索杆张力结构刚度的动力监测理论研究

发布时间：2018-03-29 21:14

本文选题：索杆张力结构　切入点：动力测试法　出处：《浙江大学》2017年博士论文

【摘要】：最早建成的大跨度索杆张力结构服役时间已经超过30年,但目前此类结构在健康监测与性能评价方面的研究还不多,深入性的工作更是少见。索杆张力结构的刚度构成机理与常规结构存在重要区别,需要依靠预应力来提供结构刚度并维持稳定性。经验表明,索杆张力结构的设计一般由刚度验算控制,相比之下结构强度性能往往容易满足。由于担心结构刚度出现较大偏差,因此预应力也成为在役期索杆张力结构性能监测的重点。目前索杆张力结构实际工程中,预应力基本是通过直接测试索力来实施监测,但有限的索力测试数据通常较难对结构的整体预应力和刚度变化情况进行有效评价。本论文提出采用传统动力测试法来直接监测索杆张力结构刚度的新思路。考虑到索杆张力结构存在频率分布密集、模态参数对预应力偏差敏感等特点,对采用动力测试法进行此类结构刚度监测时存在的重要理论问题开展研究,具体包括以下五方面内容:(1)对于特定的荷载,各种刚度成分是否能够对该荷载所产生的变形起到主要抵抗作用应该能够被定量分析。这不仅有助于分析结构的传力路径,而且是有效实施刚度监测的依据。本文提出了一种能量形式的索杆张力结构刚度成分贡献度评价指标。该指标可以在结构和构件层面描述弹性刚度和几何刚度对结构整体刚度的贡献大小。更重要的是,以刚度矩阵各特征方向为对象,利用该指标还可以评价各特征方向刚度对特定荷载所产生变形的抵抗能力,且最终可从结构整体刚度中分离出“关键刚度”并作为结构刚度监测的重点。(2)传统方法一般基于理想结构(无预应力偏差)的模态信息来开展动力测试,但这仅适用于实际结构与理想结构模态差异较小的情况。索杆张力结构的频率分布密集且模态参数对预应力偏差敏感,因此有必要考虑实际结构模态参数(频率和振型)受随机预应力偏差的影响。本文建立了构件长度误差与模态特征值、振型之间的灵敏度关系。考虑随机预应力偏差的影响,提出了索杆张力结构模态参数变化幅值的估计方法以及是否会发生模态跃迁的判断准则。借助算例,还分析了随机预应力偏差对索杆张力结构模态特性的一般性影响规律。(3)基于模态空间的贡献度分析,本文提出了一种对索杆张力结构关键刚度实施动力监测的目标模态确定方法。由于索杆张力结构的模态参数对预应力偏差敏感,因此仅基于理想结构模型确定的目标模态往往不足以充分反映实际结构的关键刚度。针对此问题,本文进一步针对随机预应力偏差造成结构振型变化和模态跃迁这两种情况,提出了适用于实际结构关键刚度监测的目标模态扩展策略。算例分析表明,根据该扩展策略确定的目标模态可完全包络随机预应力偏差对索杆张力结构模态特征的影响,能充分反映实际结构的关键刚度。(4)正是由于模态参数对预应力偏差的敏感性,仅考虑理想结构模型信息的传统非完备振型扩展方法在应用于索杆张力结构时容易失效。本文提出一种考虑随机预应力偏差影响的索杆张力结构振型扩展的新方法。该方法的基本思路是将实际结构的振型近似表示为少数提供主要贡献的理想结构振型的线性组合,只要能合理确定这些贡献模态并对它们的组合系数进行最佳估计,则可完成目标模态的振型扩展。文中考虑振型变化和模态跃迁的影响提出了贡献模态的确定方法。此外,还进一步提出了一种改进的有效独立法(CMEI法)来获得贡献模态组合系数的最佳估计,同时也解决了传感器的最优布置问题。算例分析表明,本文所提方法能够有效进行存在随机预应力偏差影响的索杆张力结构的振型扩展。即使对于变化剧烈的模态和存在高噪声污染的情况,该方法也能保持很高的计算精度。(5)不考虑激励优化的常规模态测试方法对密集模态的识别精度较低。为提高识别精度,配合模态识别的时域法本文提出一种基于优化阶跃激励的密集模态测试方法。该方法通过调节阶跃激励的布置和大小,来增强结构自由振动中待识别模态的贡献且同时抑制邻近模态的贡献,并采用遗传算法来寻求最优激励模式。算例分析表明,所提方法能将密集模态识别问题转化成为孤立模态识别问题,可有效提高索杆张力结构密集模态的识别精度。
[Abstract]:The earliest built large span cable strut tensile structure service time has been more than 30 years, but at present the research in structure health monitoring and performance evaluation is not enough, the depth of the work is more rare. Cable strut structure stiffness and the formation mechanism of the conventional structure there are important differences, need to rely on to provide the prestressed structure the stiffness and stability. Experience shows that the design of cable strut structure is generally controlled by stiffness checking, compared with structural strength performance is often easily satisfied. Due to concerns about the structural stiffness big deviation, therefore has become the focus of prestressed cable strut structure performance monitoring in service period. The cable strut structure in practical engineering the basic is through direct test, prestressed cable to monitor, but the limited test data of cable force is often difficult to whole prestressed structure and stiffness changes of effective evaluation Price. This thesis puts forward new ideas to direct monitoring of cable strut structure stiffness test method using the traditional power. Taking into account the existence of dense frequency distribution of cable strut structure, modal parameters are sensitive to the deviation characteristics of important theoretical problems of prestressed, existing stiffness monitoring of this kind of structure on the dynamic testing method to carry out specific research including the following five aspects: (1) for a given load, the deformation is various stiffness of the components can play a major role in the load resistance should be quantitatively analyzed. The force transmission path which not only helps to analyze the structure, but also is the effective implementation of stiffness monitoring basis. This paper proposes a a form of energy the stiffness of cable strut structure component contribution evaluation index. The index can describe the elastic stiffness and geometric stiffness small contribution to the overall stiffness of the structure in the structure and component level More importantly, the stiffness matrix of the characteristic direction as the object, using the index can also evaluate the characteristics of directional stiffness deformation resistance to specific load, and ultimately from the overall stiffness of the structure to separate "stiffness" and as a key