Experimental Investigations into Progressive Interfacial Waves

The majority of experimental investigations into wave motion have been concerned with surface gravity waves rather than internal waves. This is because surface waves are easier to produce and measure in a wave tank and easier to measure in a field experiment. This is for a number of reasons.

1. It is easy to fill a tank, with a single fluid, in order to study surface waves. It is more complicated to produce a desired density distribution in the tank. The following methods can be used to produce a density difference.
   1. Two different liquids can be used. This has the drawback that the surface tension between the two liquids can act to damp the waves. This does not happen in the ocean or the atmosphere where the density difference is produced by a thermocline or a salinity change.
      
   2. A change in the density can be introduced by using a saline solution for the denser fluid and pure water for the less-dense fluid. Such a configuration is stable but the tank will need to be emptied and refilled at regular intervals. This can be very time consuming particularly if more than two layers are being used. Each layer, with its own salinity, must be allowed to settle before the next layer is added. The fluid must enter the tank at a slow speed so that it does not mix with the other layers.
      
2. Surface waves can be produced in the laboratory using a paddle system at one end of the tank, the movement of the paddle initialising the wave motion, which then propagates along the tank. If interfacial waves are to be produced the wavemaker normally moves in a vertical plane and the mean position of the wavemaker is matched to the interface height.
   
3. The high rate of damping observed in interfacial waves means that the measurement area must be close to the production area to prevent the wave being excessively damped. This is not a problem with surface waves.
   
4. Measurements of the wave amplitude can be made, in a laboratory or in a field experiment, either using a floating buoy or using a wave gauge which measures the resistance, or capacitance, between two metal probes and relates it to the immersed length of the probes. More complex methods need to be used for internal waves.
   

Early investigations were carried out [85, 86] by dying one of the fluids and making photographic records of the waves in the tank. This allows the shape of the interface to be measured. The movement of the wave down the tank can also be found. To obtain information about the wave velocities it is preferable to use a non-intrusive measuring technique. Until relatively recently the main technique for measuring the velocity in a multi-layer fluid has been laser Doppler anemometry (LDA) [87, 88]. LDA can be used to measure the velocity of a fluid at a single point were two laser beams intersect. The main drawback of the LDA technique is that it can only make a measurement at one position in space. To record velocities at different positions the apparatus must be realigned to each position. The flow must be repeatable since each measurement is being taken at a different time. There are additional problems associated with LDA applied to a multi-layer systems because of the change in the refractive index within the flow [89]. This can be overcome, to some extent, if the density difference is small. Hannoun et al. [90, 91] suggest that, provided a suitable solvent is used, the refractive indices can be matched between the layers of a two layer system for density differences of up to 2%. The matching technique allows LDA measurements to be made.

A recent development in measurement technology was the development of particle image velocimetry (PIV) [92]. This uses a pulsed bean laser or a continuous laser beam which is scanned across the fluid, illuminating a vertical plane. Small, neutrally buoyant particles are suspended in the fluid. The laser light reflected from the particles is recorded on film, each frame containing several images of each particle from successive pulses or scans of the laser. This information can then be analysed to produce a two-dimensional velocity map in the plane which was illuminated by the laser. PIV has been applied widely to surface gravity waves [93] but has only recently been applied to internal waves. One of the difficulties encountered in applying PIV to internal waves is the presence of zero and low velocities where the suspended particles move an insignificant distance, relative to their size, between exposures. This is overcome using an image shifting technique [94], this also eliminates any directional ambiguity in the velocity. Such a PIV system was employed at Edinburgh University to produce the experimental results [10] which will be compared with the simulations in the rest of this chapter.