特征含水率对轻量土基本性质的影响规律

为了研究最优含水率、流动上、下限含水率对轻量土性能的影响，通过密度、无侧限抗压强度试验研究了混合土的工程性质。结果表明：采用流动性指标 控制混合土的流动性基本可行。混合土处于流动性上、下限含水率时，无侧限应力-应变关系曲线几乎重合，流动性上、下限含水率范围内土壤性质较为接近。无侧限抗压强度随含水率增加而衰减，但流动上、下限含水率对应强度差别不大。不管含水率高低，强度随龄期增长，都可以采用双曲线模型进行预测，并且总结了7 d、90 d强度与28 d强度之间的经验关系。当含水率为最优含水率时，混合土基本不收缩；在流动性上、下限含水率范围内，线收缩率范围为1.53%～4.71%，体积收缩率范围为4.53%～13.46%，收缩性受含水率、水泥剂量等因素影响。理想密度模型可以近似预测混合土的湿密度，高含水率时预测值有一定误差，误差范围为3.834%～8.231%。

Influence law of characteristic water content on basic properties of light weight soil

To research the influence of water content(optimum water content,flow lower limit water content,flow upper limit water content) on basic properties of light weight soil,engineering properties of mixed soil are investigated by density test and unconfined compressive strength test.The results show liquidity of mixed soil can be controlled feasibly by liquidity indexes.When water content are on the points of flow lower limit water content and flow upper limit water content,unconfined stress-strain relation curves are almost coincident.So when water content is between flow lower limit water content and flow upper limit water content,engineering properties of mixed soil are approximative.Unconfined compressive strength decreases rapidly with increased water content,but there are no significant differences for the strength on the points of flow lower limit water content and flow upper limit water content.No matter how much the water content is,strength-age relation can be predicted by hyperbolic model.Empirical relations between strength of 7 d and 90 d and strength of 28 d are summarized.When water content is optimum water content,there is only a little shrinkage.When water content is between flow lower limit water content and flow upper limit water content,the range of line shrinkage ratio is [1.53%,4.71%],and the range of volume shrinkage ratio is [4.53%,13.46%].Shrinkage property can be influenced by water content and cement dosage.Wet density of mixed soil can be predicted approximately by ideal density model,there is some error for predictive value,and error range is [3.834%,8.231%].