We have developed methods to measure flow velocities in metallic liquid gallium. We use the DOP 1000 ultrasonic Doppler apparatus from Signal Processing. Our test configuration consists in a cylinder filled with gallium. The cylinder is 80 mm in diameter and 130 mm high. A 40 mm-diameter crenelated disk spins at the top of the cylinder, with angular velocities up to 3000 rpm. 8 mm-diameter ultrasonic transducers are placed on the outside walls of the cylinder, on flat portions machined at some angle from tangent. We therefore retrieve profiles of the velocity along a line of sight that is a cord of the cylinder.

We have obtained very nice such profiles for various disk velocities. The observed bell-shaped profile is consistent with the expected flow pattern. The maximum measured velocity is proportional to the disk velocity. The profiles are perfectly superposable with profiles measured with water in the same cylinder. A comparison with velocities derived from streak photographs in water shows very good agreement.

In order to obtain profiles in liquid gallium, we have had to overcome a number of technical problems. The main problem is due to the very fast oxidation of gallium. Apparently, oxides tend to cluster on the walls of the container, thereby making it impossible for 4 MHz ultrasonic waves to probe the inside. The first step to overcome this problem is to clean the gallium with an HCl-ethanol mixture. However, once oxides stick to the walls, it is very difficult to get rid of them. We therefore had to give up using polycarbonate, nylon, and copper for the cylinder. Good and lasting results were obtained using copper coated with a cataphoretic black deposit.

We will report on our latest results and recipes.