For design validation of offshore structures and conceptualisation of wave energy converters, physical model testing performed in wave basin laboratories is often applied. In such cases, knowledge about the wave conditions is of great significance. For validation of the wave condition in such tests, different methods for estimation of the directional wave spectra may be applied. However, different assumptions are imposed in the methods and deviations here from provide uncertainties in the results. The following paper quantifies the influence of nonlinear effects on the accuracy of the estimated directional wave spectra. This is done by analysis of idealised, synthetically generated waves based on second order wave theory and secondly with simplified amplitude dispersion included. The present analyses show that the uncertainties of the directional wave spectra are proportional to the level of nonlinearity present in the wave field.
To meet the increasing demand for renewable energy, the wave energy sector has been in a continuous development the past years. In the development of wave energy converters, it is crucial to know the conditions of the waves to assess the possible power take-off and determine the loads on the structures. Physical model testing is a commonly applied design and validation tool for offshore structures in general. In such tests, the structure is exposed to various short-crested sea states and the structural response is measured.
The generation of waves in laboratory tests can be performed from either Fast Fourier Transform-based approaches like the Random Phase Method or the Random Complex Spectrum, where the energy of the wave spectrum is introduced at discrete frequencies. Alternative, methods based on digital filters may be applied such as the White Noise Filtering Method which may produce continues spectra. To verify if the generated sea state matches the site-specific conditions, analyses of the wave field is required. A variety of methods have been proposed for the analysis of directional wave spectra, among which most aims to determine the Directional Spreading Function (DSF). This type of method includes for instance the Bayesian Directional Method (BDM) (Hashimoto and Kobune, 1988), the Maximum Likelihood Method (MLM) (Capon et al. (1976); Isobe et al. (1984); Krogstad (1988)) and the Extended Maximum Entropy Principal (EMEP) (Hashimoto et al., 1994). Such methods assume a double summation sea state, meaning that each frequency of the irregular wave field has several different directions. Miles and Funke (1989) however found that the double summation model is related with problems regarding phase-locking. The phase-locking causes spatial differences in the energy across the tank, thus a non-ergodic wave field, wherefore it is often preferred to use the single summation model for wave generation in laboratory tests.
