Font Size: 

The presence of a crack or any other discontinuities can significantly reduce the fatiguelife of a component or structure. The crack may be initiated by fatigue, be amanufacturing defect, caused by an impact, or similar event (e.g., a thermal shock.).However, in most engineering cases, the initial size of a crack or discontinuity is notlarge enough to cause catastrophic failure. Once a crack is present in a structure, it tendsto grow under the influence of cyclic loading until its critical size is reached. At thispoint, fracture occurs.Many fatigue crack growth studies available in literature have been carried out underconstant amplitude loading, otherwise referred to as steady state conditions. As a result,constant amplitude fatigue crack growth data is in general repeatable and wellunderstood. The problem of predicting fatigue crack growth life becomes increasinglymore complex when the loading spectrum is not constant amplitude. This is commonlyreferred to as variable-amplitude or spectrum loading and produces what is known asmemory effects or load-history effects.The UniGrow fatigue crack growth model initially proposed by Glinka and Noroozi [1] isbased on the elastic-plastic stress-strain material response at the crack tip region. Thismodel can be also related to the group of so-called ‘residual stress based models’,according to the Skorupa’s classification [2]. Residual or compressive stresses ahead ofthe crack tip can either delay or accelerate subsequent fatigue crack growth dependingupon the current crack length and load history.The present work is a modified version of the UniGrow fatigue crack growth model, andmakes it applicable to all kinds of variable amplitude loading spectra. The modificationsinclude a very important feature called the ‘memory effect’.