|
Abstract
|
In this thesis, the estimation of thermo-mechanical moving load in functionally graded axisymmetric shells is investigated. The temperatures and displacement components at the outter surface, which are obtained experimentally by using efficient number of sensors or are simulated by using the direct problem solution, are utilized as input data for the inverse algorithm. Since the experimental data is not available, in this thesis the direct problem is solved to simulate the measuring data. The governing mechanical equations are derived based on the first-order shear deformation theory and the thermal equation is derived by Fourier’s heat condunction law. To benefit from the high accuracy and low computational efforts of the differential quadrature method (DQM) in conjunction with the effectiveness of the finite element method (FEM), a combination of these methods is employed to discretize the thermal equation in the spatial domain. The finite element method (FEM) is employed to discretize the equations of motion in axial direction and the Newmark’s time integration scheme is utilize to solve the equations in time domain. The material properties are assumed to be temperature-dependent and vary smoothly and continuously in the thickness direction. In order to check the validity of the results, the free vibrations results of shells are compared with the existing solutions in literature in the limit case. After validating the formulation and method of solution, the thermo-mechanical moving load in functionally graded axisymmetric shells under different boundary conditions is estimated. The inverse algorithm composed of the conjugated gradient method and adjoint equations which is one of the most stable methods. This method is not sensitive to initial guesses and measurement errors and the result of this thesis obviously shows this fact. The applicability of the proposed inverse algorithm is demonstrated through different examples.
|