chapter_01_1_exp_desc.md

Experimental Setup

{ #fig:scheme-coriolis width=52%} {#fig:exp-top width=45%}

{#fig:movement width=50%}

Experimental setup

The Coriolis platform is an experimental facility at LEGI, Universit\text{\'e} Grenoble Alpes, which is 13 metres in diameter and is mounted on motors allowing the entire platform to rotate. Both rotating and non-rotating experiments were conducted in this project. The platform was filled with a linear stratification of salinity up to a height of 81 cm using a double-bucket method. A nine metre long and six metre wide rectangular enclosure, as shown in @fig:scheme-coriolis, contained most of the equipment, including the oscillating carriage which forces the fluid.

Carriage

Anticipating a forward energy cascade [@riley_dynamics_2003;@Lindborg2006], we chose to force at large scales. In the literature, various approaches can be found to inject KE into the system. For instance in @AugierBillantNegrettiChomaz2013 a series of flapping plates stationed along the boundary are used to generate dipoles. A drawback of this method is that it also produces smaller scales of motion in comparison with the size of the plate. In @PraudFinchamSommeria2005, a rake consisting of series of flat plates were towed through the fluid which also has the same drawback. Here we replaced the rake with a carriage attached with a comb of six vertical cylinders, with diameter 0.25 metre, attached to a motor-driven carriage towed through the fluid. The oscillating comb injects kinetic energy in the form of eddies as shown in @fig:eddies. The wake of the comb has a characteristic length scale of the cylinder diameter. In contrast to @PraudFinchamSommeria2005, who carried out decaying turbulence studies, here we make measurements while the carriage oscillates back and forth along the rectangular enclosure. A chronogram of the movement is shown in @fig:movement. Several experiments with different levels of stratification ($N$) and carriage velocity ($U_c$) were conducted. In order to characterize the experiments we define a Froude number and a buoyancy Reynolds number, $$ F_{hc} = \frac{U_c}{NM},\quad \R_c = \frac{U_c^3}{\nu N^2M} $$ \noindent respectively where $N$, the Brunt-\text{V"ais"al"a} frequency is evaluated a priori while the fluid is quiescent and $M$ is the characteristic size of the vortices.

Measuring instruments

Density probes

Five conductometric probes calibrated for ambient temperature conditions and the expected salinity range were installed at different locations, indicated with blue dots in @fig:scheme-coriolis. The probes $P_p^1$ and $P_p^2$ are mounted on vertical profilers and these dive into the fluid at regular time intervals at a moderate speed, providing us with accurate one-dimensional estimates of the stratification. Probe $P_c$ is attached on the carriage, at a constant height of 0.395 meters. Since $P_c$ makes measurements along the direction of the carriage oscillation, the data output from this probe is decomposed and interpreted differently. When it is ahead of the carriage it measures decaying turbulence and when it is in the wake of the carriage it measures forced turbulence. $P_f^1$ and $P_f^2$ are stationary probes which are placed at the top and bottom of the fluid to measure temporal evolution of the mixed layer.

Particle Image Velocimetry (PIV)

{#fig:exp-side width=67%}

{#fig:eddies width=73%}

Visuals from a test experiment with the laser sheets illuminating the flow.

Two sets of PIV techniques are applied in this setup. Firstly, a two-dimensional, two-component (2D-2C) scanning PIV system is used for measuring velocities along quasi-horizontal planes with a maximum tilt of 1.5$^\circ$ about the true horizontal plane. This system relies on a single laser sheet deflected by an oscillating mirror controlled by a servo motor. For the same experimental parameters two sets of measurements are carried out, one with five horizontal planes and small scan amplitudes to obtain fine vertical resolution, and another with typically eight planes and wider scan angles which tend to be vertically decorrelated. The images are captured by three cameras, one placed at a high altitude capturing a large field ($2.2 \times 2.5 \text{ m}^2$), a second one placed at an intermediate altitude capturing a smaller field ($1.3 \times 1.0 \text{ m}^2$) and a third one placed under the platform capturing an even smaller field ($1.18 \times 0.53 \text{ m}^2$). The fields of vision are demarcated in pink in @fig:scheme-coriolis. The cameras are placed progressively closer to the fluid to get more illumination from the particles seeded and potentially increased resolution. Secondly, the velocity field in the vertical plane is measured using a two-dimensional, three-component (2D-3C) stereoscopic PIV system. This system captures a field of area $0.45 \times 0.45 \text{ m}^2$. For both the PIV systems, up to 1300 images per oscillation of the carriage are captured.