工程现场珊瑚礁砂场地主要以珊瑚砂、砾组成的宽级配珊瑚礁砂，其砾粒含量分布从20%～90%，其液化特性与普通石英砂有较大区别，如仍采用现行液化判别方法评估珊瑚礁砂场地液化潜势，则容易导致工程场地的抗液化处理设计不经济或无法满足要求。以中国南海岛礁和东帝汶珊瑚礁砂为研究对象开展了原级配大动三轴试验分析，建立了基于抗液化强度（cyclic resistance ratios，简称CRR）与相对密实度Dr关系液化判别方法，并通过离心机振动试验进行对比分析。结果表明，当采取相同地震动工况时，由动三轴试验产生的超孔压比相比模型试验超孔压比大；当持时增加到30周时（对应震级8级），土体液化深度达20 m，有效证明了珊瑚礁砂场地遭遇强地震动时具有液化潜在风险。此外，通过液化判别计算，验证了基于CRR-Dr关系的液化判别方法准确率达82.5%，且判别不一致工况的判别结果偏保守，进一步验证了此方法可应用于工程抗液化设计。The coral sand in the project site ismainly composed of coral sand and gravel with wide gradation characteristics

and its coral gravel content is distributed from 20% to 90%. Its liquefactioncharacteristics are quite different from quartz sand. The application of thecurrent liquefaction evaluation method to the coral reefsand site will lead to the design of liquefaction remediation beinguneconomical or unacceptable. In this study

the sands from islands and reefsof the South China Sea and the coral reefs of East Timor were taken as theresearch objects to conduct the original gradation dynamic triaxial test. Aliquefaction evaluation method was developed based on the relationship betweenthe liquefaction resistance CRR and the relative density Dr

and a comparative analysis was carried out through the centrifuge vibrationtest. The results show that under the same ground motion conditions

the excesspore pressure ratio generated by the dynamic triaxial test is larger than thatof the model test; when the duration increases to 30 weeks (corresponding tomagnitude 8)

the soil liquefaction depth is up to 20 m

which effectivelyproves that the coral reef sand site has the potential risk of liquefactionunder the action of strong earthquakes. In addition

through the calculation ofliquefaction discrimination

it is verified that the accuracy rate of theliquefaction evaluation method based on the “CRR-Dr”relationship is 82.5%

and the inconsistent results are conservative for theworking conditions

which can be applied to engineering liquefaction mitigationdesign.