Measurement of temperature of drops when in flight
by
Francois Feuillebois
PMMH, ESPCI, 10 rue Vauquelin, 75005 Paris, France
Coauthors: F. Hervy

Measurement of temperature of drops
when in flight

F. Hervy ^{(1, 2)} and F. Feuillebois ^(1)
(1) PMMH, ESPCI, 10 rue Vauquelin, 75005 Paris, France
⁽²⁾ CEPr, Saclay, 91895 Orsay Cedex, France

Simulation tests of aircraft icing have been usually performed with clouds of supercooled water drops whose median volume diameter was about 20 \mum. For these sizes, drops were assumed to be at the static temperature of the flow. There is a recent interest in simulating freezing drizzle (diameters up to 500 \mum) and it is necessary to measure drop temperatures before impaction on obstacles because the cooling time increases with drops size and may become larger than the transport duration of drops.

There exist some techniques which can be used for droplet temperature measurement, like rainbow angle measurement, dual burst technique or extended phase doppler anemometer []. These techniques are applicable only if the refractive index variation of the liquid varies significantly with temperature. However, this is not the case for water from the supercooled state (around -40 ^o C) up to 50 ^o C. So, we choose another method based on Laser Induced Fluorescence (LIF) of a dilute organic dye.

Guilbault [] has pointed out that fluorescent molecules like rhodamine B are very sensitive to a temperature variation (temperature quenching), and Nakajima et al. [] have used this organic dye to measure the temperature of a liquid flow by Planar LIF (PLIF). We show here that LIF could be employed for temperature measurement of a liquid phase dispersed in a gas. A theoretical analysis using quantum mechanics, geometrical optics and generalized Lorenz-Mie Theory (GLMT) yields, for low rhodamine B concentration C and a drop diameter D smaller than the beam diameter w₀, the following expression for the intensity of fluorescence emitted by a drop crossing a laser beam at a distance y from the axis :

If=KoptKspectI0CD3e[(\beta)/T]e-([(\surd2y)/(w0)])2 ,

(1)

where I₀ is the incident intensity on the beam axis, T is the volume average drop temperature (in K), K_opt and K_spect are constants depending on the optical setup and characteristics of fluorescent molecules, respectively. The constant \beta has to be determined by calibration.

A validation of equation (1) was performed with a simple experimental setup. A monodisperse injector, filled with a water solution containing 10^-4 mol.l^-1 of rhodamine B, was placed vertically and produced periodically droplets with a nominal diameter of 92 \mum. The initial temperature of droplets was known and adjustable. Tests on polydisperse sprays were done using the injector in some peculiar modes. Falling drops were illuminated by a 60 mW Ar-ion laser (with \lambda = 514.5 nm) and fluorescent red light was collected on a photomultiplier after passing through a pass band filter centered at 600 nm (maximum peak of rhodamine B fluorescence). Signals were then analyzed on an oscilloscope.

The results confirmed that the equation (1) is correct for temperature above 293 K. Although accuracy is limited here to 0.5 K due to thermal probe uncertainty, the measured value of coefficient \beta indicated that the accuracy on temperature measurement could reach 0.1 K.

References

[]

Riethmuller M. L. Développements récents dans les techniques de granulométrie-laser. In 4^e Congrès Francophone de Vélocimétrie Laser, Poitiers-Futuroscope, 1994. Paper 1.1.

[]

Guilbault G. G. Practical Fluorescence: Theory, Methods and Techniques. Dekker, 1973.

[]

Nakajima T. Utsunomiya M. Ikeda Y. and Matsumoto R. Simultaneous measurements of velocity and temperature of water using ldv and fluorescence technique. In 5^th Int. Symp. on Appl. of LASER Tech. to Fluid Mech., Lisbon, 1991. Paper 12.1.

Date received: January 13, 1999