Failure Mechanisms of Granular Columns

Failure Mechanisms of Granular Columns


Ground improvement is one of the most useful technique adopted in past two to three decades in order to improve the poor condition of foundation soils. Construction of structures on such underlying soil is often challenging due to the problems associated with excessive settlement and low bearing capacity of the foundation soils. The increased cost of the conventional foundations and different environmental constraints, greatly encourage in-situ improvement of the weak soil deposit. Depending upon the type of soil, various ground improvement methods are used to improve the strength of the soft foundation soil, reduce the total and differential settlements of the structures, shorten the duration of construction, and reduce the construction costs. The choice of a particular ground improvement technique mainly depends on the type of soil and overlying structure. Furthermore, the economic feasibility and environmental acceptability of a particular ground improvement technique is also an important parameter.


Ground improvement

In practise various types of columnar inclusions are used such as stone / granular columns, sand compaction piles, lime or cement columns which are stiffer and stronger than existing soil. In case of the soft clay soil stone/granular columns are most useful method in comparison to the other available columnar systems. Granular columns are composed of the compacted sand or gravel installed in soft clay by full displacement methods.

Granular columns are used for the construction of various flexible rigid structures such as buildings, oil storage tank, embankment etc. (Murugesan and Rajagopal 2006, 2007, 2008; Gniel and Bouazza 2009, 2010; Ali et al. 2012; Shahu and Reddy 2014). Granular column reinforced ground behaves as a composite with higher strength and stiffness compared to the virgin soil (Priebe 1995; Alamgir et al. 1996; Murugesan and Rajagopal 2010). Since the major component of the columns is granular material with high permeability, the column can also accelerate consolidation process of the soft ground and consequently accelerate the strength gain of the surrounding soft soil. Thus, use of granular columns not only reduces the excessive settlement, improves the stability and increases the bearing capacity of soft foundation soil, but also increases the rate of consolidation and resistance to liquefaction (Balam and Booker 1981; Bouassida et al. 1995; Alamgir et al. 1996; Babu et al. 2013; Najjar 2013). In earlier days, the length of the column used to vary between 4 m to 10 m (Barksdale and Bachus 1983) but with the introduction of modern technologies granular column can be installed to larger depths ( 30m) (Black et al. 2007).


The amount of the soil replaced to construct the granular column is defined as the area replacement ratio (as), which is defined as ratio of the area of granular column (As) to the total area within the unit cell (A) i.e., the influence zone of the column. Optimum value of as is decided by considering the cost of project and the type of structure. In present practise the value of as ranges from 10% to 30% (Barksdale and Bachus 1983). For the construction of granular column well graded aggregates of size between 2 mm to 75 mm shall be used (IS: 15284-2003). In practise granular columns are usually installed in square or triangular pattern, the equilateral triangular pattern is more commonly used as it gives better coverage of the influence area.

In case of very soft clay, (cu<15 kPa) the installation of granular column is difficult owing to the loss of aggregates. Moreover, the load carrying capacity of granular column also depends on the lateral confinement provided by the surrounding soil (Hughes and Withers 1974; Hughes et al. 1975). Owing to the absence of confining pressure from the surrounding in very soft clays, the strength and stiffness of the granular column decreases. The discharge capacity of granular columns can also reduce due to clogging from the surrounding soft clay (Murugesan and Rajagopal 2008; Weber et al. 2010; Castro and Sagaseta 2011; Indraratna et al. 2012). In such scenarios encasement of the ordinary granular columns (OGC) with geosynthetic is a good option. Encasement of the granular columns provide an additional confinement and prevents the loss of aggregates. Geosynthetic encased granular columns (EGC) have higher strength and stiffness compared to OGC (Raithel et al. 2002; Murugesan and Rajagopal 2007, 2008). Encasement also helps to achieve higher degree of compaction in the column. Figure 1 compares the strength and stiffness of OGC and EGC with respect to virgin soil and the rigid compaction pile.

