Modeling the compression behavior of structured soils considering the effects of sample disturbance on the pre-compression pressure

Natural soils pose an inter particle bonds that makes their compression and failure behavior different than the compression and failure behavior of remolded soils. Different reasons have been suggested for the creatin of inter   particle bonds, including natural cementation, carbonating agents, and aging. These classes of soils are called structured soil. Structured soils could also be produced from artificial cementation by cement, lime, and other chemical agents. Overconsolidated generally show different compression behavior in overconsolidated state and normally consolidated state and their compression index in both states are different. Structured materials show a highly nonlinear compression behavior after initial yielding in the normally consolidated state. Their compression index has a large value at the beginning of the virgin yielding and decreases as the structure of the soil crashes. This highly nonlinear behavior prevents from adapting the conventional linear compression models in both semilogarithmic or fully logarithmic scale in e-log(p) or ln(1+e)-ln(p) spaces. On the other hand, sampling disturbs the soil samples and destructs the soil structure. Although there are several new sampling methods and apparatuses to reduce the disturbance of the samples, however the disturbance could not be completely removed from sampling procedure. Disturbance of the samples makes the determination of the precompression pressure of the soil samples where the compression regime of the soil changes, very complicated. There are several methods for determination of the precompression pressure of disturbed samples that most of them are based on graphical procedures. This paper presents a contentious compression model for volumetric compression and yielding of structured soils considering the effect of the sample disturbance on the determination of the precompression pressure. The compression behavior of the structured soil in low stress levels where the structure of the soil is relatively intact (RI) is known and can be measured in the laboratory. Also, the compression behavior of a structured soil at very high stress level where in fully adjusted (FA) state is similar to the compression behavior of the same soil in remolded state. The compression behavior of the structured soil with some degree of the disturbance must lays between these two reference states. Based on the Disturbed State Concept (DSC) the behavior of any complex phenomenon between two reference states of RI and FA could be completely described using a coupling mechanism called the state function. In this paper, the compression curve of the soil in low stress levels at overconsolidated state was considered as RI state and the compression curve of the same soil in the remolded state was considered as FA state.  A continuous sigmoid form state function was proposed for description of the continuous change of the e-log(p) curve of the soil from overconsolidated state to normally consolidated states. The effect of the sample disturbance was introduced in the state function by an intactness parameter. High values of the intactness parameter produce distinctive changes from overconsolidated regime to normally consolidated regime that is the main character of intact samples. Based on the proposed model the precompression pressure could be determined minimizing the deviations of the predicted results from observed results in compression test. The proposed model was verified by data reported in the literature and also laboratory tests conducted by the authors. The verification of the model showed the ability of the model in determination of the precompression pressure of artificially overconsolidated samples with known precompression pressures more precise than the other methods.