Study on the interface behavior between unsaturated soil and steel surface

Abstract

An interface formed between a structural material and an unsaturated soil is common in numerous civil engineering projects. The ultimate shear strength at the interface is an important parameter for the design and safety assessment of the structures in the soils and is also a key factor in the design and analysis of the structural interfacial interactions with soil. Matric suction has a predominant effect on the shear strength and volume change behavior of the soil and soil-steel interfaces, thereby proper characterization of interface behavior is important for its accurate performance predictions. The experimental study of the interface behavior plays an important role in advancing the understanding of the complex behavior of soil-steel interfaces. It is believed that the critical interface plane which possesses the minimum shear strength exists on the surface of the structural material (counterface). Also, no defmite criterion for the selection of interface thickness (i.e., the layer which possesses the minimum shear strength) for different soil-structure interfaces is readily available. The main focus of this study is to investigate the behavior of the interface between a compacted completely decomposed granite (CDG) soil and steel counterfaces at (a) different shearing planes (b) and different counterface roughness under the influence of different matric suctions and net stresses. Considering the main focus of the study in view, an effort has been made to examine the variation of shear strength for pure soil and soil-steel interface by employing a modified direct shear testing device. Firstly, to investigate the elementary behavior of the soil-steel interface (rough) sheared at different shearing levels and compare the interface shear behavior with the soil shear behavior, a number of single-staged consolidated drained direct shear tests were carried out on interface and pure CDG soil specimens, under different matric suctions and net normal stresses. The results confirm that the suction and net normal stress significantly influences the shear behavior of pure soil and soil-steel interfaces sheared at different shearing planes. The behavior of stress-displacement curves of soil-steel interface tests is similar to those of soil tests. The suction envelopes were noted to be nonlinear. The degree of dilatancy contributes to the gain in shear strength and is dependent of both matric suction and net normal stress. It is noted that the matric suction plays an important role in evaluating the critical shear strength, as it directly affects the volume change behavior. It is found that the critical failure plane is dependent on the matric suction and an increase in suction reduces the critical interface layer thickness thereby gradually shifting the critical shear plane from the soil towards the counterface. The experimental results compare well with shear strength model proposed by Hossain (2010), which considers the influence of suction and dilation. Secondly, to examme the elementary interface behavior with different counterface roughness values and compare it with the behavior of pure CDG soil, a series of interface direct shear tests were performed between the CDG soil and a steel plate with different counterface roughness values under the same matric suctions and net normal stresses. The test results show that the counterface roughness, matric suction and net normal stress have significant influence on the hardening-softening and contractive-dilative interface behavior. An increase in the value of the net normal stress results in a partial stick-slip behavior during the shearing of soil-steel interfaces. The suction envelopes for different counterface roughness were observed to be non-linear and the apparent interface friction angle increases with matric suction. The experimental shear strength data are compared with an analytical model, that considers the influence of suction and dilation on apparent interface friction angle, and it is noted that the analytical model works well for rough interfaces under all the applied stress-state variables. Finally, a series of model pile pullout test were performed under different unsaturated conditions to evaluate the influence of surface roughness and water content on the elementary interface behavior. To measure the axial strain along the model pile during pullout, the model piles were installed with fibre bragg grating (FBG) sensors. The pullout test results are in good agreement with the observation from direct shear tests. The pullout forces are noted to be directly proportional to the surface roughness and decreases with increase in water content of the soil. The axial strain value along the model pile obtained from FBG sensors reveal that the strain at the top of the model pile is notably greater than that at the base. It is opined that it is worth to employ FBG sensors for examining the interface behavior of piles in unsaturated soil. Further studies are recommended to ascertain the performance of FBG sensors in unsaturated soil and testify the non-linearity of axial strain, shear stress and skin friction of pile. To summarize, this study attempts to investigate the interface behavior formed between CDG soil-steel counterfaces under different stress state variables. Different types of test suction as (a) suction controlled direct shear test to study the several key factors (such as matric suction, net normal stress and counterface roughness) that influences the interface shear behavior (b) FBG sensor based model pile pullout test were conducted to study several aspects of soil-steel interface behavior such as pullout behavior, axial strain, and skin friction of pile. The results show that the critical interface shear plane is greatly influenced by the soil suction and the critical interface layer thickness reduces as soil suction increases. The gain or loss interface shear strength is a function of particle grain size and counterface roughness. Variation in counterface roughness influences the interface shear strength non-linearly. However, further studies are required to investigate and identify the optimum counterface roughness for unsaturated soil-steel interface. Furthermore, this study also demonstrates successful implementation of FBG sensors in determining the axial strain and skin friction of pile in unsaturated soil. The axial strain induced in the pile is significantly influenced by the counterface roughness and the degree of soil saturation. The axial force along the length of the pile is noted be greater at upper end of model piles as compared to the lower end. Also, the variation of skin friction is non linear throughout the length of the pile. Summary and important findings from this study are presented in Chapters 5 to 7. Chapter 8 presents the major conclusions from this research along with the recommendation for further research work related to this research topic area.