Theory For Gas Sensing Technology Based On Acoustic Spectra

Acoustic spectra, i.e. the frequency-dependent sound absorption and speed spectrum, vary with gas composition in gas mixtures. With the outstanding ability of directly obtaining the molecular structure of gases, the gas sensing technology based on acoustic spectra becomes one of the most potential approaches for gas composition monitoring. This technique has wide promising applications in process industry, occupation health and safety, environment and emission monitoring, etc., attributing to its attractive features of in-line measurement, non-invasive measurement, rapid response, low power consumption, simple hardware, excellent long-term stability and high accuracy.The establishment of theoretical physical model to predict acoustic spectra in multi-component gas mixtures, is the theoretical foundation for this gas sensing technology; the formulation of analysis model to find the qualitative and quantitative relationship between acoustic spectra and gas composition in gas mixtures, is the key theory for the realization of this technology; and acquiring the simplified method of measuring acoustic spectra is a necessary step for the practical applications of it.This thesis begins with tracing the development of the gas sensing technology based on acoustic spectra and briefly introduces its principle, typical applications. The problems in the existing theory for the gas sensing technology based on acoustic spectra are also discussed in detail. Subsequently, the research of this thesis focuses on three aspects as follows:Firstly, the traditional thermal relaxation theory can not applied to gas mixture including three or more components. To solve this problem, the relaxation equations given by Schwartz et al., Tanczos, Lueptow et al., and the effective heat capacity of relaxing gas given by Herzfeld&Litovitz are theoretically connected together by means of the temperature ratio between molecular vibrational mode and external degrees of freedom. As a result, a theoretical physical model to predict acoustic spectra in multi-component gas mixtures is presented, and the analytical expressions for the relationship between multi-relaxation spectra and the relaxation characteristics of gas molecules are obtained. The predicted absorption spectra of various gas mixtures, consisting of methane, nitrogen, carbon dioxide, chlorine and oxygen, are consistent with the experimental data very well, and the peak errors are lower than1%. Moreover, the simulation results illustrate that only one or two peaks generally appear in a multi-relaxation absorption spectrum. This analytical model can directly obtain the analytical expressions of characteristic points of relaxation spectrum in gas mixtures. So, the analytical model provides an effective approach to analyze the relationship between sound propagation and molecular vibrational relaxation of gas mixtures under different ambient conditions.Secondly, the existing models can neither quantitatively explain the relation between sound absorption spectra and relaxation processes, nor analyze vibrational mode-contribution for sound absorption in multi-component gas mixtures. To solve those two problems, a decoupling model for multimode vibrational relaxations and a decomposing model for multi-relaxation spectra are developed respectively. On one hand, the decoupling model is provided to decouple the Vibrational-Vibrational coupled energy into each Vibrational-Translational deexcitation path, and analyze how multimode relaxations form the peaks of sound absorption spectra in gas mixtures. It is proven that a multimode relaxation is the sum of its decoupled single-relaxation processes; the amplitude of absorption peak is proportional to the isochoric molar heat of decoupled process, and the position of peak in frequencies is inversely proportional to the relaxation time of decoupled process. The decoupling model clarifies the essential processes behind the peaks in sound absorption spectra arising from multimode relaxations in multi-component gas mixtures. On the other hand, the decomposing model demonstrates that the sound relaxational absorption spectrum under a multi-relaxation process can be decomposed into the sum of spectra ascribed to the decoupled single-relaxation processes. By this decomposable characteristic, a Contribution-weight Analysis Matrix is obtained to analyze the contribution of each vibrational mode to the sound absorption, and thus the qualitative and quantitative analysis on the relationship between molecular vibrational modes and acoustic absorption spectra is achieved in multi-component gas mixtures.Thirdly, the traditional method of measuring acoustic spectra needs to vary the pressure in the gas cell, resulting in this method is not available for practical application. To solve it, a characteristic-point algorithm to reconstruct acoustic spectra in the whole frequency range is presented. This algorithm only needs to measure the sound absorption and sound speed of two operating frequencies at single pressure, without the necessity of detecting the gas density. The algorithm is developed from the fact:the frequency-dependent sound absorption spectrum due to a single-relaxation process can be reconstructed from the maximum absorption point (i.e. the characteristic-point); the characteristic-point corresponds to the two values of the relaxation frequency and the maximum relaxational absorption, and they can be synthesized by the acoustic measurements at two frequencies. Moreover, by acquiring the high-frequency sound speed, those two synthesized values can be used to reconstruct the sound dispersion spectrum. The reconstruction preciseness of a single-relaxation acoustic spectrum by the characteristic-point algorithm just depends on the accuracy of two-frequency acoustic measurements. The errors of reconstructed spectra are linearly dependent on the measurement errors at two operating frequencies. This linearity means that, via multi-measurement to obtain the average, the errors of two synthesized values in reconstructing acoustic spectra could be effectively reduced. Hence, a feasible and practical method of reconstructing acoustic spectra is acquired.This thesis was supported by the National Natural Science Foundation of China (Grant Nos.60971009,61001011), the Specialized Research Fund for the Doctoral Program of Higher Education of China (Grant No.20090142110019), the Natural Science Foundation of Hubei Province, China (Grant No.2010CDB02701), and the Fundamental Research Funds for the Central Universities, China (Grant No.2012QN083). Overall, this work provides some fundamental theory for the gas sensing technology based on acoustic spectra, and would promote the feasibility of the practical applications for this gas sensing technology.