One significant shortcoming of simplified liquefaction analysis is that it does not account for dynamic interactions during ground shaking and development of liquefaction in a deposit. The lack of consideration of cross-layer interactions results in excessively rapid evolution of the liquefaction response and diminished capacity of simplified analysis to distinguish between varying severities of liquefaction responses associated with different soil profiles and deposit characteristics. This paper presents a framework for incorporating dynamic interaction effects in simplified analysis and introduces preliminary models for dynamic response and diffusion effects in liquefying deposits. The improved accuracy of the modified simplified analysis due to incorporation of interaction effects is demonstrated and discussed through comparisons with liquefaction case histories and results from rigorous dynamic analyses.

Here, we introduce three innovative challenges to liquefaction prediction technology that the authors have addressed in recent years. The first is an innovation in the test procedure for testing the liquefaction phenomena in element tests. Elemental tests for investigating liquefaction were tests to determine the undrained cyclic shear strength (liquefaction strength). Over the past few decades, liquefaction resistance design has shifted to performance-based design, and there is an increasing demand for soil deformation performance during liquefaction rather than strength tests. A new test method that focuses on residual deformation characteristics is proposed. The second is the innovation of in situ survey methods. In this section, we introduce the challenges of developing a cyclic pressure meter test system. Current in-situ survey technologies are based on standard penetration test (SPT) and cone penetration test (CPT); however, the accuracy of liquefaction prediction has hardly improved in recent decades. Therefore, a fundamental shift in thinking is required. Third, we introduce a technology for detecting the degree of liquefaction from seismic motion without site ground profile information. In today's information society, it has become easier to install seismometers, and technologies using internet of things (IoT) and artificial intelligence (AI) are rapidly developing. When a major earthquake occurs, using the proposed technology, it is possible to instantly predict the extent of liquefaction damage based on earthquake records. In addition, it is possible to detect moderate liquefaction from the records of small and medium earthquakes, making it possible to predict the liquefaction risk in the event of a large earthquake. The challenges introduced here are just beginning and are yet to be resolved. Researchers and engineers involved in geotechnical earthquake engineering are encouraged to address the challenges of developing innovative technologies without being tied to existing ones

Modernization of ageing motorway infrastructure poses a major engineering challenge worldwide. With respect to existing pile groups, any upgrade is not only a challenging, costly, and time-consuming operation, but it can also lead to significant indirect impact due to traffic deterioration. As part of a research project funded by the Swiss Federal Roads Office, this work examined the potential of avoiding such major operation by taking advantage of nonlinear foundation response. Evaluating a plethora of Swiss bridges, a simple –yet realistic– 2 × 1 bored pile group on saturated sand is selected as the prototype problem. The latter is examined in a combined experimental and numerical study. Initially, a series of model tests is conducted at the ETH Zurich drum centrifuge. The test campaign provides not only fundamental insights on pile group response, but also benchmark results for subsequent validation of finite element (FE) models. The numerical study initially focuses on model scale, showing good agreement between the FE model and the centrifuge tests. The unavoidable scale effects of centrifuge modelling are critically discussed and a careful transition is proposed from model to prototype scale. Back to reality, a parametric numerical analysis allows for the identification and quantification of the key resisting mechanisms. The conclusions of this work contribute towards optimized design of new and existing reinforced concrete (RC) pile groups.