PIV (Particle Image Velocimetry) measures a complete velocity field using short interval images [15] [16]. This technique can be used to study a gas or a liquid. PLIF (Planar Laser-Induced Fluorescence) is a technique widely used in the study of fluids. This technique makes it possible to observe the flows but also the fields of concentration and temperature in a two-phase flow. The luminous excitation of fluorescein and crystals is done using a Laser Argon (Light Amplification by Stimulated Emission of Radiation) Figure 2. This Laser was chosen because it is capable of emitting at several lengths of wave between 457.9 - 514.5 nm. Fluorescein as the crystals are excited at about 500 nm, wavelength corresponding to the blue-cyan color, a filter is placed at the output of the laser beam to have only one wavelength emitted on the tank [17]. However, this filter will lose 80% - 90% of the power of the Laser, but it does not prevent the fluorescein and the crystals are excited because one of the ideas used to solve this problem is to place a reflective mirror at opposite side of the Laser to double the power output. Note also that the Laser must be cooled due to a large heat emission during operation. The flow rate of the cooling system is adjusted to around 10 L/min, thus avoiding instability of power and temperature. The normal power of use of the Laser rotates around 8 W. The intensity can be adjusted to optimize the exposure time at the camera and to have better images.

In order to study the behavior of bubbles in different solutions, we chose to take images with different flow rates and needle diameters for four glycerin concentrations. In our experiments, three different types of diameter were used: Æ0.8 mm; Æ0.6 mm; Æ0.4 mm. For PIV, Dantec dynamics particles were used. These particles are spherical crystals HGS-10, Hollow Æ10 μm. They are used at a wavelength around 500 μm. Four 20 L solutions were used in our experiments, which are defined by their glycerin percentage by volume. For the demineralized water solution and for the 20% solution of glycerin, solutions with a low viscosity,

we find that the rise of the bubbles takes a spiral trajectory (3D) Figure 3 when injecting low flow rates. On the other hand, the bubbles rotate about themselves and each of them has its own trajectory. For solutions of 30% and 40% glycerin we can notice that the rise of the bubbles follows a zigzag trajectory but in the case of the solution 30%, this trajectory is in 2D Figure 3 whereas for the solution 40% it is in 3D and less marked (Figure 4). In this last case, we can also see

that the upstream bubble crashes against the downstream bubble, the one coming out of the plane in which it was and can follow a new trajectory with a different speed. The terminal velocities in Figure 4 have been calculated. It is noted that this decreases exponentially, the higher the concentration of glycerin increases.

In order to study the dissolution of CO₂, we use fluorescein. Indeed, when the concentration of CO₂ increases, the pH decreases and the re-emission of fluorescein photons also appears in Figure 5. Fluorescein is a dye which has the particularity of returning a fluorescent light when it is excited at a wavelength of 494 nm [18].

It is observed that the maximum is reached is reached at around 5.10 - 7 mol/L (Figure 6). For a water depth of 40 cm or 18 L, we used 100 mL of fluorescein at a concentration of 0.034 g/L. In order to exploit PLIF results, it is necessary to make a calibration. Thus, a series of images is produced at different concentrations and the corresponding light intensity and the pH are measured. It is therefore possible to obtain curves that bind the pH, the light intensity and the CO₂.

• Camera: Exposure time: 2500 μs, Frequency: 206 FPS, Image size: 1399 × 500, Dynamic gain: 3000 DN

• Particles: When brightness is high, it is best to use 10 μm glass particles. In the case of a lack of brightness, polyamide particles of 50 .mu.m are used.

• Pump flow syringe: 100

• Needle diameters: Æ0.4; Æ0.6 and Æ0.8.

• Column filled up to 40 cm with 17.96l of demineralized water and 0.1l of fluorescein.

• Gas: 99.9% CO₂ of liquid air.

Figure 7 shows the average speed fields. These measurements were taken for air flow rates equal to Q 60 = 2.548 cm³/min and a Æ0.8 mm.

Note that the average displacement value of the liquid in the column is a function of the viscosity of the liquid used. Indeed, for a fairly viscous liquid (40% glycerin) the speed undergoes an increase of about 50% compared to that of pure water.

In order to characterize the distribution of the velocity field in the whole of the studied volume, the transverse profiles in different planes of the column were determined (Figures 8-11). The results show that these profiles are more and more flattened by decreasing the viscosity. This is mainly due to the value of the volume density of the liquid which increases as a function of the viscosity and consequently the displacement of the gas bubbles cannot cause a displacement on the totality of the liquid contained in the column. It is for this reason that we notice quite significant speeds around the central axis of the column for the most viscous solution.

In Figure 4, it was noted that bubble velocities decreased as glycerin concentration increased. However, when we observe the transverse velocity profiles, we notice that the velocity of the fluid increases with the concentration of glycerin. This is because the slip speed decreases with increasing glycerin concentration. As a result, the fluid is more driven by the bubble when there is viscosity, which explains the higher fluid speeds.

For lack of time we could not take quality picture series for all needles. However, the programs are now functional and the protocol is clearly established. Here is the result for a Æ0.6 mm needle. The MATLAB program processes only one image. Here are the results of the different stages of treatment. Figure 12 shows the raw image of the PLIF and PIV and Figure 13 shows the velocity field result processed but Figure 16 shows the superimposed result PIV-PLIF deals with Davis 8.1.

The particularity of our work is the use of a doubler to acquire PIV and PLIF. This allowed us to follow the dissolution of CO₂, determine its concentration field, and then visualize the velocity profiles. But also to superimpose the two images PIV and PLIF.

During our work, we found that at t = 0, there is no bubble on the water table Figure 14(a), but at t = 0.30 s we observe the first bubble on the water table then at 0.43 s it goes out of the Laser Plan Figure 14(b).

Bubble columns have long been used in water treatment to capture different types of particles. More recently, it is their exchange capacity that has been used to produce chemical reactions between gas phase and liquid phase [13]. The vacuum bubble column studied dissolves carbon dioxide in water (Figure 14 and Figure 15) [7]. This carbon-rich water is an ideal medium for growing micro-algae. The production of these micro-algae is today an issue in many fields since it can be intended for medicinal, food or energy use in the form of biofuel. It is therefore important to understand the phenomena and parameters at stake before scaling up (Figure 16).

These theoretical results of PLIF and PIV will be applied to the culture of a blue-green algae: Spirulina Platensis. The choice of this plant-bacterium is based on several criteria: short culture cycle, applications in cosmetics, medicine and high nutritional value. Thus, we were able to study the influence of culture parameters such as light intensity, CO₂ velocity and concentration field, and temperature on its growth. In addition, we have implemented two methods based on bubble velocity and CO₂ dissolution to quantify the biomass produced by spirulina.