机构地区:[1]西南交通大学土木工程学院,四川成都610031
出 处:《岩石力学与工程学报》2020年第8期1671-1681,共11页Chinese Journal of Rock Mechanics and Engineering
基 金:国家重点研发计划(2017YFC0504901)。
摘 要:为了提高冻土试验系统的控温精度、实现分凝冰发育过程的实时监测,基于固态制冷技术和线阵CCD(Charge Coupled Device)扫描成像技术,研发一种新型冻土可视化试验系统。该系统的核心部件包括固态控温模块和CCD可视化模块。在固态控温模块中,引入人工神经网络对控温算法的控制参数进行优化选取,使得试验系统对控温对象的温度状态具备自适应调节能力。而在CCD可视化模块中,使用CCD感光元件构建了操作简便、成像精度高的线阵扫描结构,其最高分辨率可达4 800 dpi×4 800 dpi。性能测试结果显示系统的恒温控制精度为±0.002℃。在变温幅度为0.05℃的控温过程中,其温度过冲只有0.009℃;当变温幅度增加至0.5℃时,其温度过冲仅为0.089℃。而在线性降温过程中(降温速率-1.2℃/h),其控温误差未超过±0.01℃。在此基础上,开展正弦温度波动下的应用试验。从试验结果上看,该系统运行稳定、控温精度高,可视化模块可观测到0.5mm宽度的分凝冰发育过程。并且所得数据表明土体过冷过程和冻融历史对温度场、分凝冰发育状态存在显著影响。因此,当前冻胀试验中为了消除过冷影响而采取的预冻结处理方式会改变土样的初始状态,从而影响土体冻胀特性的评估结果;而且所采取的单次冻结的试验方式无法反应冻融历史对冻胀特性的影响,尤其是对于室内重塑土样。相关研究有待深入,而该试验系统为开展此类研究提供了条件。In order to improve the temperature control accuracy and to monitor the growth process of ice lens in real-time,a novel visualization apparatus for freezing soil test,mainly including a solid-state refrigeration module and a CCD visualization module,was designed based on solid-state refrigeration and linear array CCD(Charge Coupled Device) scanning technology. The neural network was adopted in the solid-state refrigeration module to optimize the parameters of control algorithm so that the test apparatus has the self-adaption capability to temperature conditions of the controlled object. In the visualization module,a linear array scanning structure with the maximum resolution of 4 800 dpi× 4 800 dpi was assembled,which has the advantages of simple operation and high imaging accuracy. Performance test results show that the developed apparatus has a constant temperature control accuracy of ±0.002 ℃. In the variable temperature control process with an amplitude of 0.05 ℃,the temperature overshoot is 0.009℃. When the amplitude increases to 0.5 ℃,the temperature overshoot is only 0.089 ℃. The deviation does not exceed ±0.01 ℃ even in the linear cooling process with a cooling rate of 1.2 ℃/h. The application test with sinusoidal temperature boundary was carried out. The test results show that the control system is robust with high accuracy and that the ice lens as small as 0.5 mm width can be observed by the scanning imaging module. It is also indicated from the test data that both the supercooling of pore water and the freezing-thawing history have a significant influence on the temperature profile and the growth of ice lens. Hence,the prefreezing of soil samples,aiming at eliminating the supercooling process,will change the initial state of soil samples and affect the test results. Furthermore,the single freezing process cannot simulate the influence of freezing-thawing history on frost heave characteristics,especially for remoulded samples. The novel apparatus can provide test conditions for further
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...