海洋音響学会誌
Online ISSN : 1881-6819
Print ISSN : 0916-5835
ISSN-L : 0916-5835
32 巻, 1 号
選択された号の論文の2件中1~2を表示しています
  • 土屋 健伸, 遠藤 信行
    2005 年 32 巻 1 号 p. 18-29
    発行日: 2005/01/01
    公開日: 2011/08/16
    ジャーナル フリー
    The numerical analysis method of sound propagation is developed in ultrasonic field because calculating performance of the personal computer improved in very high pace. Finite Difference Time Domain (FDTD) method is very useful method to calculate the sound propagation in the field, although it is requires very large memory and long calculation time. In order to shorten the calculation time of FDTD method, we proposed a new algorithm of parallel calculation using Message-Passing Interface (MPI).
    In this paper, we proposed a new algorithm of parallel computing using MPI in Personal Computer (PC) clusters to shorten the calculation time of FDTD method. PC clusters with 8 nodes in connected by gigabit Ethernet is used for calculation. In order to confirm the accuracy of calculation, we calculated sound propagation in homogenous media, and calculated propagation of sound pulse in shallow water. As a result, it is shown that FDTD method with parallel processing can be calculated accurately in sound propagation model. The calculation speed becomes about 6 times of normal personal computer in shallow water model. It is clear that FDTD method with parallel processing is very efficient method for numerical analysis of sound propagation.
  • 貞安 一廣, 安部 幸樹, 澤田 浩一, 高尾 芳三, 向井 徹, 飯田 浩二
    2005 年 32 巻 1 号 p. 30-39
    発行日: 2005/01/01
    公開日: 2011/03/02
    ジャーナル フリー
    Understanding the variability of fish acoustic target strength (TS) potentially improves the accuracy of fisheries acoustic survey. This study examines the effect of fish body bending by swimming activity on fish target strength. Bending shape of fish body and swimbladder were simulated from soft X-ray images of anesthetized walleye pollock. Target strength of bent fish was estimated using a Kirchhoff Ray-Mode model from simulated morphology as a function of fish pitch angle. TS patterns of whole fish and swimbladder did not change near the pitch angle at which maximum TS was obtained. But at the other pitch angle, TS patterns of bent fish differed from that of no bent fish. TS patterns of fish body changed at all the pitch angle. Decrease of maximum TS caused by fish bending was within 0.7 dB. Present study indicates that it is not necessary to take care of fish bending on TS estimation, when fish swimming behaviour is normal except for feeding time that fishes are more active.
feedback
Top