Importance of volumetric shrinkage curve (VSC) for determination of soil–water retention curve (SWRC) for low plastic natural soils

Highlights

• Influence of no volume change assumption on the SWRC for low plastic soils is evaluated.

• Low plastics soils undergo significant amount of volume change (about 60% of total volume) during drying.

• A new methodology is proposed to determine unambiguous SWRC irrespective of the initial water content.

• Consideration of VSC becomes mandatory after soil plasticity and clay content above 15 and 11%, respectively.

Abstract

The study of unsaturated behavior of soils pivots around the relationship between soil suction and the amount of water content present in the soil mass, which is termed as soil–water retention curve (SWRC). The drying SWRC is generally determined by neglecting the volume change behavior of the soil for low plastic soils. Such a procedure is bound by an inherent assumption that the soil exhibit zero or negligible volume change and can lead to approximations in important SWRC parameters, including saturated volumetric water content, residual water content, and fitting parameters, which are essential for the estimation of unsaturated hydraulic characteristics of the soil. This study deals with the determination of SWRC parameters for a range of soils starting from non-plastic to high plastic soils, with and without considering volume change characteristics. For this purpose, the SWRC of six different natural soils were experimentally measured using a miniature tensiometer (T5) and a dew point potentiometer (WP4C), while the VSC of the used soils were measured using the balloon method. Efforts were made to critically understand the influence of no volume change assumption on the SWRC parameters, unsaturated hydraulic conductivity function, and flow of water through 1D soil column. The experimental results indicated that the no volume change assumption underpredicted the water retention curves of low plastic soils, and this error in SWRC increases with the soil plasticity. The measured VSC of the used soils further revealed that there could be a range of saturated volumetric water content (VWC) depending on the initial state of the soil sample, which also influences the sensitivity of the SWRC parameters. Therefore, the present study proposed a new methodology to determine unambiguous SWRC (in terms of VWC-suction) for a particular soil type combining both the measured VSC data and SWRC data. Based on the obtained results, it was recommended that the details of VSC must be incorporated along with the measured SWRC data for soils having plasticity and clay content higher than 15 and 11%, respectively.

Introduction

Understanding the basic principles of unsaturated soil mechanics is extremely important for finding solutions to the wide range of geoenvironmental problems associated with the unsaturated slope instability, rainfall infiltration (Kim et al., 2004, Yang et al., 2006, Gavin and Xue, 2008, Lee et al., 2011), waste containment facilities or nuclear waste disposal (Miller et al., 2000, Durmusoglu et al., 2006, Ng et al., 2016, Rahardjo et al., 2016), pavement engineering projects (Yang et al., 2008, Cary and Zapata, 2014, Turco et al., 2017) and solute transport through the vadose zone (Mohamed et al., 2000, Malaya and Sreedeep, 2012). Any study related to unsaturated soil mechanics necessitates the soil–water retention curve (SWRC) as the primary input, which is a constitutive relationship between soil suction or negative pore water pressure and soil moisture content (presented in terms of either gravimetric or volumetric or degree of saturation). The SWRC parameters are very important for the estimation of unsaturated soil properties, such as shear strength (Fredlund et al., 1996, Zhai et al., 2019, Zhai et al., 2020a, Zhai et al., 2020b) and unsaturated hydraulic conductivity (van Genuchten, 1980, Mbonimpa et al., 2006, Ghanbarian et al., 2015, Haghverdi et al., 2018). Therefore, proper understanding and determination of the SWRC parameters are most important (Zhai and Rahardjo, 2012). One of the major issues while determining the SWRC parameters is that the soil is often considered as rigid mass (i.e., ignoring the volume change during drying-wetting), especially for low to medium plastic soils.

The measurement of void ratio (e) of the soil sample (during drying process) is required to incorporate the influence of volume change on the SWRC. Such an information can be obtained from the volumetric shrinkage curve (VSC), which is a graphical representation of volume of the sample (or void ratio) and the corresponding water content (gravimetric or volumetric) during different stages of drying (Bensallam et al., 2012, Wijaya et al., 2015). A typical representation of VSC was presented in Fig. 1, which consists of four distinguished phases, including structural shrinkage, normal shrinkage, residual shrinkage, and zero shrinkage (from the wet side to dry side). In the structural shrinkage and residual shrinkage, the reduction of sample volume is less than the amount of water loss. The water loss in the structural shrinkage phase is governed by the extraction of free water localized in the inter-aggregate pores, whereas the water extraction occurs from the intra-aggregate pores in the residual shrinkage phase (Bensallam et al., 2012). The intersection point between normal shrinkage and residual shrinkage is known as air entry point (AEP) because the desaturation of soil sample starts after this point (Cornelis et al., 2006). The soil volume remains almost constant in the zero shrinkage phase, and the water content at which soil volume becomes constant is known as the shrinkage limit.

