1Senior Researcher

Technical Note

Soil Compressibility Models for a Wide Stress Range Song-Hun Chong1 and J. Carlos Santamarina2

Abstract: Soil compressibility models with physically correct asymptotic void ratios are required to analyze situations that involve a wide stress range. Previously suggested models and other functions are adapted to satisfy asymptotic void ratios at low and high stress levels; all updated models involve four parameters. Compiled consolidation data for remolded and natural clays are used to test the models and to develop correlations between model parameters and index properties. Models can adequately fit soil compression data for a wide range of stresses and soil types; in particular, models that involve the power of the stress σ 0β display higher flexibility to capture the brittle response of some natural soils. The use of a single continuous function avoids numerical discontinuities or the need for ad hoc procedures to determine the yield stress. The tangent stiffness—readily computed for all models—should not be mistaken for the small-strain constant-fabric stiffness. DOI: 10.1061/(ASCE)GT.1943-5606.0001482. © 2016 American Society of Civil Engineers.

Author keywords: Soil compressibility models; Compression index; Remolded clays; Natural sedimentary clays; Tangential stiffness; Small-strain stiffness; Yield stress.

Introduction

Soils subjected to either isotropic compression or ko (where ko is under zero lateral strain condition) compression experience volume contraction. Contraction depends on soil type, formation history, diagenesis, prior stress history, porosity, and stress conditions.

A soil compressibility model used for settlement analysis often needs to justify the data in a relatively narrow stress range. How- ever, many geotechnical problems involve soils subjected to either extremely low or extremely high effective stress or a wide effective stress range. Examples include self-weight consolidation (Been and Sills 1981; Cargill 1984; Bartholomeeusen et al. 2002; Stark et al. 2005), seafloor engineering [suction casings (Houlsby et al. 2005) and skirted foundation (Bransby and Randolph 1998)], gradual movement of pipelines resting on the seafloor and lakebeds (Krost et al. 2011; Randolph et al. 2011), pile tips (Yang et al. 2010; Tsuha et al. 2012), blast loads (Wang et al. 2005), methane recovery by depressurization from hydrate bearing sediments, and filter-cake formation in drilling mud (Sherwood and Meeten 1997).

Soil compressibility models are sought in this study to analyze field conditions over a wide stress range. Models must be able to fit compressibility data for diverse soils, have physically correct asymptotic values at low stress σ 0 → 0 and high stresses σ 0 → ∞, and involve a small number of physically meaningful parameters (Ockham’s criterion).

Compressibility: Stress Regimes

Soil compressibility is briefly reviewed next. Three stress regimes are tentatively identified in reference to the standard stress

range in common geotechnical applications, namely 10 kPa < σ 0z < 1 MPa.

Low Stress Regime (σ 0z < 10 kPa)

Individual grains or flocs form a granular skeleton with a characteristic finite porosity that depends on grain geometry and pore fluid chemistry (Klein and Santamarina 2005; Palomino and Santamarina 2005). Compressibility at low stress reflects the for- mation fabric and postdepositional diagenetic changes triggered by preloading, moisture fluctuations, thermal history, fluid-mineral in- teraction, dissolution, and reprecipitation (Mitchell 1956; Burland 1990; Santamarina et al. 2001; Rinaldi and Santamarina 2008).

Intermediate Stress Regime (10 kPa < σ 0z < 1 MPa)

Soil compression in this stress regime remains affected by forma- tion conditions such as initial water content (Hong et al. 2010, 2012) and temperature (Campanella and Mitchell 1968; Baldi et al. 1988; Leroueil 1996; Sultan et al. 2002), and is affected by diage- netic processes such as cementation and aging (Mesri et al. 1975; Schmertmann 1983, 1984, 1991). This stress regime is of main in- terest to classical geotechnical practice; therefore, many studies have explored correlations between compressibility and index properties. In particular, the compression index Cc has a strong cor- relation with the liquid limit LL, or the void ratio at the liquid limit eLL (Skempton 1944; Terzaghi and Peck 1948; Burland 1990; Sridharan and Nagaraj 2000). Disturbance and/or remolding de- structures natural soils, and remolded soils exhibit a compression curve that plots at a lower void ratio than the undisturbed natural soil and with a less pronounced yield stress. Measured consolida- tion curves are affected by experimental procedures such as sam- pling disturbance (Casagrande 1936; Terzaghi and Peck 1948; Schmertmann 1955; Rochelle et al. 1981; Hight et al. 1992; Santagata and Germaine 2002), seating and boundary effects, and strain rate (Hanzawa 1989; Leroueil 1996; Leoni et al. 2008).

High Stress Regime (σ 0z > 1 MPa)

The void ratio decreases at a gradually lower rate at high stress (Athy 1930; Aplin et al. 1995), and the prevailing deformation mechanisms become particle compliance, pressure dissolution,

1Senior Researcher, High Speed Railroad Systems Research Center, Korea Railroad Research Institute, 176, Cheoldo bangmulgwan-ro, Uiwang-si, Gyeonggi-do 437-757, Republic of Korea (corresponding author). E-mail: [email protected]

2Professor, Earth Science and Engineering, King Abdullah Univ. of Science and Technology, Bldg. 5, Thuwal, Saudi Arabia 23955-6900.

Note. This manuscript was submitted on May 18, 2015; approved on December 3, 2015; published online on March 3, 2016. Discussion period open until August 3, 2016; separate discussions must be submitted for in- dividual papers. This technical note is part of the Journal of Geotechnical and Geoenvironmental Engineering, © ASCE, ISSN 1090-0241.

© ASCE 06016003-1 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 06016003