Doppler Global Velocimetry in flames using a newly developed, frequency stabilized, tunable, Long pulse, Nd: YAG Laser

Laser anemometry is an important tool in modern combustion research and in the development of industrial combustors. Its purpose is to deliver data for code validation and to give an overview on the flow field to enhance the understanding of the burner and to identify flow phenomena. In the last two decades laser Doppler anemometry (LDA) and phase Doppler anemometry (PDA) were used for this purpose. LDA and PDA are both point techniques and therefore quite time-consuming to apply. Planar light sheet techniques like particle image velocimetry (PIV) and Doppler global velocimetry (DGV) offer much higher data rates. Today there is high interest to use planar velocimetry techniques in flames. One further reason, why especially DGV might be the favourable technique for this purpose is, that LDA, PDA and also PIV are very sensitive against strong fluctuations of the index of refraction, especially at high pressures at which they often fail. DGV in contrast has not to image single particles and therefore is less affected by fluctuations of the index of refraction. In addition to this, DGV also does not require windows with brilliant optical qualities. Furthermore, DGV can –to a certain extenddistinguish between the velocity of the kerosene particles and the gas velocity, an important feature in optical diagnostics for combustion. Despite these advantages of DGV, the technique was not used for combustion research yet. Measurements in flames were hindered by the fact that the scattering signal has to overcome the background luminosity of the flame and the incandescence of solid particles (soot as well as tracers). The light intensity of cw Ar + lasers, which were often used for DGV, is normally not sufficient for this purpose, especially at high temperatures and high pressures. To overcome this problem, pulsed Nd:YAG lasers can be used in combination with gated CCD-cameras. However, commercially available, pulsed Nd:YAG lasers have a rather wide bandwidth of about 100 MHz, or even more, due to their short pulse duration. This large bandwidth may result in a serious reduction of the measurement sensitivity. These lasers also tend to show an uneven frequency distribution over the light sheet hight (Forkey, 1996; McKenzie, 1996), reducing the measurement accuracy even more. A new kind of narrow band frequency stabilized, tuneable, long pulse Nd:YAG laser was developed in the frame of a co-operation between the German Aerospace Center (DLR) and the Laser Center Hannover (LZH). The new laser is capable to fulfil the requirements for a successful DGV application. This paper describes the set-up and the performance of this laser, with special emphasis on its frequency stabilization. Furthermore the DGV camera system with its intensified cameras is presented and the special aspects of the image processing are discussed. Finally results from an atmospheric kerosene combustor are presented. These first sucessfull DGV measurements in a combustion experiment demonstrate the capability of DGV and may open a new field of applicability. PRINCIPLE of DOPPLER GLOBAL VELOCIMETRY Like LDA or PIV, DGV also measures the velocity of tracer particles which need to be added to the flow. With one orientation of the laser light sheet and one direction of observation, one component of the flow velocity is measured. DGV takes advantage of the fact, that the frequency of the scattered light is shifted due to the Doppler effect: ∆ν=ν-ν0 (1) ν0 : Laser frequency ν : Scattered light frequency ∆ν : Doppler shift This shift depends on the particle velocity v r , the light sheet direction l r and the direction of observation o r : ( ) v c l o o r r r − ν = ν ∆ (2) The basic idea of DGV is to measure the scattered light frequency ν by transmitting the scattered light through an iodine cell (Fig. 1). Iodine has strong absorption lines, which are used as a frequency to transmission converter. These lines interfere with the 514 nm line of the Ar laser as well as the 532 nm line of the frequency doubled Nd:YAG laser. Assuming the frequency ν to be on the slope of one absorption line, then ν can be determined by measuring the iodine cell transmission of the scattered light. Therefore, two detectors are required to measure the light intensity before and after the cell. To correlate ν and T, the transmission profile T(ν) of the iodine cell must be known. l o ref sig v v v v T I I − ⇒ ∆ = − ⇒ = v , 0 Fig. 1: Doppler Global Velocimetry setup and transmission profile of the iodine cell. Direction of the measured velocity component depending on the direction of laser light propagation and the direction of observation. The laser frequency ν0 has to be known and precisely stabilized, so that the Doppler shift ∆ν can be calculated according to equation (1). With equation (2), one component of the vector v r can be calculated. It is the component in the direction of l o r r − , the bisector of the angle formed by the direction of the laser light and the direction of observation (Fig. 1). At a scattering angle of 90° a velocity of 1 m/s corresponds to a frequency shift of 2.7 MHz. Since the frequency width of the slope of the absorption line is between 300 to 600 MHz (depending on the operation conditions of the gas cell), the dynamic range of velocity measurement is between 100 and 200 m/s. Another basic idea of DGV is to use two CCD-cameras as detectors, both watching the same section of a laser light sheet. By pixel wise division of the two pictures and further post processing a map of one velocity component in the light sheet is obtained. Depending on the type of laser (cw or pulsed), the result is either a time averaged or a frozen velocity image. The second and the third velocity components can be measured by changing the arrangement of the optical setup. There are two alternative ways to accomplish this: signal image light sheet