Load condition of granular column

Granular columns installed below an embankment are subjected to various loading conditions as shown in Figure 2. From the figure it can be seen that columns below the crest of the embankment are primarily subjected to vertical loading, whereas columns near the toe of the embankment are primarily subjected to shear loading (Mohapatra et al. 2016, 2017; Mohapatra and Rajagopal 2017). The granular columns need be designed for both vertical and shear load for better performance in the field. This will help to enhance the safety and serviceability of the super structures.


Mechanism of granular column subjected to vertical loading

Granular columns in soft soil act as semi-rigid piles with higher modulus (Mitchell 1981). When the granular column treated ground is subjected to vertical loading, the load is shared by the aggregate column and the surrounding soil. The granular columns carry higher percentage of the applied load owing to their higher modulus and thus reducing the load carried by the surrounding native soil. Due to the lesser confinement at top of the soil, bulging of granular column takes place. Bulging usually happens up to a depth of 4D (D=diameter of granular column) and maximum bulging usually occurs at a depth of 2D (Figure 3) (Hughes and Withers 1974; McKelvey et al. 2004; IS: 15284-2003; Black et al. 2006). According to Hughes et al. (1975) critical length of the column is the shortest column length which can carry the ultimate load regardless of the settlements. As the confinement increases with depth, so at depths greater than 4D, no significant bulging occurs in the granular column due to higher confining pressures. Due to the bulging, the granular column pushes the surrounding soil laterally which in turn increases the lateral stresses in the clay and provides an additional confinement for the column. The confinement provided by the surrounding soil increases the stiffness of the granular column due to which the overall load carrying capacity increases.

Most of the studies reported the bulging failure of granular column due to vertical loading (Hughes and Withers 1974; Bergado and Lam 1987; McKelvey et al. 2004; Ambily and Gandhi 2007). Ambily and Gandhi (2007) carried out laboratory experiments using unit cell tank and reported that, when the column area alone was loaded, failure was governed by bulging with maximum bulging occurring at a depth of about 0.5 times the diameter of granular column. Bergado and Lam (1987) observed from the full-scale load test, that the maximum bulging occurs at a depth of one pile diameter below the ground surface. From laboratory model test McKelvey et al. (2004) observed that bulging failure was significant in case of long columns, whereas for short columns punching failure was prominent. Wood et al. (2000) reported different types of failure in sand columns depending on the geometric configuration (Figure 4)

As discussed earlier, the OGC may not function well in the case of extremely soft foundation soils due to the lack of adequate confinement and contamination of aggregate by the clay soils. In such scenarios, the granular columns were observed to fail by excessive bulging and impede the drainage due to the contamination by surrounding clay soil. Chummar (2000) has highlighted the following drawbacks with the OGC foundation systems:

Limiting bearing capacity of the improved ground cannot be increased beyond 25 times the initial undrained strength of the clay.


Entire settlement is not eliminated.

In highly sensitive clays (even with cohesive strength higher than 15 kPa) vibro-floatation techniques may cause remoulding leading to significant reduction in the shear strength of the clay soil.

To overcome the above limitations of OGC, geosynthetic encasement is wrapped around the column in tubular form (Figure 5), which provides an additional confinement and prevents the contamination of aggregates by surrounding clay soil. The concept of encasing the granular column by wrapping with geotextile was proposed by Van Impe in the year 1985. When the undrained shear strength of soil is very less (cu < 15 kPa) the required additional confinement can be given by wrapping the granular column by a suitable geotextile.

When the geosynthetic encased granular column is subjected to compressive load the granular aggregate tends to bulge because of the dilating properties of the aggregates. This dilation is subdued by the geosynthetic which wraps around the column. The dilating aggregates cause hoop tension force in the geosynthetic which offers the all-round confinement to the granular column. This confinement is in addition to the passive pressure from the surrounding soft soil. As the geosynthetic confines the granular column, the bulging, which tends to occur on the top portion of the column is controlled and the stresses are redistributed to a deeper depth of the granular column. This makes the bulging also to be redistributed to lower levels depending upon the stiffness of the geosynthetic and the density of the aggregate. As the bulging is controlled, the surface settlements are reduced. The failure of the encased granular columns can be in three following possible modes.


