Final Report Summary - SOMEF (Critical State Soil Mechanics Revisited: Fabric Effects)
Soils are related to the built environment and civil infrastructure, constituting the foundation (buildings), the construction material (embankments) or the medium requiring retaining (walls). Thus, the mechanical response and failure characteristics of soils are of paramount importance for efficient and safe design of infrastructure. The predominant failure theory for soils of the last half century is Critical State Theory (CST) developed at Cambridge University, which provides two Critical State (CS) conditions, one on the ratio of shear to normal stress and a second on the relation between void ratio (measuring density) and mean pressure, that must be satisfied for CS to occur. Critical State implies failure, which means that soil continues to deform with no volume change.
However, besides stress and density there is another aspect of soil aggregates, not easily measured, that of fabric. In simple terms fabric addresses a feature of a granular or particulate aggregate related to the statistical orientation of directions associated with particles, such as the direction of normal to contact planes between particles, of long axes of particles or elongated voids formed in between the particles or grains. One may think of fabric (mathematically expressed by the so-called fabric tensor) as a measure of internal microstructural arrangements of an aggregate of particles, having an orientation distribution of appropriately defined directions and a norm measuring the intensity of such orientation distribution. Fabric naturally evolves during deformation all the way to CS failure as particles reorient themselves.
Yet, fabric is not included in the failure conditions of CST. This is exactly the main point that this project challenged, claiming that the classical CST is incomplete because it did not include a third condition related to fabric, in addition to the two conditions on stress ratio and void ratio. Such third condition, motivated by numerical simulations using the Discrete Element Method (DEM), was proposed and together with the previous two conditions, constituted the newly developed Anisotropic Critical State Theory (ACST). The DEM allowed an in-depth study of fabric characteristics and concluded that the fabric measure (fabric tensor) attains a specific value norm-wise and direction-wise at CS that in association with the direction of external loading provides the aforementioned third CS condition of the ACST. The DEM conclusions were confirmed by advanced X-Ray Computed Tomography experimental studies of granular samples during loading. Furthermore, it was possible to prove that this new third condition is absolutely necessary for the CS to be reached and maintained by performing a virtual DEM experiment on a sample at CS, where the new third condition was violated while maintaining the two classical CS conditions satisfied, and yet observing that CS is abandoned.
Within the framework of ACST constitutive models were developed for both sands and clays, namely mathematical relations between increments of stress and strain, necessary to describe the mechanical response of soils, which were able to simulate experimental results much better than models within the classical CST because of fabric effects that were included. Subsequently these constitutive models were implemented into numerical codes and used to analyze specific boundary value problems of geo-structures, where the importance of including fabric was demonstrated at that system level.
The impact of the main finding, namely the necessity to consider the effect of fabric at CS failure is enormous. It actually establishes a new paradigm, that of ACST that must replace the old paradigm of classical CST, and therefore it changes the way Critical State Soil Mechanics must be taught and applied in Universities and the profession.
However, besides stress and density there is another aspect of soil aggregates, not easily measured, that of fabric. In simple terms fabric addresses a feature of a granular or particulate aggregate related to the statistical orientation of directions associated with particles, such as the direction of normal to contact planes between particles, of long axes of particles or elongated voids formed in between the particles or grains. One may think of fabric (mathematically expressed by the so-called fabric tensor) as a measure of internal microstructural arrangements of an aggregate of particles, having an orientation distribution of appropriately defined directions and a norm measuring the intensity of such orientation distribution. Fabric naturally evolves during deformation all the way to CS failure as particles reorient themselves.
Yet, fabric is not included in the failure conditions of CST. This is exactly the main point that this project challenged, claiming that the classical CST is incomplete because it did not include a third condition related to fabric, in addition to the two conditions on stress ratio and void ratio. Such third condition, motivated by numerical simulations using the Discrete Element Method (DEM), was proposed and together with the previous two conditions, constituted the newly developed Anisotropic Critical State Theory (ACST). The DEM allowed an in-depth study of fabric characteristics and concluded that the fabric measure (fabric tensor) attains a specific value norm-wise and direction-wise at CS that in association with the direction of external loading provides the aforementioned third CS condition of the ACST. The DEM conclusions were confirmed by advanced X-Ray Computed Tomography experimental studies of granular samples during loading. Furthermore, it was possible to prove that this new third condition is absolutely necessary for the CS to be reached and maintained by performing a virtual DEM experiment on a sample at CS, where the new third condition was violated while maintaining the two classical CS conditions satisfied, and yet observing that CS is abandoned.
Within the framework of ACST constitutive models were developed for both sands and clays, namely mathematical relations between increments of stress and strain, necessary to describe the mechanical response of soils, which were able to simulate experimental results much better than models within the classical CST because of fabric effects that were included. Subsequently these constitutive models were implemented into numerical codes and used to analyze specific boundary value problems of geo-structures, where the importance of including fabric was demonstrated at that system level.
The impact of the main finding, namely the necessity to consider the effect of fabric at CS failure is enormous. It actually establishes a new paradigm, that of ACST that must replace the old paradigm of classical CST, and therefore it changes the way Critical State Soil Mechanics must be taught and applied in Universities and the profession.