).
Pereira EM (2004) Swelling tests on material from the Guabirotuba Formation, Parana, Brazil, with suction control. Tese de Doutorado, Escola de Engenharia de São Carlos, São Carlos, SP, Brazil
Pooni J, Giustozzi F, Robert D, Setunge S, O’Donnell B (2019) Durability of enzyme stabilized expansive soil in road pavements subjected to moisture degradation. Transp Geotech 21:100255. https://doi.org/10.1016/j.trgeo.2019.100255
Ribeiro Júnior I, Futai MM, Conciani W (2006) Investigation of a design solution for foundations on expansive soils. Anais do XIII Cobramseg, ABMS, Curitiba, PR, Brazil
Seed HB, Woodward RJ Jr, Lundgren R (1962) Prediction of swelling potential for compacted clays. J Soil Mech Foundation Div, ASCE 88(SM3):53–87
Seixas S (1997) Dinamic Behavior of the Runway Pavement Materials of the new Airport of Rio Branco. Dissertação de Mestrado, COOPE/UFRJ, Rio de Janeiro, RJ, Brazil
SEMA – Secretaria de Estado de Meio Ambiente, Geology, Geomorphology and Soils of Acre. Livro temático II: Recursos Naturais I ‑ Programa Estadual de Zoneamento ecológico‑econômico do Acre Fase II ‑ Escala 1:250.000. Rio Branco, Acre, 2010 (in Portuguese).
Soares JMD, Pires GM, Conterato TM (2013) Expansive Soils: Case Study in Santa Maria. RS, GEORS, Santa Maria, RS, Brazil
Souza LS. Analysis of mechanical behavior of an expansive soil from the oil province of Urucu-AM for paving. Dissertação de Mestrado, UFG, Goânia, GO, Brazil, 2009 (in Portuguese).
Terzaghi K, Peck RB, Mesri G (1996) Soil Mechanics in Engineering Practice, 3rd edn. John Wiley & Sons Inc, New York
USDA. Soil Survey Staff. 1999. Soil taxonomy: A basic system of soil classification for making and interpreting soil surveys. 2nd edition. Natural Resources Conservation Service. U.S. Department of Agriculture Handbook 436.
Funding
No funding was received to assist with the preparation of this manuscript.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Barbosa, V.H.R., Marques, M.E.S. & Guimarães, A.C.R. Predicting Soil Swelling Potential Using Soil Classification Properties. Geotech Geol Eng 41, 4445–4457 (2023). https://doi.org/10.1007/s10706-023-02525-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10706-023-02525-2