View non-flash version
en-US en-US en-USJuly 2012 en-USwww.sname.org/sname/mt e controlling factor in any acoustic prediction is having accurate noise and vibration (structureborne) levels for the machinery and propulsors. Typical ship -board machinery includes propulsion diesels and gas turbines, diesel and gas turbine generators, compressors, pumps, transformers, bow/stern thruster, hydrau -lic power units, fans and other auxiliary equipment. These data should span the acoustic range of interest, which is 31.5 to 8000 Hz in nine octave bands. Measured acoustic data for the equipment (or from a piece of equipment similar in type, power, and rpm) is one source for this information. Vendor-supplied data are another source, especially if the data were collected to a reg -ulatory standard such as ISO 10816 or ANSI S2.16-1997. Empirical prediction methods are available from SNAME?s De sign Guide for Shipboard Airborne Noise Control, T&R Bulletin 3-37. Other inputs needed for the sources are their physical size, weight, and location within the vessel. Other signicant sources are the hydro-acoustic excitation from propulsors and thrusters and bow/ wave impact. Empirical methods have been used successfully to dene these acous -tic sources and their interaction with the ship 3-D model. For a propeller, typically the tip speed, cavitation inception speed, and number of blades control the acoustic source level. The model also should reflect treat -ments such as absorption; joiner bulkheads placed over structural bulk -heads; isolation mounts for equipment; and damping treatments. Again, measured or vendor-provided treatment acoustic perfor -mance data should be used primarily. As with the acoustic sources, SNAME?s T&R Bulletin 3-37 has predicted treatment per -formance information. Given the topology, ship?s structural details, mechanical and hydro-acous -tic sources, and location of sources and locations of interest, accurate noise predictions can be provided by the appro -priate SEA software. FEA modeling considerations Vibration analysis of a ship typically consists of two analysis components: hull natural fre -quency and forced response. ese two can be performed using a single FEA model with minor modications. is model will have signicant dierences with a model devel -oped for stress applications, as described further on, but standard FEA practices such as the use of quality (not skewed, warped, or of high aspect ratio) quadrilateral elements where possible still apply. A picture of an FEA model produced for vibration model -ing purposes is shown in Figure 2. e rst and most important modeling considerations for FEA models are the ele -ment size and type to be used. e plating Figure 2: FEA model of a ferry, produced for vibration mo deling purposes. e controlling factor in any acoustic prediction is having accurate noise and vibration (structureborne) levels for the machinery and propulsors. Approach ASound