Non-linear unsteady wave loads on large high-speed wave piercing catamarans

TLDR: In this article, a reverse engineering approach is introduced to predict slamming loads during sea trials by using the capabilities of Finite Element Analysis (FEA) using the well known software PATRAN/NASTRAN1.

Abstract: The current work investigates the slamming characteristics of wave piercing catamarans through the analysis of sea trials data of the 98 m Incat sea frame “Hull 061”, built in Tasmania, Australia and currently serving in the US Navy combat fleet. The importance of this sea trials data is that the ship was tested in severe sea conditions to assess her suitability for military operations and to define her operational envelope. New signal processing techniques such as Wavelet Transforms are used in analysing slamming data for two main purposes, slamming identification and modal analysis in time and frequency domains simultaneously. The Wavelet Transforms were found superior to conventional signal processing tools such as Fast Fourier Transform and Short Time Fourier Transform. The structural strength of wave piercing catamarans is studied by introducing a novel sea trials analysis for structural performance assessment in an attempt to simulate real loading conditions. The methodology was tested on normal linear wave loading (without slamming) and was found satisfactory. A “Reverse Engineering” approach is introduced to predict slamming loads during sea trials by using the capabilities of Finite Element Analysis using the well known software PATRAN/NASTRAN1. To increase the efficiency of this approach, the load parameters, spatial location and distribution, were investigated through model tests of a similar but larger 112 m Incat hydro-elastic model in the Australian Maritime College towing tank facility. Based on pressure measurements, proper slam load models can be more accurately and efficiently introduced in the finite element analysis. Quasi-static analysis was first performed to examine its suitability to analyse such fast time varying loads. Difficulties in comparison procedures between numerical simulations and trials data have strongly highlighted the need for dynamic analysis. Direct transient dynamic analysis was performed using the dynamic solver of the same software package. Good agreement with trials data was found. The suggested procedure and slamming loading patterns used in the numerical simulation is then verified and can be regarded as a solid base for verification of other theoretical design models.