A NUMERICAL ASSESSMENT OF SCALE EFFECTS ON WAVE BREAKING MODELLING

Scale effects are a key issue in hydraulic laboratory modelling and specially so in wave action studies; it is therefore only natural that a great deal of work in the past has gone into evaluating the different physical phenomena which concur to the difference between test and real life coastal applications. For instance Tirindelli et al. (2000) reviewed the relative influence of such diverse effects like viscosity, surface tension and compressibility. It is normally assumed that Froude number is the key parameter in wave laboratory experiments, so that results from a model can be safely scaled up if such a parameter is kept constant; this, of course, implies – among other things – that the Reynolds number, strongly dependent as it is on the physical size of the experiment, only plays a minor role. This assumption should be properly verified, and the present paper provides some insight in this regard. The examples considered in the paper are all related to the transformation of waves over a shallow beach, an important problem from the application point of view and one which has been often been studied experimentally and numerically. Indeed, thanks to recent developments, the numerical simulation techniques of wave dynamics are now an efficient tool to enquire into the effects of spatial scale on the parameters that are of interest to the design of coastal structures. Numerical Navier Stokes integration with VOF surface algorithms, first developed by Lin and Liu‘s (1998) has now evolved into a fully reliable technique (see for instance Christensen 2006). Also, innovative Lagrangian SPH methods are quickly evolving, and it is most likely that they will soon be able to compete with Eulerian methods (Dalrymple et al 2005; Shao et al 2006), even though at present there is not enough available experience to support their application to real wave action problems. Given this, a well proven Navier Stokes Eulerian code has been used for this work, after careful calibration with published results; besides, since the problem being dealt with is basically the effect of viscosity, results have been checked against spurious computational effects linked to numerical viscosity by carefully studying the sensitivity of results to the mesh size. Results are presented for test cases taken from recent bibliography, computations being performed with the same geometry at different model scales but with the same Froude number. Parameters such as wave height, particle velocity and turbulent energy profiles are considered to assess the difference between real life and model scale testing