State Key Laboratory of Precision Measurement Technology and Instrument, Department of Precision Instrument, Tsinghua University, Beijing 100084, China
Fund Project:Project supported by the National Key R&D Program of China (Grant Nos. 2020YFB2010701, 2020YFC2200101).
Received Date:17 June 2021
Accepted Date:02 July 2021
Available Online:15 August 2021
Published Online:05 November 2021
Abstract:Temperature, as an important parameter in combustion diagnostic process, will directly affect the combustion efficiency and the generation of combustion products. The accurate measuring of combustion temperature and then controlling of combustion state can not only contribute to avoiding the generation of harmful waste gas, such as carbon monoxide (CO) and oxynitride (NOx), but also improve the combustion efficiency, thereby saving the energy. However, in practical applications, dynamic and high-temperature combustion field has strict requirements for measurement accuracy and response speed of the thermometry technology. As an advanced spectral thermometry technology, coherent anti-Stokes Raman scattering (CARS) has a much higher spatial resolution, and can achieve accurate temperature measurement in high-temperature environment, so CARS has the potential applications in complex combustion field. For the temperature measurement requirements in the complex dynamic and high-temperature combustion field, we demonstrate a hybrid femtosecond/picosecond coherent anti-Stokes Raman scattering thermometry method through using the second harmonic bandwidth compression method, and achieve accurate measurements and dynamic response to temperature in dynamic and high-temperature combustion field. By using the narrow-band picosecond pulse obtained from the sum frequency process of femtosecond pulse in the BBO crystal as a probe pulse, this thermometry method can achieve single-shot, 1-kHz temperature measurement in high-temperature flame. We utilize the standard burner to simulate dynamic combustion field in a range of 1700–2200 K by changing the equivalence ratio quickly, and carry out continuous temperature measurement in 70 s by our thermometry method in this simulated dynamic and high-temperature flame. The least square method is used to fit the theoretical spectrum library to the actual single spectrum, and the fitting temperature corresponding to the actual single spectrum is obtained from the curve of fitting error. The continuous temperature measurements in 70 s exhibit superior performance in dynamic and high-temperature flame with a temperature inaccuracy less than 1.2% and a precision less than 1.8% at four different temperatures, and can track the temperature variation process within 0.2 s dynamically. These results verify the accuracy, stability and response speed in dynamic and high-temperature environment, and provide a new system scheme for thermometry in practical harsh combustion field. Keywords:coherent anti-Stokes Raman scattering/ thermometry/ dynamic measurement
表14种流速配比及其对应参考温度[36] Table1.Four flow velocity ratios and their corresponding reference temperatures.
-->
4.1.光谱拟合
通过将实际实验中的参数代入2.2节中的理论模型, 可获得符合实验条件的、不同温度下的理论光谱库, 理论光谱库的温度范围为300—2500 K, 温度间隔为1 K. 利用最小二乘法计算采集到的单幅光谱与理论光谱库中不同温度下的理论光谱的误差, 从而得到每个单幅光谱在不同温度下的拟合误差曲线, 理论拟合温度为单幅光谱拟合误差曲线最小值所对应的温度. 下面选择一幅光谱为例具体说明. 该幅光谱原始光谱如图5(a)所示, 原始光谱信噪比约为2373.5, 高于此前报道的高温下的单幅信噪比[33], 侧面说明了本系统实现单脉冲稳定测量的能力, 这里信噪比是指光谱峰的最大值除以背景噪声的标准差. 对原始光谱进行处理, 减去背景噪声, 将波长转换为对应的拉曼频移, 就可以得到如图5(b)所示的归一化光谱. 按照上文所述, 利用最小二乘法, 计算2201幅理论光谱与处理后的归一化光谱的拟合误差, 误差曲线如图5(c)所示, 误差最小值约为0.05832, 对应温度为1705 K, 则该幅光谱的拟合温度便为1705 K. 实际光谱与误差最小温度下的理论光谱的对比如图5(d)所示, 可以看出, 实际光谱与理论光谱拉曼峰的位置与强度基本一一对应. 图 5 (a) 原始光谱; (b) 处理后的归一化光谱; (c) 拟合误差曲线; (d) 拟合光谱与实际光谱的对比 Figure5. (a) Original spectrum; (b) normalized spectrum after processing; (c) curve of fitting error; (d) comparison of fitting spectrum with actual spectrum.
24.2.动态结果分析 -->
4.2.动态结果分析
按照4.1节的步骤, 对70000幅光谱进行了分析, 得到70 s内连续测量结果如图6所示. 图 6 70 s动态温度测量结果 (a) 28.6—29 s局部放大图; (b) 65.2—65.6 s局部放大图 Figure6. Results of dynamic temperature measurement within 70 s: (a) Local enlarged between 28.6 s and 29 s; (b) local enlarged between 65.2 s and 65.6 s.
图6中各段的红色线条为该段流速配比所对应的参考温度, 可以明显地看出, 在约8.5 , 28.6, 48.8 和62.2 s时间位置处, 由于改变甲烷和空气的流速配比, 火焰温度发生了变化. 图6(a)和图6(b)分别为28.6—29 s以及65.2—65.6 s时间段温度变化趋势的局部放大图, 可以看出, 改变流量后温度变化过程发生在0.2 s内. 千赫兹采集到的单幅光谱的拟合温度准确描述了火焰温度递增与递减的改变过程, 与温度本身过程量的特征相符合, 验证了本系统对于动态变化的高温火焰的响应能力. 图6中前4个流速配比下单幅光谱的测量温度的分布直方图如图7所示, 图7中红色直线表示该流速配比的参考温度, 左上角为该流速配比下的参考温度与平均温度的具体数值. 图 7 不同流速配比下测量温度的柱状分布图 (a) 0—8 s段; (b) 9—28.5 s段; (c) 28.9—48.8 s段; (d) 49—65.3 s段 Figure7. Histograms of temperature measurements in different flow velocity ratios: (a) From 0 s to 8 s; (b) from 9 s to 28.5 s; (c) from 28.9 s to 48.8 s; (d) from 49 s to 65.3 s.
为进一步说明混合飞秒/皮秒CARS测温的原理, 从4个流速配比下各选择1个单幅光谱, 对比其与最佳拟合的理论光谱, 结果如图8所示, 其中第1段所选择光谱与4.1节中选择光谱一致. 图 8 单幅光谱拟合结果 (a) 0—8 s段; (b) 9—28.5 s段; (c) 28.9—48.8 s段; (d) 49—65.3 s段 Figure8. Fitting results of single shot: (a) From 0 s to 8 s; (b) from 9 s to 28.5 s; (c) from 28.9 s to 48.8 s; (d) from 49 s to 65.3 s.