Date of Award

8-2018

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

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. Weichiang Pang

Abstract

Uncertainties associated with the assessment of diagenesis or aging effects on soil liquefaction are investigated in this dissertation. Liquefaction-induced ground failure is a major cause of damage during earthquakes. Current liquefaction triggering evaluation procedures are primarily based on field case histories where liquefaction occurred in soil deposits that are less than a few thousand years old. Without correction for diagenesis, these evaluation procedures can provide excessively conservative predictions, often resulting in unnecessary and costly ground improvements in natural and older man-made soils. The assessment includes a comprehensive review of several field case histories indicating greater resistance to liquefaction in aged soils than in young uncemented soils during earthquakes, a discussion of the mechanisms that increase liquefaction resistance with time, an evaluation of proposed methods for quantifying the influence of diagenesis on liquefaction resistance (KDR), and an evaluation of proposed predictor variables for KDR. The published literature indicates that physical diagenetic processes tend to dominate the nature of interactions at the grain-to-grain contacts of sand deposits in the absence of sufficient cementing agents, while chemical processes tend to dominate where sufficient cementing agents are present. Variables proposed for predicting KDR include: time since deposition or last critical disturbance; ratio of measured to estimated small-strain shear wave velocity (MEVR); ratio of small-strain shear modulus to cone tip resistance (Gmax/qc); adjusted Gmax/qc (KG); and ratio of measured to estimated adjusted Gmax/qc (MEKG). MEVR, Gmax/qc, KG and MEKG are shown to be ratios of measured to estimated or reference shear wave velocity. MEVR appears to be a more robust predictor of KDR than time, Gmax/qc, KG and MEKG. Time should only be used at sites where deposit age can be accurately determined. The influence of diagenesis on liquefaction resistance of Holocene and Pleistocene soils near Christchurch, New Zealand is evaluated in this dissertation. Many Holocene alluvial and marine deposits in and around Christchurch experienced minor to severe liquefaction during the 2010-2011 Canterbury earthquake sequence. Liquefaction and permanent ground deformations were most severe in the city during the 22 February 2011 event. Permanent horizontal ground displacements up to 0.35 m also occurred at several locations in moderately sloping Pleistocene loess-colluvium in the Port Hills area, but no sand/silt boil has been connected to those deposits. Values of MEVR determined for the critical layers in Holocene deposits in four different areas agree well with the study by McGann et al. (2015) and are similar to ratios of about 1.0 computed for recently liquefied deposits in other areas of the world. On the other hand, MEVR values determined for the loess-colluvium typically range from 1.0 to 5.0, with the zone of lowest values occurring below the groundwater table and above a depth of 6.5 m. Average KDR values determined for the Holocene and Pleistocene deposits are in good agreement with the MEVR-based relationship developed by Hayati and Andrus (2009) from a global database. These findings support the use of MEVR and KDR for accurate site-specific liquefaction assessment in aged soil deposits. The liquefaction resistance of the 200,000- to 240,000-year-old Ten Mile Hill beds sands (Qts) near Charleston, South Carolina is also characterized. The characterization includes an evaluation of relative liquefaction susceptibility using MEVR and liquefaction potential index (LPI), an assessment of liquefaction potential in terms of liquefaction probability curves, and a back-calculation of KDR. Computed MEVR and LPI indicate that there is no significant spatial trend in the liquefaction susceptibility of Qts with respect to distance to the source zone of the 1886 Charleston earthquake. Without correction for diagenesis, liquefaction potential based on the 1886 Charleston earthquake is over predicted. Average back-calculated KDR for Qts deposits is consistent with MEVR-KDR data and relationships by Hayati and Andrus (2009). The evaluations at the New Zealand and South Carolina sites provide strong supporting evidence for the conclusion that MEVR yields more robust predictions of the effects of diagenetic processes on liquefaction resistance, than using time since deposition or last critical disturbance. Analysis of additional KDR case histories based on laboratory and field tests further indicates that MEVR is a better predictor of KDR followed by the ratio of small-strain shear modulus (Gmax/qc). The adjusted Gmax/qc (KG) and the ratio of measured to estimated adjusted Gmax/qc (MEKG) are the least robust predictors of KDR.

Share

COinS