Use this URL to cite or link to this record in EThOS:
Title: Development of improved predictive tools for mechanical soil root interaction
Author: Duckett, Natasha-Rass
ISNI:       0000 0004 5352 0763
Awarding Body: University of Dundee
Current Institution: University of Dundee
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Access from Institution:
Plant roots can stabilise a soil through two key mechanisms, namely: mechanical interaction and suction (drawing water from the soil). As a result, they offer a sustainable alternative to traditional soil stabilisation techniques, such as soil nailing or piling, and are becoming increasingly sought after as concerns for global climate change increase. In practise, however, use of vegetation in infrastructure (termed bioengineering) is little used due to a lack of understanding of root functions and, therefore, a lack of confidence from engineers. Predicting the response of soil root systems to mechanical loading is therefore of significant importance in the development and improved use of bioengineering techniques. Previous research has primarily focused on predicting a root cohesion factor, which can be used to estimate the ultimate limit state but not pre-failure behaviour (e.g Wu et al., 1974). This Thesis reports an extensive series of laboratory uprooting1 and shear box tests, which were carried out to quantify soil root interactions, and to provide a database of results to develop and test predictive numerical models. The laboratory tests used root analogues made from either rubber or wood, to span a wide range of root stiffness, whilst avoiding the natural variability associated with plant roots. These included full section-centre tests, where the roots were located in the centre of the soil sample, and novel cross section-front tests, where the roots were halved along their length and placed at the edge of the soil sample to provide a window into the system during loading. The latter allowed the soil and root deformation to be digitally photographed during loading and thus the displacement fields to be measured using GeoPIV analysis, a computer program designed to trace the movement of pixels through a series of digital images. With such data, the forces acting within the roots could be assessed during loading and the interface friction between the root and the soil could be quantified for input into the numerical models (using t-z and p-y pile analysis theory). The full section-centre direct shear box tests considered the impact of the following factors on the reinforcing potential of roots: root area ratio (the ratio of root area to soil sample area), root length, root diameter, root stiffness and root spacing/distribution. The numerical models were developed in line with the p-y and t-z pile analysis techniques, used to model lateral and axial loading, respectively, and were constructed in Abaqus/CAE. They consider the root as a beam-column and the mechanical soil root interaction as a series of discrete non-linear springs. The properties of the springs were back calculated from the cross section-front laboratory shear box and uprooting tests, as well as being determined theoretically (using standard pile design codes). The results of the numerical models show that the p-y and t-z analysis techniques can be successfully applied to the study of soil root interaction, provided appropriate springs and root properties can be defined. Moreover, they show that the proposed tools improve substantially upon existing root analysis models by accurately predicting the uprooting force (axial) or shearing contribution (lateral) as a function of applied axial or lateral displacement of the soil root system during deformation. Standard pile design codes, however, were shown to require adjustment for the application of soil root interaction. The output of the laboratory and numerical testing revealed a number of interesting findings, including: (i) A stress related parameter, such as dilation, provides a better representation of a roots contribution to soil shear strength than a root cohesion factor, which is currently used. (ii) The root area ratio, often used to define a root cohesion factor, is not directly related to root contribution (e.g. two samples with the same root area ratio, but different root lengths, stiffness's or diameters, will not necessarily have the same shear strength) (iii) Root bending capacity is significant in defining its reinforcing potential. Overall, the predictive tools developed in this Thesis have advanced their predecessors by: incorporating the effects of root bending; modelling the progressive contribution of roots during soil deformation, and; utilising an analysis technique that is already well established in industry. At present, however, they are in the early stages of development and require considerable improvement (such as the development of theoretical design codes for estimating t-z and p-y springs suitable for plant root soil interaction) before they can be considered a useful tool in practice.
Supervisor: Knappett, Jonathan Sponsor: James Hutton Institute
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available