【作者】 殷玉梅；

【导师】 赵德有； 洪明；

【作者基本信息】 大连理工大学 ， 船舶与海洋结构物设计制造， 2015， 博士

【摘要】 船舶是进行水上运输和海洋开发的重要工具,船舶在航行过程中总会受到主机、螺旋桨和波浪等外界激励的作用。船体结构在这些激励的作用下,会发生不同程度的振动甚至有害振动。过大的振动可导致结构发生疲劳破坏,影响人员的身体健康和仪表设备的正常使用。为了避免船舶发生有害振动,需要在设计阶段较准确地预报船体结构的固有频率及螺旋桨脉动压力。本文主要针对船体振动固有特性预报及螺旋桨脉动压力预报进行研究,论文主要完成以下几部分工作：1)对船舶结构上的设备建模方法进行研究,分别采用集中质量模型、杆模型和空间结构模型来模拟船舶结构上的设备,通过对几种常用模型的分析,在杆模型的基础上利用刚度等效法提出了一种较为实用的等效模型。参照上层建筑对船舶总体振动固有频率影响系数计算公式的形式,给出设备对船体局部振动固有频率影响系数的计算公式。数值算例结果表明,采用本文给出的等效模型和设备对局部振动固有频率影响系数计算公式可以较准确地计算船体局部振动固有频率。2)对船舶上层建筑整体纵向振动固有频率预报进行了系统地研究。首先对船舶上层建筑整体振动有限元建模方法进行研究,讨论了不同计算模型、边界条件、附加水质量以及装载情况对上层建筑整体纵向振动固有频率的影响,对于上层建筑整体振动有限元分析得到了一些有益的结论,并给出了较为实用的计算模型。考虑更多的影响因素,提出了一种新的船舶上层建筑整体纵向振动固有频率预报公式,重点研究了上层建筑刚体回转振动固有频率计算公式中等效刚性系数的取法,给出了等效刚性系数曲线；并对剪切振动固有频率计算公式进行了修正,得到考虑弯曲振动影响的修正系数计算公式。智能方法具有很强的辨识能力,能够很好的处理不确定性和非线性问题,可以应用智能方法预测船体振动固有频率。支持向量机(Support Vector Machine, SVM)与其它智能方法相比具有小样本学习、全局寻优和泛化能力强等特点,能较好的解决小样本和非线性等实际问题。本文应用SVM对船舶上层建筑整体纵向振动固有频率进行预测,取得了比较好的预测效果,为船舶上层建筑整体纵向振动固有频率预报提供了一种新思路。3)对于船舶总体垂向振动固有频率预报,提出了一种新的船舶总体垂向振动固有频率预报公式。以Timoshenko梁理论为基础,引入剪切系数来修正剪力均匀分布假设,得到两端自由等直梁固有频率计算公式；根据大量船舶总体垂向振动固有频率实测数据,按照两端自由等直梁固有频率计算公式的形式,采用统计分析方法,得到船舶总体垂向振动固有频率计算公式,为船舶总体垂向振动固有频率预报提供了一种新方法。应用SVM对船舶总体垂向振动固有频率进行预测,考虑到影响船舶总体垂向振动固有频率的参数较多,利用灰关联分析理论建立船舶主要参数和垂向振动固有频率之间的灰关联模型,通过计算各参数的灰关联度,来确定各参数对船舶总体垂向振动固有频率的影响大小；将关联度大的参数作为SVM的输入数据,取得了比较好的预测效果,为船舶总体垂向振动固有频率预报提供了一种新思路。4)采用CFD方法对螺旋桨诱导的船体表面脉动压力进行预报。对船体和螺旋桨进行整体建模,基于Workbench平台进行双向流固耦合分析,计算得到船体表面脉动压力幅值。将CFD方法计算得到的螺旋桨诱导的船尾脉动压力幅值与试验值进行比较,两者结果较为接近,采用CFD方法可以较准确地预报螺旋桨诱导的船体表面脉动压力。船体振动预报涉及到结构动力学、弹性力学和流体力学等多个学科领域的交叉,本文对船体振动预报进行了较系统地研究,为船体振动固有频率预报提供了一些新方法和新思路。更多还原

