基于水化度的泵送混凝土温升模型及参数反演

Temperature rise model for pumped concrete based on hydration degree and parameter inversion

针对泵送大体积混凝土绝热温升高、早期水化速率快的特点,结合绝热温升试验,构建了基于水化度的水化温升速率模型(简称水化度模型),采用人工蜂群算法对水化度模型、双曲线模型、单指数模型和双指数模型参数进行反演对比分析,将水化度模型应用于5组绝热温升试验结果拟合中,使用4种温升模型对上海某泵闸工程混凝土试块温度场进行了反馈仿真分析。结果表明,水化度模型仿真结果与试验结果吻合较好;水化度模型能反映不同阶段的水化热变化特性,其参数采用分阶段拟合,稳定性好,避免了一般模型参数过多时出现的过拟合问题,能为泵送混凝土结构施工期智能温控提供较准确的温度场数值模拟结果。

In view of the high adiabatic temperature rise and rapid hydration heat release rate of pumped concrete,a hydration temperature rising rate model based on hydration degree(hydration degree model)was constructed combined with the adiabatic temperature rise tests.Parameter inversion of hydration degree model,hyperbolic model,single exponential model and compound exponential model was performed and compared by the ABC(artificial bee colony)algorithm.The hydration degree model was applied to the fitting of five sets of adiabatic temperature rise test results,and the four temperature rise models were performed a feedback simulation analysis on the temperature field of a concrete test block in a pump gate project in Shanghai.The simulated results show good agreement with the experimental data.Hydration degree model can fully describe the characteristics of the hydration heat release in different stages,and the parameters are fitted in steps,which can greatly avoid the over-fitting problem when too many parameters exist.The model can provide more accurate temperature simulation results for intelligent temperature control during the construction of pumped concrete structures.