Experimental and numerical investigation of the internal kinematics of a surf-zone plunging breaker

Abstract

Over the last couple of decades both the qualitative and quantitative understanding of breaking waves in the surf zone have greatly increased. This is due to the advances in experimental and numerical techniques. However, few comparisons between these two different investigative techniques for surf-zone breaking waves have been reported. In this study, a comparison is made between the experimental and numerical investigation of the internal kinematics of a surf-zone plunging breaker. The full-field velocity measuring technique known as Particle Image Velocimetry (PIV) is used in the experiments. In the hybrid numerical scheme, the main model solves the NavierStokes equations using a Finite Volume method and the free-surface is simulated using a Volume of Fluid (VOF) method. An important feature of this work is that, unlike in most other comparisons between numerical and experimental results, the exact geometry of the physical wave flume and the exact motion of the physical wavemaker are duplicated in the numerical wave tank. To achieve this, an additional numerical model using a Boundary-Integral Method (BIM) is employed to generate the input conditions for the NavierStokes solver. Very good agreement was found for all comparisons: free-surface elevations, velocity vector maps, velocity profiles and velocity-magnitude contours. However, some small discrepancies were observed. In the free-surface elevation comparisons, a slight time lag was observed in the numerical results and it is suggested that this was due to the small amount of smoothing applied in the BIM to enable it to continue to supply input data to the NavierStokes solver well beyond the breaking of the wave. In addition, some small differences were also found between the numerically predicted velocity distributions and those measured in the experiments. These disagreements occurred mostly in the aerated region and it is proposed that they could be caused by errors in the PIV velocity data due to air bubble effects. However, they could also be attributed to the fact that no turbulence model is used in the numerical scheme and it is these aerated areas where the turbulence levels are the highest. © 2011 Elsevier Masson SAS. All rights reserved.