Influence of fatigue crack closure on the growth rate of surface flaws

William Alan Troha, Purdue University


Inconsistent experimental results of fatigue crack closure studies over the past 15 years have caused recent investigators to conclude that the fundamental phenomena controlling the crack driving mechanisms are not understood. A major contributor to this misunderstanding is an inadequate understanding of the complex three-dimensional nature of closure. This is true for two-dimensional thru-the-thickness cracks as well as for three-dimensional part-thru surface cracks. To assess the three-dimensional influence of fatigue crack closure on the growth rate of surface flaws, an experimental investigation was conducted using a rectangular plate of polymethylmethacrylate (PMMA) material loaded in four-point bending. The use of transparent PMMA material and a Newton optical interferometer system with a laser light source provided the capability to simultaneously measure three-dimensional crack opening displacement (COD) profiles and closure loads (P$\sb{\rm cl}$). Experimental results are presented for a constant load test, a variable block loading test, and two constant stress intensity factor tests controlled along the crack free-surface and into the specimen thickness. The formation of a "void" area internal to the crack surfaces was found to be a major factor in defining three crack types which directly influenced COD, P$\sb{\rm cl}$ and fatigue crack growth. The three experimentally determined crack types and three distinct closure loads are used to establish corresponding definitions of effective stress intensity factor ($\Delta$K$\sb{\rm eff}$). In the closure load region, current prediction models are hypothesized to underestimate K along most of the crack tip boundary due to the different crack types. The crack P$\sb{\rm cl}$, P$\sb{\rm cl}$/P$\sb{\rm max}$ and fatigue crack growth rate (FCGR) data are presented as a function of crack length. FCGR data are compared using applied $\Delta$K and several definitions of $\Delta$K$\sb{\rm eff}$. Selection from several effective load ranges definitions for calculating $\Delta$K$\sb{\rm eff}$ is based on the best correlation of FCGR data. The use of $\Delta$K$\sb{\rm eff}$ for predicting fatigue cycles was found to be of superior accuracy over applied $\Delta$K. COD predictions for surface cracks using the weight function approach were found to underestimate displacement amplitude and to not adequately model profile distributions away from the crack tip.




Grandt, Purdue University.

Subject Area

Aerospace materials

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server