论文部分内容阅读
对不同加载龄期的混凝土圆柱体进行长期徐变试验,结果表明混凝土圆柱体内徐变系数较体外徐变系数普遍偏小,加载龄期对混凝土体外徐变系数影响较大,对体内徐变系数影响较小。国内外徐变预测模型对短期加载龄期(9d)的徐变系数预测精度较高,对长期加载龄期(39d)的徐变系数预测精度较差,特别是对后期徐变系数预测普遍不足。GZ(1993)模型对徐变初期预测存在不连续现象,但整体预测效果最好。CEB-FIP(1990)模型预测效果最差,预测徐变系数的变化规律与实测值偏离较大。采用最小二乘法对实测数据进行非线性回归分析,分析表明CEB-FIP(1990)模型对名义徐变系数预测偏低,这主要是由CEB-FIP(1990)模型中加载龄期影响系数预测偏低造成的。根据不同加载龄期的徐变系数实测值,对CEB-FIP(1990)模型进行了修正,从而大大提高了徐变预测模型的预测精度。
Long-term creep tests on concrete cylinders with different loading ages show that the creep coefficient of concrete cylinders is generally smaller than that of in-vitro creep coefficients, and the loading age has a greater impact on the creep coefficient of concrete in vitro. The creep coefficient Less affected. Creep prediction models at home and abroad have higher prediction accuracy for creep coefficient of short-term loading (9d), prediction accuracy of creep coefficient of long-term loading (39d) is poor, especially for late creep coefficient prediction . The GZ (1993) model has discontinuities in the initial prediction of creep, but the overall prediction is the best. The CEB-FIP (1990) model has the worst prediction effect and the predicted variation of creep coefficient deviates from the measured value. The least squares method was used to conduct non-linear regression analysis of the measured data. The analysis shows that the prediction of the nominal creep coefficient of the CEB-FIP (1990) model is low, mainly due to the prediction of the loading age impact coefficient in the CEB-FIP (1990) model Low caused. The CEB-FIP (1990) model is modified according to the measured values of creep coefficients at different loading ages, which greatly improves the prediction accuracy of the creep prediction model.