Bursting of the encasement:

Due to the excessive load on the column, the developed hoop tension force may increase the tensile strength of the geosynthetic and leads to the tearing of the geosynthetic. This will promote excessive bulging in the portion where the geosynthetic has failed and hence leads to higher settlement at the top level of column.

Excessive bulging due to yielding of geosynthetic:

When the geosynthetic encasement is of high yielding nature the bulging in the granular column cannot be controlled effectively leading to settlement beyond the service criterion.

Punching failure:

When the encased granular columns are not resting on the firm stratum and are acting like floating pile, the column as a whole may sink in to the soft soil due to higher load levels.

Many studies have been conducted in the literature to understand the behaviour of granular columns subjected to vertical loading. (Prisco et al. 2006; Murugesan and Rajagopal 2006, 2007, 2010; Yoo and Kim 2009; Gniel and Bouazza 2009, 2010; Lo et al. 2010; Khabbazian et al. 2010; Pulko et al. 2011; Elsawy 2013; Ali et al. 2012, 2014; Keykhosropur et al. 2012; Dash and Bora 2013; Ghazavi and Afshar 2013; Almeida et al 2015). However, the granular columns located near the centreline of an embankment are only subjected to vertical loads, whereas the columns near the toe of the embankment are predominantly subjected to shear loads (Murugesan and Rajagopal 2008).


Ground improvement

Mechanism of granular column subjected to shear loading

Murugesan and Rajagopal (2008) have carried out laboratory tests using plane-strain tank and observed that EGC performed better than the OGC when subjected to shear loading. Abusharar and Han (2011) have carried out two-dimensional slope stability analysis of stone columns supported embankment using FLAC2D. From their study it was concluded that, shear failure is the most common mode of failure for granular columns.

Mohapatra et al. (2016) carried out several large direct shear tests on the granular columns with and without encasement in a shear box having plan size of 305mm × 305mm. To simulate the embankment loading, the tests were carried out at different normal pressure varying from 15 kPa to 75 kPa. Mohapatra et al (2017) carried out 3-dimensional (3D) numerical modelling of the direct shear test reported by Mohapatra et al. (2016) to understand the complete mechanism using FLAC3D (version 3.1). From the experimental and numerical study, it was observed that OGC undergo complete rupture failure along the shear plane. Similar failure mechanism was also observed from the numerical study (Mohapatra et al. 2017). Due to shear failure of OGC, top and bottom portion of the column gets separated, leading to reduction in vertical load carrying capacity of the system (soil +column). This may lead to generation of larger total and differential settlements of the super structures. Therefore, the columns are recommended to be encased with a geosynthetic encasement. In addition to the increase in the shear loading capacity, the geosynthetic encasement also prevents rupture failure of the granular columns.

Depending upon the modulus and strength of the geosynthetic encasement two modes of failure were observed in case of EGC. Encasements with high initial modulus and low tensile strength were found to shear completely along with the granular column. This is shown as Mode-1 failure in Figure 6(a). In case of encasement having higher initial modulus and higher tensile strength or lower initial modulus and high rupture strain Mode-2 failure was observed (Figure 6b). In case of Mode-2 failure, the encasement did not undergo any rupture. Since, the EGC did not shear completely in case of a Mode-2 failure, they can still act as vertical drains because of the continuity of flow path at large deformations. Figures 6 (c) and (d), show the schematic of Mode-1 and Mode-2 failure. The ability of a geosynthetic encasement to prevent shear failure of granular columns is important in scenarios, where the earthquake-induced liquefaction causes the foundation soil to undergo large deformations. In such events, the EGC will continue to drain the liquefied ground and restore the effective stresses rapidly after the earthquake, thereby preventing complete failure of the structure.


Figure 7 shows the mode of failure of a single granular column installed at the centre of shear box obtained from the numerical analysis (Mohapatra et al. 2017). They reported that the failure mechanism of EGC depends on the applied normal pressure (n). At low normal pressures (e.g. 15 kPa), the EGC was found to tilt like a rigid body owing to the lateral soil movement (Figure 7a). In field, such behaviour is expected for granular columns near toe of the embankment (Chen et al. 2015). With an increase in the normal pressure, the tilting of granular column is restricted and a distinct flexural deformation is observed along the predefined failure plane, Figure 7 (b). From the deformed shapes of EGC (Figure 6 and 7), it can be concluded that the geosynthetic encasement having higher initial modulus and higher tensile strength or lower initial modulus and high rupture strain holds the granular materials together even after large shear deformations.

