AFGROW User Workshop 2022

AFGROW, Release 5.4 includes several new features and capabilities. The most important new capabilities/features include: a new, advanced solution for combinations of a corner crack and through crack on either side of a hole; new MSD solutions; new weight function solutions for single corner crack and a single corner crack at hole with the stress distribution in C direction and double/single through crack(s) at hole.

Available finite width correction functions F_w that have been used in aircraft community for decades were developed based on a very limited number of FE-solutions, often of unknown accuracy and by mixing 2D/3D solutions with partly unverified assumptions used for inter-/extrapolation. The accuracy of the F_w-functions obtained in this way was in most cases estimated from comparisons with the FE-solutions that was used to derive the F_w-functions. Hence, the accuracy of the existing F_w-functions in the entire parameter space (R/t, W/R, a/t, a/c) has remained unknown until now.

In the present lecture we derive very accurate finite width correction functions F_w (R/t, W/R, a/t, a/c) for a single quarter-elliptical crack (dimensions (a, c)) at a straight shank hole (radius R) in a plate (thickness t, width 2·W and height 10·W) subject to tension, bending and pin loading.

In 1966, Schijve found that the transition from flat-to-slant (plane-strain to plane-stress) crack-growth behavior occurred at a "constant" crack-growth rate, independent of the stress ratio (R). Others had proposed that ΔK or Kmax controlled the transitional behavior. Newman (1966) had provided additional test data from the NASA Langley Research Center to support Schijve’s conclusion that crack-growth rate was the controlling parameter. In 1968, Elber discovered the crack-closure phenomenon, whereby the crack surfaces are partially-closed under tensile loading. Cracks only grow when the crack tip is fully open. Elber proposed to modify the Paris expression as

dc/dN = C (ΔKeff)ⁿ (1)

Since the crack-closure concept correlated fatigue-crack-growth-rate data on a unique ΔKeff-rate curve, independent of R, the flat-to-slant crack transition is controlled by ΔKeff. In 1992, Newman developed an expression for the flat-to-slant (constraint-loss) regime in terms of ΔKeff, as

(ΔKeff)T = 0.5 σo sqrt(B) (2)

where σo is the flow stress (average between yield stress and ultimate tensile strength) and B is sheet or plate thickness. Currently, the range of the constraint-loss regime (rate at the start of transition and rate at the ending of transition) and the associated constraint factors have to be determined by a trial-and-error method from test data under variable-amplitude loading. The objective of the current study is to “experimentally” determine the constraint-loss regime from constant-amplitude tests by measuring crack-opening stresses under low stress ratio (R = 0 or 0.1) conditions. In addition, it has been determined that the crack-growth delays after a single-spike overload may be caused by constraint-loss behavior. Thus, experiments on both 2024-T3 and 7075-T6 thin-sheet aluminum alloys will be conducted to measure the constraint-loss regime and to conduct repeated single-spike overload tests on M(T) specimens. A method will be developed to experimentally determine the constraint-loss regime either by measurements and/or by single-spike overload tests.

A primary goal of this project was to generate continuing damage test data for comparison to different continuing damage modeling approaches. Two different coupon forms were fatigue tested to specimen failure with crack length measurements taken throughout the testing process. The first sample form was manufactured with an offset hole and had two cracks embedded on each side of the hole. The flaw size closer to the free edge measured 0.050-in., while the crack opposite the short ligament side measured 0.005-in. These coupons more closely replicate the continuing damage model process. The second coupon form had a severed ligament from a previous test project. No artificial flaw was embedded on the side of the hole opposite the short ligament. Several models, including the currently used USAF approach, were compared with the test data.

This presentation will give an overview of the upcoming Spectrum Manager v 1.2. Some of the features include Spectrum Generation from Exceedance data, a new preview selector and spectrum chart, improved application speed, and COM support.