Identification technique for nonlinear boundary conditions of a circular plate

https://doi.org/10.1016/j.jsv.2005.01.058Get rights and content

Abstract

As a basic study for developing an identification technique for boundary conditions of machines and structures, a new technique for a circular plate is proposed. This technique has features that do not require data measured on the boundary and is applicable to nonlinear boundary conditions. In the proposed technique, the boundary is modelled by springs and dampers as well as effective mass and moment of inertia. Then their characteristics are determined by using the analytical solution together with the experimental data. Since the technique is based on the analytical solution, it is applicable to any structure, provided that its analytical solution can be derived. Numerical simulation is conducted to show that the procedure determines the boundary conditions accurately.

Introduction

To analyse the dynamic behaviour of machines and structures, numerical methods such as finite element method are often used. Such numerical methods, however, do not always yield accurate results. One of the reasons for this is its difficulty to specify the actual boundary conditions accurately.

As a means to grasp the characteristics of boundaries, experimental identification techniques have attracted interests of engineers, and many techniques have been proposed. Among them are techniques by Zhao et al. [1], Ren and Beards [2], [3], Ahmadian et al. [4], Pabst and Hagedorn [5], Xiang et al. [6], Takahashi [7] and Sanayei et al. [8], and Zhu and Huang [9]. All these techniques, however, have a restriction that they can be applied only to the case in which boundary conditions are linear. In practical cases, the boundary conditions often become nonlinear due to clearance, friction, material properties and so on. Thus, techniques applicable to nonlinear boundary conditions are desired. Sato et al. [10] proposed an identification technique for nonlinear boundary conditions. This technique, however, requires data measured on the boundaries. There are many cases in which measurement of data on the boundaries is difficult.

In previous papers [11], [12], [13], [14], [15], with the aim of developing an identification technique which is applicable to nonlinear boundary conditions and does not require data measured on the boundaries, the authors proposed techniques for one-dimensional structures. In this paper, as a continuation to the previous study, a technique applicable to two-dimensional structure is developed. Here a circular plate is considered.

In the first part of this paper, as a preparation, analytical solution for a circular plate with nonlinear boundary conditions is derived. Then an identification technique is proposed. In the technique, the boundary is modelled by springs and dampers. Depending on the case, effective mass and moment of inertia of the boundary are included in the model. Then their characteristics are determined by using the analytical solution together with the experimental data. Since the technique is based on the analytical solution, it is applicable to any structure, provided that its analytical solution can be derived. Finally, numerical simulation is conducted to discuss the applicability of the technique. In the simulation, responses of a circular plate are obtained from the equations of motion numerically, and random numbers are added to them as noise. With these data being regarded as experimental, identification of the boundary conditions is performed. In this way, it is shown that the proposed technique yields accurate results.

Section snippets

Formulation of the problem

A problem of identification of nonlinear boundary conditions for a linear circular plate is considered. Dynamic properties of the plate, except the boundary conditions are assumed to be known. In practice, it is often difficult to apply excitation to, or to measure responses on the boundaries. So, in developing a technique, it is imposed that the excitation to and measurement on the boundary are not required.

To formulate the problem, the boundary is modelled, as shown in Fig. 1, by springs and

Numerical simulation

To show applicability of the proposed technique, numerical simulation is conducted. First, dynamic response is calculated by solving the equations of motion of a plate whose parameters are given appropriately. Then using the data thus obtained, identification is performed. Finally, the obtained results are compared to the original values of the parameters used for calculation of the response.

Conclusions

As a basic study for developing an experimental identification technique applicable to nonlinear boundary conditions of a two-dimensional structure, a circular plate has been considered. A technique for it has been proposed. In the proposed technique, the boundary is modelled by springs, dampers, effective masses and moments of inertia. Their characteristics are determined by use of the data of steady-state deflection. By numerical simulation, it has been confirmed that the proposed technique

References (19)

There are more references available in the full text version of this article.

Cited by (6)

  • Boundary condition identification of tapered beam with flexible supports using static flexibility measurements

    2016, Mechanical Systems and Signal Processing
    Citation Excerpt :

    Using the combination of traveling wave method and BP neural network, Deng et al. [18] proposed a boundary condition identification method for fluid-filled piping system using the measured lower natural frequencies, and the experimental identification results shown that the boundary mass and stiffness were identified correctly. By solving the linear equations established by boundary characteristics and steady-state deflection, Suzuki et al. [19] developed an identification technique applicable to nonlinear boundary conditions of a two-dimensional structure, and illuminated the method by numerical examples of a circular plate with general boundary conditions. By expanding the beam response using its nonlinear normal modes, Ahmadian et al. [20,21] proposed a nonlinear boundary condition identification method for a clamped-free beam structure with local nonlinearities due to micro-slip at the clamped end in [20] and a fixed-frictionally supported beam with a relatively high constant normal preload at the interface of the frictional contact support in [21], and the experimental results shown that the identified boundary parameters can be utilized to predict the experimental observations accurately.

  • Vibrations in an elastic beam with nonlinear supports at both ends

    2015, Journal of Applied Mechanics and Technical Physics
  • Model-based identification of anatomical boundary conditions in living tissues

    2014, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
  • Atlas-based transfer of boundary conditions for biomechanical simulation

    2014, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
View full text