834707674

IY3-6 in the excited State. For an equilibrium system the intensity ratio of the Stokes to anti-Stokes lines is dependent upon the temperaturę of the molecule. The laser frequency v_L must be well removed from the molecular absorption frequency. For non-resonant scattering the intensity of a Raman linę is proportional to the frequency change, the number density of molecules, and a transition probability for the energy change. The number density is determined from a calibration experiment carried out under the same conditions of temperaturę and pressure as those in the test experiment. High laser intensities are required with the exciting radiation at high frequencies.

The relatively much morę intense scattered Rayleigh radiation at the exciting frequency is removed by filtering. Filtering can be acomplished with a high resolution Raman spectrometer using double monochromators. Filtering can also be accomplished with high resolution interference filters or by polarizing the exciting frequency.

Substantial progress in the measurement of temperaturę by laser Raman scattering has been madę by Lapp and co-workers²⁵’²⁶. Observation is madę of temperaturę dependent effects in the spectral distribution of the Stokes Q-branch vibrational scattering. These effects arise predominantly from the vibration-rotation interaction and from significant population of excited vibrational levels. Upper-state bands originate from these excited levels and these are usually shifted towards the blue region of the spectrum. Observations were madę using an argon ion laser, operated at 1.5 watts with 4880-A. The scattered light was analysed by a double monochromator with 5000-A blazed gratings. Scattering data for H₂0 and 0₂ were obtained from lean H₂-0₂ flames and data for N₂ were obtained from a lean H₂-air flame. For the diatomic molecules considered in these experiments the Stokes Q-branch fundamental series profiles are calculated. These calculated profiles are used to fit experimental profiles in order to determine the scattering gas temperaturę.

Figurę 11 shows calculated Stokes Q-branch fundamental intensities for nitrogen over a rangę of temperatures from 300°K to 3500°K. Vibrational temperatures are proportional to the integral of intensity for particular bands, while rotational temperatures are proportional to the profile on the short wavelength side of each band via the influence of the vibration-rotation interaction. For calculations madę of nitrogen profiles, the spectral width can be seen to inerease relatively with inerease in temperaturę (Fig.l 1). The N₂ temperaturę was determined by fitting theoretical prefiles to experimental profiles obtained from the flame. An example of this fitting procedurę is shown in Figurę 12. The Raman scattering signatures are direct measurements of the relative populations of the molecular intemal modes and, for equilibrium situations, these relative populations correspond to the fundamental definition of temperaturę. This form of temperaturę diagnostics has the potential for becoming the most funda-mentally accurate scheme for non-perturbing three-dimensional measurements. Temperatures measured by Raman spectroscopy are shown to agree with independently measured temperatures utilizing a fine wire thermocouple to within 2 per cent.

Relatively few measurements have been madę of gas concentration in flames but those reported recently by Regnier et al²⁷ have demonstrated the feasibility of making gas concentration measurements with laser probes. In order to improve the sensitivity they used a stimulated scattering process called “parametric four-wave mixing” or “coherent Raman anti-Stokes scattering’’. This scattering is much morę intense than spontaneous Raman scattering. Two co-linear light beams of frequencies cj, and cj₂ generate a co-linear (anti-Stokes) wave at frequency 2cji - co₂ vibrational frequency co_v = cj, — cu₂ . The intensity of the new wave is proportional to the square of the number density of resident molecules. In the experiments of Ragnier et al²⁷ a single modę ruby laser was used, together with a stimulated Raman scattering oscillator. This oscillator consisted of a high pressure celi filled with a mixture of 80% H₂ and 20% He at a total pressure of 30 atm. This pressure celi converts a fraction of the punp pulse at oo, and produces a pulse at the required frequency cu₂ . The high pressure celi contains the same gas that is being detected. The detection capability of the instrument was shown by making measurements on H₂ diluted in N₂ at concentrations ranging from 10 ppm to 100%. Measurements were madę in a flame of natural gas containing 75% methane and 25% ethane, pre-mixed with air. The results of the H₂ distributions in a horizontal gas flame are shown in Figurę 13. The peaks of H₂ concentration are in the vicinity of the reaction zonę. The results shown in Figurę 13 demonstrate the type of new information that can be obtained using laser optical concentration methods. The H₂ concentrations would not normally be detected by a gas sampling probe. H₂ is generated as a result of the cracking of hydrocarbons as they pass through the high temperaturę gradients immediately prior to the main reaction zonę.

Developments in tunable dye-lasers will allow measurements to be madę of a number of species in the flames and will probably eventually replace the high pressure celi used by Regnier et al²⁷.

CONCLUSIONS

Measurements in turbulent flows with Chemical reaction need to be examined critically, taking into account the instruments used for making the measurement. The detailed flow structure can be complex, as illustrated by the photograph in Figurę 1. Measuring instruments cannot give a fuli description of the physical phenomena and the process of time-mean averaging can conceal the detail structure.

In turbulent flow systems instruments with frequency response in the kHz rangę are required in order to supplement time-mean average measurements with rms values of the turbulent fluctuating components of velocity,