(35a) Mixture Analysis of Gases by Raman Imaging

Introduction:

Various processes in the field of chemical engineering are rate-determined by mixture generation and not by the subsequent reactions. The mixing degree of two different flows is represented by the two scalar quantities ?mixture fraction? and/or ?mixture temperature? while the mixing driving force is represented by the vectorial quantity ?velocity field?. Hence there is the necessity for a measurement technique that simultaneously acquires the temperature, composition and velocity field for a comprehensive characterization of the mixture formation. Furthermore this measurement technique needs to be non-intrusive in order to prevent any distortions of the flow field. Current trends in laser imaging techniques for flow analysis point towards high speed and multi-parameter imaging. By means of the latter, it is aimed to investigate the interaction of scalar fields and vector fields during the turbulent mixing process in a multi-component system. In this context spontaneous Raman scattering is introduced as a potential optical technology suitable to achieve this goal.

Theoretical Background:

The term ?Ramanography? is known since 1974 when Hartley measured the distribution of one certain species molecule using planar Raman scattering [1]. Ramanography is based on the Raman Effect, which in fluids denotes, that a photon is inelastically scattered from a molecule. Consequently the scattered photon is spectrally shifted from the incident photon by the energy of the molecular transition. This so called Raman shift is species specific and its intensity is directly proportional to the number of molecules. Ramanography was applied for the analysis of jets and turbulence in reacting and non-reacting flows under ambient pressure and in pressurized chambers at ambient temperature and under cryogenic conditions [2]. Due to low detection signal intensities usually methane and hydrogen are probed because of their rather high Raman scattering cross sections. In addition to multi-species investigations and local density measurements [3] the possibility of acquiring the temperature and the concentration field simultaneously was developed [4] and already used in a hydrogen internal combustion engine [5]. The composition of the jet mixing experiment is characterized by the H₂ mole fraction x(n_H2) which is defined as the number of H₂ molecules n(H₂) over the total number of H₂ and N₂ molecules n(N₂). As the Raman signal is directly proportional to the number of molecules, the corresponding H₂ Raman signal fraction x(I_H2) is the H₂ Raman signal intensity I(H₂) over the total intensity of the H₂ and the N₂ Raman signal I(N₂).

To correlate x(I_H2) and x(n_H2) a calibration is carried out at defined pressures by filling the injection chamber with varying H₂/N₂ compositions. The benefit of considering Raman signal ratios is that pulse-to-pulse laser energy fluctuations, inhomogeneities in the excitation light sheet, beam distortions and window fouling effects cancel out. The temperature acquisition is based on probing the temperature sensitive distribution of molecules among their rotational energy levels. Figure 1 shows the theoretically calculated populations of several rotational J and vibrational v levels of hydrogen and nitrogen and the corresponding Raman shift in wave numbers. The population distribution across the rotational and vibrational energy levels follows the Boltzmann statistics.

Figure 1: Population of vibrational levels (v) of hydrogen and nitrogen as well as the population of different rotational levels (J) of hydrogen for three different temperatures; according to [4]

The population of the hydrogen rotational levels within temperatures between 200 and 800 K is temperature sensitive. Contrary, the population of the nitrogen and hydrogen vibrational levels is almost constant in the considered temperature range. Hence the number of species i molecules in the vibrational ground state v₀ is equal to the total number of species i molecules present between 200 K and 800 K,

which is true for H₂ and for N₂. Thus the temperature can be obtained by measuring the signal intensities of the hydrogen rotational levels. High temperature resolution can be achieved, if the population ratio of J₃ and J₁ is measured.

The combination of Ramanography and particle image velocimetry (PIV) allows the simultaneous acquisition of scalar and vectorial data. PIV is an optical measurement technique where a fluid is seeded with a trace of particles small enough to follow the gas flow without disruption. The flow is illuminated subsequently by two laser pulses. The incident light is scattered elastically from the particle and is termed Mie scattering here. The Mie scattering of both incident laser pulses is detected one after the other by a CCD-camera which can take two images synchronized to the time delay between the incident laser pulses. A cross correlation algorithm computes the shift of the scattering patterns in both images and represents this shift as a vector field.

