Soils experience reductions in shear strength when pore pressures increase, which can happen under different types of loadings such as static loadings or earthquake-induced cyclic loadings. At present, widely used correlations for soil strength loss have inconsistencies, especially as it relates to some soil types and their amounts of soil strength loss and associated strains. For example, since the early 1970s, geotechnical engineers worldwide have largely relied upon empirical correlations to predict soil liquefaction susceptibility, triggering, and consequences/damage due to earthquakes. Such analyses are necessary to ensure the stability and adequate performance of structures like buildings, bridges, retaining walls, and engineered soil slopes in the event of an earthquake. While the geotechnical community largely has accepted these empirical correlations as the state of practice, numerous inconsistencies exist in their application. Existing correlations have primarily been developed for clean quartz and silica sands. The application of these correlations and approaches to other soils, including silty soils, gravel-rich soils, sands with variable mineralogy, and even low-plasticity silts and clays is largely unknown. These knowledge gaps contribute to our inability to accurately predict the dynamic behavior of these soils, resulting in empirical correlations to assess liquefaction susceptibility, triggering, and consequences that are inconsistent and result in significantly different remediation designs. Recent cyclic (and even monotonic/static) laboratory testing of various transitional soils validate these concerns. Importantly, these inconsistencies can result in both overly-conservative (expensive) and under-conservative (dangerous) outcomes, neither of which is optimal. To improve the accuracy of static and dynamic soil behavior predictions, there is a need for a universal, consequences-based approach for all soils that focuses on the compressibility and undrained shear behavior of the soil. This may include use of the “delta Q” approach, which is less affected by the overconsolidation ratio and stress history of soil, for improved soil identification and characterization using cone penetration test (CPT) data (Saye et al. 2017, ASCE JGGE), as well as other recent advances in geophysics.

The overall objective of this multi-year, multi-phase effort is to create a true performance-based model to evaluate the consequences of undrained response in all soils, including consequences resulting from earthquake-induced liquefaction and cyclic softening. Through this overall project, a more robust method for estimating field performance of soils during undrained events (including earthquakes) will be developed and tested. Due to the ability of the CPT to collect nearly continuous profiles of data in most soil types, for these studies we will focus initially on using CPT data for analyzing undrained shear behavior and liquefaction hazards. The framework is intended to be adaptable to other methods such as Standard Penetration Test (SPT), laboratory testing and analysis, and shear wave velocity (Vs) data. The objective of this Phase 1 study is to fill critical data gaps to document the undrained shear behavior of sands, silts, and clays for both static and dynamic loadings, and to provide a preliminary set of predictive models for the undrained shear response of soils. We anticipate that several state DOTs would be interested in participating in this initial pooled fund study. Later, in separate pooled fund studies, Phase 2 would focus on additional development of the models for consequences-based analysis of the undrained shear behavior of soils, and Phase 3 would focus on testing and validation of the models.

Planned tasks for this Phase 1 study are as follows: (1) Perform a literature review of related studies. (2) Conduct a field sampling program for transitional soils at sites in several states. (3) Perform conventional laboratory tests. (4) Perform advanced laboratory tests. (5) Compile a database of soil in-situ resistance and corresponding undrained shear strength and strains from across the United States. (6) As part of the Next Generation Liquefaction (NGL) Project modeling effort, develop and deliver a preliminary set of predictive models based on available field-case-history data for the monotonic and cyclic undrained shear response of soils. In addition to liquefaction triggering, these models may include excess pore pressure ratio, maximum shear strain, volumetric strain, lateral deformations, and degraded undrained shear strength. (7) Prepare the Phase 1 Final Report. (8) Meet with the multi-state study panel to discuss results, including whether further research is recommended, and what it might entail. (9) Conduct education, outreach, and training.

Additional agencies are welcome to join this TPF study while it is going. Please contact David Stevens (davidstevens@utah.gov) if you are interested in having your agency join the study.  The Utah Department of Transportation (UDOT) will be the lead agency for this Phase 1 study, with Grant Gummow (ggummow@utah.gov) as the UDOT Champion. We intend to hire a firm or university as the prime consultant through qualifications-based selection in the UDOT General Engineering Services Pool, Research Work Discipline, after sufficient funds are committed by study partner agencies. The Phase 1 study is planned to begin in spring or summer of 2022, with study completion in approximately three years’ time. The minimum partner commitment expected for the Phase 1 study is $60,000 in FFY 2021 or 2022, or the $60,000 total can be split between FFY 2021, 2022, and 2023 (or 2024 or 2025) at a portion per year.