Model validation for structural dynamic analysis: an approach to the Sandia structural dynamics challenge (Q929060)

Nowadays the model validation for structural dynamics is actively studied in industrial applications. Most validation methods are based on deterministic concepts. A deterministic structural model is usually validated by comparison of model predictions with reference data, and by examining the predicted dynamic properties such as stiffness, mass and damping. The stochastic methods are more complex and require more computing resources than deterministic methods. Here the authors extend some deterministic methods taking into account uncertainty and variability in the model validation. A mathematical model of a 3 DOF subsystem can be expressed by the set of basic equations of motion \[ \begin{multlined}\left[\begin{matrix} m_1&&\\&m_2&\\&&m_2\end{matrix}\right]\left\{\begin{matrix} \ddot x_1\\ \ddot x_2\\ \ddot x_3\end{matrix}\right\}+\left[\begin{matrix} c_1+c_2&-c_2&0\\ -c_2&c_1+c_3&-c_3\\ 0&-c_3&c_3\end{matrix}\right]\left\{\begin{matrix} \dot x_1\\ \dot x_2\\ \dot x_3\end{matrix}\right\}\\ +\left[\begin{matrix} k_1+k_2&-k_2&0\\ -k_2&k_1+k_3&-k_3\\ 0&-k_3&k_3\end{matrix}\right]\left\{\begin{matrix} x_1\\ x_2\\ x_3\end{matrix}\right\}+\left\{\begin{matrix} g(x_1,\dot x_1)\\ 0\\ 0\end{matrix}\right\}=\left\{\begin{matrix} f_1(x_1)\\ 0\\ 0\end{matrix}\right\}, \end{multlined} \] where \(g(x_1,\dot x_1)\) represents the weak nonlinearity in the model. By applying such approach to noise-free complete data and to a linear system, it is possible to derive all spatial properties exactly. For a nonlinear structure, the nonlinear structural elements can be considered as elements with equivalent linear stiffness, damping and mass coefficients. P.S. As a rule, in complex mechanical systems are essentially important only one, two, or three eigenvalues. This explains why the considered model systems possess only three DOFs.