Author
Abstract
When a structure is subjected to a random loading, its dynamic response changes, characterized by shifts in the eigenvalues and modification of eigenvectors. The peaks in the dynamic response spectrum will give indications of the natural frequencies of the structure. Any damage in the structure will be reflected in the spectrum as a shift in natural frequencies.
Crack initiation and early growth of fatigue cracks in 0.4% carbon steel and 2024-T4 aluminum are investigated using spectral density approach for applied random loading. Energetic consideration and spectral density approach lead to the formulation of a model used to predict the behavior of short and long cracks initiation, propagation, and paths taking into account the microstructural variables relevant to fatigue crack initiation and early crack growth.
The model indicates that crack arrest occurs when crack-tip is blocked by a grain boundary. In the short crack region, propagation and non-propagation occurs depending on random loading stress level and slip band energy released. The application of the present model to cumulative damage is compared with the experimental data and a reasonable agreement is found. The introduction of material specification shows a quantitative description of the parameters which affect the reliability of a structural component subjected to random loading. The power of the analytic model and the simulation analysis in the present work give some insight to the behavior of the structure under random loading showing the different mode shapes, eigenvalues/vectors, deformation, propagation and non-propagation of cracks, and the stresses caused by such random loading that can lead to fatigue failure. A life predication model is presented for long-life fatigue , to control the hardware design and to find a proper combination of random load and life .
)