structural stiffness monitoring. (2 the traditional method is based on the ideal structure) (unprestressed deviation) modal information to carry out dynamic test, but this only applies to the actual structure and mode of the ideal structure of smaller differences. The structure of cable frequency distribution is dense and modal parameters of prestressed partial differential sensitivity, so it is necessary to consider the actual modal parameters (frequency and vibration type) affected by random deviation. This paper established the prestressed member length error and modal eigenvalue sensitivity relationship between modes. Considering the influence of random deviation of prestressed cable strut tension, the structure of die Estimation method of amplitude of state parameters and the criterion of whether there will be a modal transition. With a numerical example, also analyzed the general impact of cable strut tensile structure modal characteristics of the prestressed random deviation. (3) analysis of modal space based on the contributions of this paper puts forward a method of cable strut structure key the implementation of stiffness and dynamic monitoring mode. The modal parameters of cable strut structure sensitive to prestress deviation, so only the target mode to determine the ideal structure model is often not sufficient to fully reflect the actual key based on the stiffness of the structure. Aiming at this problem, this paper further studies the random deviation caused by the two kinds of prestressed structure and vibration mode change mode transition, put forward the target mode applicable to the actual structure of key stiffness monitoring expansion strategy. The example analysis shows that according to the objective model determine the expansion strategy States can be completely random deviation envelope of cable strut tension prestressed modal structure, can fully reflect the actual key structural stiffness. (4) it is because of the sensitivity of modal parameters of the prestressed deviation, only considering the ideal structure model of incomplete information of the traditional modal expansion method applied in cable strut structure easily failure. This paper presents a new method considering influence of random deviation of prestressed cable strut structure modal expansion. The basic idea of this method is the mode of the actual structure of the approximate linear combination of the ideal structure mode provides the main contribution to the minority, as long as can reasonably determine the contribution of these modal and their combination coefficient the best estimate mode can complete the target modal expansion. The method to determine the contribution of the modal vibration mode change and influence of mode transitions are considered. In addition, further Step forward an improved effective independent method (CMEI method) to obtain the best estimate with modal combination coefficient, but also solve the problem of optimal placement of the sensor. The example analysis shows that the proposed method can effectively for cable strut structure modal expansion effect of prestressed random deviation. Even for high the problem of noise intensity and the mode, the method can maintain high accuracy. (5) do not consider the conventional modal test method of optimal incentive intensive mode of low identification accuracy. In order to improve the recognition accuracy, with the time domain method of modal identification is proposed based on optimization step excitation intensive modal testing method. The method by adjusting the step excitation arrangement and size, to enhance the contribution to the identification of modal structures of free vibration and suppression of adjacent modal contribution, and the use of. The algorithm is used to find the optimal excitation mode. The example analysis shows that the proposed method can transform the dense modal identification problem into the isolated modal identification problem, which can effectively improve the identification accuracy of the cable rod tension structure with dense mode.

【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TU399

【参考文献】