The phenomena of liquefaction–induced lateral spreading of saturated cohesionless soils triggered by movements of caisson type quay walls are studied. First of all, the patterns of ground failure at the Taichung Habor in Taiwan, and Port Island and Rokko Island in Japan were investigated in detail. Then, a series of centrifuge shaking table tests were carried out to study both the deformation characteristics of backfills and the response of caisson type quay walls during liquefaction induced by strong earthquakes. A comparison of test results and observed field ground failure is made and the possible mechanisms of lateral spreading triggered by quay wall movements are discussed.
Among the methods for improving the strength of weak soils, Blasting Densification method (BDM) is characterized with fastness in constructional time and low-cost, compared to other methods. As BDM is conducted by detonating explosives in a borehole, this method enables efficient compaction of sandy soil in a great depth or any depths desired.
The historical development of BDM, the mechanism and the design concepts of BDM are presented with focus on elucidating the key factors influencing the effectiveness of compaction. Meanwhile, model tests conducted in laboratory and full-scaled in-situ tests have been conducted to further identify the deformation of the soil body after the blasting and to access the validity of adopting BDM in Taiwan. Model test results indicate that a cavity was formed at the moment of explosion, followed by collapsing of the cavity, provided sufficient overburden exists. Field test results reveal that the layout of the blasting sequence and detonation delay play an important role in inducing liquefaction of the loose stratum, which in turn results in a more compacted structure of particles and thus improves the strength of the soil.
PROBABILISTIC ANALYSIS FOR EARTHQUAKE-INDUCED SOIL LIQUEFACTION
工程上慣常使用之土壤液化(soil liquefaction)潛能評估法，係直接利用地震危害度分析所得之硬地盤最大地表加速度PGA值，及工址可能發生之最大地震規模Mu作分析，其結果往往過於保守，對於液化發生之機率未作評估。本研究利用蒙地卡羅模擬法(Monte Carlo simulation)產生與台灣地區地震危害度相符(hazard consistent)之地表運動參數，並據以評估地盤發生土壤液化損害之年超越機率，此法不僅能充分且合理地反映震區特性對工址可能產生的影響，更能直接評估土層液化之損害風險，其結果可提供大地工程規劃、設計及決策時風險管理之參考。
In engineering practices, the potential of soil liquefaction during earthquakes is evaluated based on the maximum ground acceleration and the largest magnitude of earthquake of the site. The results obtained are usually over-conservative due to the lack of probability basis. In this paper, a method based on the probabilities of earthquake occurrence and liquefaction occurrence is proposed. A large number of earthquake data, which are consistent with the hazard curve of Taiwan area, are generated by using the method of Monte Carlo simulation. Based on that, the risk of liquefaction damage of a typical soil profile is evaluated using several liquefaction-assessment models. The results thus obtained give the risk-based liquefaction potential which is consistent with the seismic hazard of the site.
Several simplified empirical methods were used to evaluate the liquefaction potential of Kaohsiung area. Microzonation map of liquefaction potential is constructed base on the liquefaction potential index proposed by Iwasaki. The result obtained from any empirical methods shows that the liquefaction potential in nearshore district of Kaohsiung, certain areas in downtown and Lotus Pond in Tsoying district are higher than the other in-land districts.
This paper gives a brief introduction to the concept of seismic design of pile foundation in a liquefied ground. The concept includes the considerations of modulus and strength reductions of liquefied soil, the effect of liquefaction-induced ground displacement on the pile, and the interactive behavior among pile, liquefied soil and superstructure. Two pseudo-static design methods are introduced to compute the kinematic force from liquefaction-induced ground displacement on the pile. A simple one dimensional finite element model is also proposed to model the interactions among pile, liquefied soil and superstructure.
This paper presents an assessment of existing probabilistic methods for liquefaction potential evaluation. Emphasis is placed on comparison of probabilities of liquefaction calculated with four different methods. Two of these methods are based on SPT, and the other two are based on CPT. In both SPT- and CPT-based evaluation, logistic regression and Bayesian techniques are applied to map factor of safety to probability of liquefaction. The present study shows that the Bayesian approach yields more conservative results than does the logistic regression approach, although results from the two approaches are quite comparable. Discussion of the procedure for risk-based liquefaction potential evaluation is also presented.