Date of Award

8-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Civil Engineering

Committee Member

Dr. Ronald D. Andrus, Committee Chair

Committee Member

Dr. C. Hsein Juang

Committee Member

Dr. Nadarajah Ravichandran

Committee Member

Dr. Qiushi Chen

Abstract

Research to quantify the the influence of aging processes (or diagenesis) on the static peak shear strength, the dilatancy, and the small strain dynamic stiffness of uncemented predominatly quartz sands is presented in this dissertation. New equations are proposed to model the dilatancy and the static shear strength due to diagenesis in natural sands as functions of either age or measured to estimated velocity ratio (MEVR). New predictive relationships between small strain dynamic stiffness and age are also recommended based on laboratory and field test results in natural sands. A laboratory investigation was performed to quantify the influence of age (or diagenesis) on the peak shear strength and the dilatancy of an uncemented Pleistocene age sand deposit at the Coastal Research and Education Center (CREC) near Charleston, South Carolina. Drained triaxial compression tests were performed on high quality intact specimens retrieved using the in situ freezing and frozen core sampling method, and on remolded specimens prepared to match the densities of the intact specimens. The stress-strain behavior of intact specimens was accompanied by dilation and a maximum or peak shear value, whereas remolded specimens generally contracted throughout shearing. The peak friction angle of intact specimens was found to be 3.0-8.6° higher than the peak friction angle of remolded specimens. A diagenesis-dilatancy term was added to the dilatancy index equation proposed by Bolton (1986) to account for the difference between intact and remolded peak friction angle. The resulting equation suggests that dilatancy caused by diagenesis and by density are both suppressed with increasing confining pressure, which has important implications for the design strength of natural deposits under heavy surcharge loads. A profile of in situ peak friction angle with depth is established from the test results and compared with values estimated from empirical relationships. The diagenesis-dilatancy term was generalized as a function of age based on a dataset of triaxial compression test results for ten different uncemented, predominantly quartz sands. Stong evidence was shown that dilatancy due to diagenesis increases with age, and that a model including age and confining pressure terms significantly improved predictions over a model with no age term. Therefore an age-dilatancy model was proposed. It was also shown that other properties such as density have little influence on dilatancy due to age. Because age of natural deposits is often difficult to accurately determine, a MEVR-dilatancy model was also proposed based on the framework of the age-dilatancy model. The age-dilatancy and MEVR-dilatancy equations were recommened to estimate intact peak friction angle from remolded peak friction angle or for predicting loss of strength during a disturbance or under large surcharges provided reliable in situ peak fricting angle estimates are available. General models for estimating peak strength are implied by the age dilatancy and MEVR dilatancy equations and can be used once the model is validated with the data presented in this study and the data compiled by Bolton (1986). Relationships for predicting the change in small strain shear modulus max() G or shear wave velocity () SV with time are reviewed. The max G -time relationship proposed by Afifi and Richart (1973) and the MEVR-time relationship proposed by Andrus et al. iv (2009) are related using a term called velocity ratio VR, which is the ratio of SV at a given time relative to its value in a deposit of similar density at a selected reference age. VR datasets were established from laboratory tests conducted on remolded sands and from laboratory tests conducted on intact sands. The VR datasets were combined to propose a VR-time relationship intended for natural sands. The proposed VR-time relationship based on laboratory results was compared with the VR-time relationship based on in situ VS and penetration resistance measurements implied by MEVR. The laboratory based relationship suggested a 3% change in VR for each ten fold change in age, while the field test based relationship suggested a 8% change with each ten fold change in age. It is found that much of the difference in the slope of the laboratory and field based VR-time relationships can be explained by the difference in fines content of the VR laboratory cases and VR field cases, which provides strong evidence for an influence of fines content on diagenesis. Much closer agreement between the VR-time relationships of field and laboratory cases with clean sands only is observed. The results indicate that field and laboratory based VR-time relationships can be used as indices for degree of diagenesis, provided the influence of fines content is accounted for. The preliminary results of a numerical study to predict the response of a Pleistocene age natural sand deposit at the CREC site during an in situ liquefaction experiment involving one of the NEES@UTexas mobile field shakers are presented. A plasticity model intended for earthquake engineering applications, was used for the Pleistocene sand deposit. Calibration of the model required considerably adjusting one of three main model inputs, called the contraction rate parameter, using the procedure recommended by Boulanger and Ziotopoulou (2015) due to the relatively low density and high predicted cyclic strength of the CREC sand. The simulation predicted concentrations of cyclic shear strain, cyclic stress ratio, and excess pore pressure that were located near the corners of the mobile shaker base plate during loading, and tended to produce a biased accumulation of shear strain toward either side of the sensor area. Below the base plate and within the zone where liquefaction sensor were installed at CREC, the excess pores pressure ratio was predicted to reach a maximum value of 12% and 18% at respective depths of 2.7 m and 3.3 m in the Pleistocene deposit. The prediction of low excess pore pressure buildup agrees with the limited field observations that were available to the author at the writing of this dissertation.

Share

COinS