【Abstract】 Ship is an important tool for water transport and ocean development. During the voyage, ship will be encountered with the external excitations from main engine, propellers and wave loads which can cause different degrees of vibration and even ship structural damage. The excessive vibration may lead to structural fatigue failure, affecting the health of human-beings and the normal use of instrument and equipment. To avoid the harmful vibration, the natural frequency of ship structure and propeller-induced fluctuating pressure are required to make accurate prediction.This paper aims to do research on prediction of natural characteristic of ship hull vibration and propeller induced fluctuating pressure, including:1) The equipment modeling method for ship’s local vibration is studied and lumped mass model, rod model and space structure model are respectively used to simulate the equipment of ship structure. Through the analysis of several common models, a new practical model is proposed based on rod model using rigidity equivalent method. Deal with the influence of equipment on ship’s local vibration according to the solution to conduct the influence of superstructure on ship’s overall vibration and provide the calculation formula of influence coefficient of equipment on natural frequency of ship’s local vibration. Numerical examples show that natural frequency of ship’s local vibration can be accurately calculated using the equivalent model and the calculation formula of influence coefficient given by this paper, which is of great practical significance.2) The prediction for natural frequency of superstructure’s overall longitudinal vibration is made systematically. The finite element modeling method of superstructure’s overall vibration is investigated and the influence of various models, boundary conditions, added mass and loading conditions on the natural frequency of superstructure’s overall vibration is discussed in this paper. Some meaningful conclusions for finite element modeling of superstructure’s overall vibration are summarized and a practical computing model for the calculation of natural frequency is also proposed. Considering more influences, the prediction formula for natural frequency of superstructure’s overall longitudinal vibration is derived. The picking algorithm of equivalent stiffness coefficient for calculating the natural frequency of superstructure with rigid-body rotation vibration is studied, drawing the curve of equivalent rigid coefficient. The formula for calculating the natural frequency of shear vibration is amended, and the correction coefficient formula of bending deformation is proposed. Intelligent method possesses many characteristics such as strong ability of identification and it can also handle the uncertain and nonlinear problems well. Therefore intelligent method can be used to predict the natural frequency of ship vibration. Compared with neural network and other intelligent methods, Support Vector Machine(SVM) which has many characteristics such as small samples study, global optimization and better generalization can handle small samples, nonlinear and other practical problems well. Natural frequency of superstructure’s overall longitudinal vibration is predicted by SVM. The proposed method in this paper is proved to be accurate and feasible, which provides a new idea for the prediction of natural frequency of superstructure’s overall longitudinal vibration.3) In terms of the prediction of natural frequency of ship’s vertical vibration, a new empirical formula is given. Based on the Timoshenko beam theory, the formula is derived for a uniform beam with free-free boundary condition by introducing shear coefficient to correcting the hypothesis of shear uniform distribution. Referring to large numbers of measured data of natural frequency of ship hull’s overall vertical vibration, a new formula for calculating the natural frequency of ship hull’s overall vertical vibration is obtained by statistical analysis based on the formula for a uniform beam with free-free boundary condition, offering a new method for predicting the natural frequency of ship hull’s overall vertical vibration. Predict the natural frequency of ship hull’s overall vertical vibration using SVM. Considering that the parameters which influence the natural frequency of ship hull’s vertical vibration are much more, a grey relation model between ship’s main parameters and natural frequency is established by grey relational analysis. Determine the influence of the parameters on natural frequency of ship hull’s vertical vibration by computing the grey correlation degree of each parameter and the parameters with greater correlation degree are used as input data in SVM. The predictions in this paper are proved to be accurate and valid, which provides a new method for the prediction of natural frequency of ship hull’s overall vertical vibration.4) Fluctuating pressure induced by propeller is predicted using CFD method. In this paper, propeller-induced fluctuating pressure on the hull surface is calculated using the analysis of two-way fluid-structure interaction based on Workbench platform. The pressure curve on the hull surface over time can be achieved directly by overall modeling with the hull and propeller. Then propeller-induced fluctuating pressure of stern can be gotten by Fourier analysis and make comparisons between the propeller-induced fluctuating pressure of stern calculated by CFD method and the experimental results. The two results are very close which indicates the method proposed in this paper can accurately predict propeller-induced fluctuating pressure on the hull surface.Vibration prediction of ship involves several interdisciplinary research including structural dynamics, elastic mechanics, fluid mechanics and so on, which is of great importance in theoretical and practical research. This paper does systematic research on vibration prediction of ship, which provides a new method for prediction of natural frequency of ship vibration.更多还原