The Hydrodynamics of a Model of a Vibrating Cable

ABSTRACT Model experiments on a simulated submersible umbilical were conducted in a large water channel at Reynolds numbers appropriate to full scale cables. The investigation was designed to produce data for semi-empirical prediction methods of umbilical and submersible behaviour. Local values of tension, drag and vibration amplitude were derived as a function of cable shape. In all cases examined, complex large amplitude strumming motions were observed with attendant large increases in normal drag coefficient. INTRODUCTION Following an initiative to replace man underwater in the context of offshore operations, the Offshore Supplies Office of the U. K Department of Energy has supported R&D aimed at improving the design and operation of remotely operated vehicles (ROV's). As part of a programme to develop a generalised computer model for ROV's, NMI were commissioned to explore the hydrodynamics of the two main elements of the system - the submersible vehicle and its umbilical cable. This paper discusses the latter study. The drag of the cable in a current or when towed plays an important part in determining the overall system capability. A design difficulty exists if the hydrodynamic cable drag consumes a significant proportion of the available power. To increase the operating range more power is needed, but this in turn requires a larger diameter cable which in itself causes increased drag with the result that an improvement may not be forthcoming. The experiments reported here were conducted to provide data on local drag forces and motions at various points on cables of different geometry. The aim of local measurement distinguishes these experiments from previous investigations. Much of the existing relevant data has been reviewed by Griffin1,2, Every3, and Kline et al4, and they reveal the lack of information on cable drag and motion as a direct function of the local cable parameters of tension, inclination and curvature. The essence of the experimental method was the traversing of tension and acceleration transducers contained in the body of the cable along the span for a range of cable lengths and end tensions. The data produced was analysed in terms of cable geometry and vibration parameters to compare with the 'global' cable (e.g. Vandiver5) and rigid cylinder coefficients commonly used in numerical models (e.g. Skop et al 6). EXPERIMENTAL ARRANGEMENT The experiments were conducted in NMI' s No 2 water channel which is a closed return facility with a working section 18m long by 3.7m wide and a normal water depth of 2.1m. Figure 1 shows a view inside the channel during the installation of the model cable.