Engineering Evaluation of Fatigue Striations in the Aircraft Engine InsdustryNick Cherolis
Rolls-Royce Corp.
"Striation Motivation" is a catchier title for this talk. Striations are the visible lines found on a fracture surface that show the progress made by a crack during a fatigue failure. In most real situations direct counting of these is impractical because of the very high numbers involved; however, small sample sections can be accurately counted, and by integrating "striation density" (striations/inch) plotted against crack depth, one can determine a fairly accurate count of the fatigue cycles. Several pitfalls must be avoided when using this approach but fortunately when one attempts to plot false striation lines no reasonable function is evident. Some example of false lines would be bench marks, tractor marks and the like. It should also be noted that high cycle fatigue, (while of great interest for certain situations) are not of interest in the situation where life is to be determined in terms of number of flights. High cycle fatigue lines will seem to remain very close together at very high counts, and can thus be identified and discounted for our purposes.
When one investigates a low cycle fatigue failure in an aircraft, the assumption is made that one cycle = one flight. Certain materials (particularly some Aluminun and Titanium alloys) are very good at developing readable striations. Some problems with the technique are when there are changes of section size, or loading redistribution occurs (both of which cause discontinuities in the attempted striation density vs. depth plots. Barring such difficulties, it is the practice to photograph the entire crack as a large mosaic, and then precisely identify where scanning electon microscope assisted counts of striations are made. These locations provide the distance axis for the plot of counted striation densities.
As an illustration of just how this works, I worked on a rotor from the famous Sious City (7-19-89) jet crash, which was caused when a tail mounted engine blew up and severed all hydraulic control functions, causing a totally fatal crash. Plotting the striation densities of many sample points along the crack, a well defined curve (high densities near origin - tapering to low densities as depth increased) was produced. Integrating the computer determined function of this curve a cycle count of 15,000 total cycles was determined (without knowledge of flight history of this engine). When records were checked, it was found that this engine had flown 15,503 times with all proper maintenance and inspections. The closeness of these two numbers means that there were essentially Zero cycles of initiation (flights before the crack began to propogate) which means that the crack itself was a manufacturing defect in the rotor which somehow excaped detection by all of the various inspections. The odds of this happening are very slim, but nevertheless it did.
To summarize, striation spacing is the reciprical of propogation rate. High striation density means low propogation rate while low density means high propogation rate (usually near the final failure point). We plot density as a function of distance from the crack origin (as determined by failure analysis). We sort out various false striation lines that may occur (such as high cycle fatigue, bench marks, etc.). Finally we fit the data to a function and integrate to determine a total cycle count (which is interpreted to be "number of flights"). This number is very useful in determining whether if there were, and how many initiation cycles were achieved before the crack began to propogate and should be detectable by inspections.