Daniel Gsell | PubMed Health

PubMed2004

Non-destructive evaluation of elastic material properties in anisotropic circular cylindrical structures.

In this paper, non-axisymmetric guided wave propagation in circular cylindrical, anisotropic structures is studied in a frequency range up to 1 MHz. The investigations are carried out with carbon fibre reinforced tubes. The aim is the experimental determination of their effective linear elastic material properties in a non-destructive way. Therefore, an analytical model of the dispersion equation is fitted to the experimentally detected dispersion curves by systematically adjusting the desired material properties. A total least square scheme accompanied by an outlier detection criterion is used for this optimization task. Since the raw data of the measured dispersion curves contain a lot of noise, these outliers have to be detected and excluded, to achieve accurate results. Good agreement is found between the measured curves and the analytically calculated curves based on the estimated parameters. This fact indicates a high accuracy of the determined material properties.

Journal Article↗

PubMed2004

Modeling three-dimensional elastic wave propagation in circular cylindrical structures using a finite-difference approach.

Wave propagation along circular cylindrical structures is important for nondestructive-testing applications and shocks in tubes. To simulate elastic wave propagation phenomena in such structures the governing equations in cylindrical coordinates are solved numerically. To reduce the required amount of computer memory and the computational time, the stress components are eliminated in the equilibrium equations. In the resulting coupled partial differential equations, in which only the three displacement components are involved, the derivatives with respect to spatial coordinates and time are approximated using second order central differences. This leads to the present new approach, which is both accurate and efficient. In order to obtain a stable scheme the displacements must be allocated on a staggered grid. The von Neumann stability analysis is performed and the result is compared with an existing empirical criterion. Mechanical energies are observed in order to validate the finite-difference code. Since no material damping or energy dissipation is taken into account in the equations of motion, the total energy must remain constant over time. Only negligible variations are observed during long-term simulations. Dispersion relations are used to check the physical behavior of the waves calculated with the proposed finite-difference method: Theoretically calculated curves are compared with values obtained by a spectrum estimation method, applied to the results of a simulation.

Journal Article↗