论文部分内容阅读
根据LNG分层翻滚的温度信号涨落,利用互信息函数和小数据量的Rosenstein算法,计算了翻滚事故样本的时间延迟、嵌入维数以及最大Lyapunov指数,得到了事故的最大可预测时间;基于盐水与LNG分层、传热以及对流过程的相似性分析,建立了盐水模化实验系统,以模拟储罐分层翻滚时的密度场。通过实验,得到了LNG分层形成的条件、翻滚发生的时间、初始密度差、充注速度和漏热等因素对分层翻滚事故的影响。实验证明:当初始密度差为17.2kg/m3,漏热为3.9W/m2时,翻滚发生的时间为28h,这一结论与液化天然气储槽中进行的翻滚实验数据相吻合。
According to the fluctuation of temperature signal of stratified tumbling LNG, the time delay, embedding dimension and maximum Lyapunov exponent of rolling accident samples were calculated by using Rosenstein algorithm with mutual information function and small data volume, and the maximum predictable time of accident was obtained. Salt water and LNG stratification, heat transfer and convection process similarity analysis, the establishment of a salt water system experimental system to simulate the tumbling tank density density field. Through the experiment, the formation of LNG stratification conditions, tumbling time, initial density difference, filling speed and heat leakage and other factors on the tumbling roll accident. Experiments show that when the initial density difference is 17.2kg / m3 and the heat loss is 3.9W / m2, the tumbling time is 28h, which is in good agreement with the tumbling experiment data in LNG storage tank.