Mechanism of granular column subjected to embankment loading

All the previous studies reported in the literature for granular columns are either restricted to pure vertical or pure shear loading conditions. These studies have shown the comparative performance of EGC with respect to OGC, which is an important factor while designing the granular column treated soft foundation soil below the embankment, but in actual field condition granular columns are subjected to combined action of vertical and shear loading (Figure 2). The scale of the test setup reported by Mohapatra et al. (2016, 2017) are small, the results cannot be directly extrapolated to the field conditions.

To understand the behaviour of the granular columns in actual field condition Almeida et al. (2015) carried out prototype scale study and observed four-fold reduction in lateral deformations at the toe of an embankment due to EGC in comparison to the unimproved soft ground. Chen et al. (2015) have carried out laboratory tests and three-dimensional (3D) numerical modelling to understand the governing mechanisms of embankment loading on soft soils reinforced with EGC and concluded that the encased columns undergo bending instead of shear failure. From the 3D, numerical modelling it was observed that the granular columns near the toe of embankments undergo larger lateral displacements in comparison to the columns closer to the centre line of the embankment.

Mohapatra and Rajagopal (2017) carried out 3D slope stability analysis of embankments supported on soft clay soils treated with OGC and EGC using FLAC3D (version 5.0). Analyses were performed by varying the diameter of the granular columns from 0.8 m to 1.4 m while keeping the spacing between the column same. The corresponding as value varies from about 8% to 25% for the aforementioned diameters of the granular columns installed in a square pattern. From Fig. 8 (a and b) it can be observed, that the large lateral deformations and ground heaving take place for smaller as value in case of OGC. It can also be observed, that the OGC near the centre line of embankment undergoes significant vertical compression, whereas columns near the toe undergo predominantly lateral deformations. In addition, it can be observed that with increase in as, the critical slip surface moves upwards. To prevent the excessive lateral deformation and shear failure of granular column the columns can be encased with geosynthetic encasement to increase their strength and stiffness. From Fig. 8 (c) it can be observed, that with the encasement of granular column the shear strength of the foundation soil increased considerably. The deformation of the embankment and heaving of foundation soil shown in Figs. 8 (b) and (c) illustrate that with lesser replacement ratios, the performance of EGC treated ground is comparable to that with larger replacement ratios of OGC treated ground. Higher value of the secant modulus of geosynthetic encasement (J) found to prevent the deep seated failure and even the critical slip surface passed through the toe of the embankment (Mohapatra and Rajagopal 2017).


Concluding Remark

The present article briefly discusses the failure mechanism of granular columns subjected to different loading conditions. An effort has been made to explain the behaviour of ordinary granular column and geosynthetic encased granular columns subjected to vertical, shear and embankment loading using various analytical, laboratory, numerical and field studies. Interested readers can refer to the research papers cited in the present article for further reading and understanding.



  • Alamgir, M., Miura, N., Poorooshasbh, H.B. and Madhav, M.R. (1996). Deformation analysis of soft ground columnar inclusions. Com-puters and Geotechnics 18 (4), 261-290.
  • Ambily, A. P. and Gandhi, S. R. (2007). Behavior of stone columns based on experimental and FEM analysis. Journal of Geotechnical and Geoenvironmental Engineering, 133(4), 405-415.
  • Abusharar, S. W. and Han, J. (2011) Two-dimensional deep-seated slope stability analysis of embankments over granular column-improved soft clay. Engineering Geology, 120, 103–110.


Dr. Sunil Ranjan Mohapatra (Ph.D, IIT Madras)
Associate Professor, Department of Civil Engineering
K L University, Vaddeswaram, Guntur-522502


  1. Good article. Very useful and informative. If possible please also provide information regarding the settlement of floating stone column. Also provide information regarding the minimum initial modulus and strength characteristics of geosynthetic materials to be used for encasing the column subjected to different vertical pressures caused by embankment loading.