Results:

The mixture formation of hydrogen and nitrogen was investigated regarding the temperature and concentration field by Ramanography [4]. The Raman process was excited with a Nd:YAG Laser at 532 nm. Hence the relevant Raman signals result at 549 nm (J₁ rotational level H₂), 564 nm (J₃ rotational level H₂) and 607 nm (v₀ vibrational level N₂). These three signals were separated by color selective (dichroitic) mirrors and detected by three CCD-cameras. Figure 2 shows the simultaneously acquired concentration and temperature field of a hydrogen jet in a nitrogen environment.

Figure 2: Mean and single-shot hydrogen mole fraction x(n_H2) and hydrogen temperature T(H₂) images of a hot (473 K) hydrogen jet which is injected into nitrogen at room temperature (297 K) at an absolute pressure of 1.1 MPa.

This measuring technique was combined with PIV to receive in addition to the scalar field also the vector field. As Ramanography is species selective the signals of different species are spectrally displaced from the incident laser pulse. Supplementary signal spectra are not overlapping for small molecules (typically gases). Accordingly Raman scattering is perfectly suited for the simultaneous application with other optical measurement techniques in multi parameter experiments. One disadvantage of spontaneous Raman scattering suffers from very small Raman scattering cross sections, which are several orders of magnitude smaller than Mie scattering cross sections from droplets or particles. Consequently for combining spontaneous Raman scattering with droplet seeded PIV experiments, the detection of pure and reliable Raman signals in the gas phase, without interference from Mie signals, is most challenging. The signal separation is achieved by an ascertain set of filters and dichroitic mirrors. Figure 3 shows the combination of a PIV vector field with a Ramanography composition field for a droplet seeded H₂ flow injected in a purged nitrogen chamber at 0.95 MPa [6].

Figure 3: Combination of the PIV vector field and the Ramanography composition field when a droplet seeded H₂ flow is injected into a N₂ purged chamber at 0.95 MPa and at room temperature (spurious vectors are removed).

Moreover the limits of Ramanography regarding the temperature range were probed by injecting hydrogen at low temperatures in a purged nitrogen chamber. The temperature was calculated using the adiabatic mixture temperature. Figure 4 shows the temperature and mole fraction distribution of the injection of hydrogen in a purged nitrogen chamber. The spatial resolution of the presented images in Figure 3 and Figure 4 is 184 µm x 184 µm. The cloud-shaped propagation of the hydrogen jet is owed to the usage of a hydrogen hollow cone injector.

Figure 4: Temperature and mole fraction distribution of the injection of hydrogen in a purged nitrogen chamber at 0.4 MPa and at injection temperatures of 200 K. The image was taken 500 µs after start of injection.

Conclusion:

By use of the inherent properties of Raman scattering like species selectivity, direct quantifiability and independency from tracers, a powerful and flexible laser diagnostic technique provides deep insight in mixing processes in the gas phase. Several parameters essential for the mixture formation are measured simultaneously. Short laser pulse durations facilitate the resolution of turbulent processes. Finally the experimental data can be used to validate numerical models. Further improvements of the measuring strategy as well as ongoing development of laser systems and cameras will extend the application range steadily. The authors gratefully acknowledge funding of parts of this work by the Bavarian Research Association (BFS) and the German Research Foundation (DFG), which also funds the Erlangen Graduate School in Advanced Optical Technologies (SAOT) in the framework of the German excellence initiative. The authors also acknowledge equipment support of BMW Group.

Reference list:

[1] D. L. Hartley, Laser Raman gas diagnostics, Plenum, 1974.

[2] a) M. B. Long, P. S. Levin, D. C. Fourguette, Opt. Lett. 1985, 10, 267; b) D. Kyritsis, P. Felton, F. Bracco, Int. J. Altern. Propul. 2007, 1, 174; c) A. R. Masri, R. W. Dibble, R. S. Barlow, J. Raman Spectrosc. 1993, 24, 83; d) R. Woodward, D. Talley, 1996; e) D. Marran, J. Frank, M. Long, S. Stårner, R. Bilger, Opt. Lett. 1995, 20, 791.

[3] S. Dowy, A. Braeuer, K. Reinhold-López, A. Leipertz, J. Supercrit. Fluids 2009, 50, 265.

[4] A. Braeuer, A. Leipertz, Appl. Opt. 2009, 48, 57.

[5] S. Engel, P. Koch, A. Braeuer, A. Leipertz, Appl. Opt. 2009, 48

[6] A. Braeuer, S. R. Engel, R. F. Hankel, A. Leipertz, Opt. Lett. 2009, 34, 3122.