A precise and simultaneous measurement of soil bulk volume over the whole range of water content is very important for an accurate determination of VSC. Past studies (Péron et al., 2007, Tripathy et al., 2014, Gapak et al., 2017) have highlighted different laboratory methods to measure soil volume using liquid displacement methods (following the Archimedes’ law) with the effective utilization of surface coating materials (such as adhesive, resin, glue, and wax) [Krosley et al., 2003, de Almeida et al., 2009, ASTM D7263-09, 2018]. The disadvantage of this method is that the coating materials do not make proper contact with the soil clod for most of the occasions, which may result in water penetration or the presence of air bubbles inside the coating layer. There are different direct measurement methods available for measuring the soil volume using vernier calipers or using displacement transducers (Berndt and Coughlan, 1976, Braudeau et al., 1999, Birle et al., 2008, Peron et al., 2009), and by digital image analysis (Julina and Thyagaraj, 2018, Li et al., 2019, Amenuvor et al., 2020). However, comparative studies have shown that the balloon method, as proposed by Tariq and Durnford (1993), is one of the most accurate methods for the measurement of both soil volume and water content using a single sample (Cornelis et al., 2006). There are not many studies present in the literature that have effectively used the balloon method to incorporate the effect of volume change during drying on the SWRC of low plastic natural soils.

The concept of SWRC was first developed for soil physics and agriculture-related discipline to quantify the amount of water storage in the soil at different suction. In those cases, a simple volumetric water content (VWC) versus suction plot with an assumption that there is no considerable volume change in soil with increasing suction can serve the purpose (Solone et al., 2012, Fredlund and Houston, 2013). The application of SWRC is quite different in subsurface environmental problems involving the fluid flow and solute transport through the vadose zone, and hence the assumption of no volume change during drying in low plastic soil may not be valid for many of the cases, where a considerable amount of volume change occurs but conveniently neglected. There are a lot of studies present in the literature that discussed the importance of considering volume change for high plastic clay and expansive soils, which undergo significant volume change during the drying or wetting stage for developing SWRC (Pham and Fredlund, 2008, Qi and Michel, 2011, Fredlund and Houston, 2013, Tripathy et al., 2014, Wijaya et al., 2015, Lu and Dong, 2017, Zhai et al., 2020c). These studies clearly indicated that ignoring the volume change of soil during drying leads to an inappropriate determination of SWRC for soils with high plasticity. However, it is also very true that even low plastic soils can undergo a considerable amount of volume change, which may have a significant effect on SWRC and its parameters. It was noted from the previous studies that the sensitivity of volume change on SWRC parameters and unsaturated soil properties had not been properly quantified for low plastic soils. Moreover, there has to be a proper guideline for low plastic soils to determine unambiguous SWRC by taking the volume change consideration that can be used for an accurate estimation of unsaturated soil properties.

The objective of this study is to investigate the importance of VSC for SWRC determination in low to medium plastic soils. Efforts were made to evaluate the range of plasticity and clay content, beyond which the consideration of VSC data becomes mandatory for the SWRC determination. For this purpose, six different soils with a wide range of plasticity starting from non-plastic to high plastic soils were collected. The SWRC of the selected soils were experimentally measured using a miniature tensiometer (for the initial portion of drying) and a dew point potentiometer, WP4C (for the higher suction range), while the VSC of the selected soils were measured using the balloon method. The measured VSC was used to quantify the error in the determination of SWRC and its parameters due to the no volume change assumption for low plastic soils. The importance of VSC on the estimated unsaturated soil properties was further illustrated by simulating one-dimensional flow of water through a 1 m soil column using HYDRUS-1D. Based on the observed results, an attempt was made to evolve a guideline on using VSC for an accurate determination of SWRC.