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; updated solution for advanced double/single corner crack(s) at a hole; updated solution for classic double/single oblique crack(s) at a hole; updated width correction for crack(s) at hole; new weight function solutions for double/single through crack(s) at hole These new capabilities will be discussed in detail, and the timeline for the release will be provided.

This presentation will review the current ASTM load-reduction test method and discuss issues of possible concern. An explanation of a compression pre-cracking threshold test methodology will also be provided. A comparison of near threshold crack growth rate data will be made for several materials. Finally, a proposed annex to ASTM E-647 will be presented.

An initial FCG Analysis Methods round robin was completed to quantify the epistemic uncertainties in the prediction of crack growth life, given a fixed set of input data, for baseline and cold expanded (Cx) fastener holes. During this initial round robin, the prediction sensitivity to the analysis inputs was highlighted with one specific case identifying the influence of error in the Mode I Stress Intensity Factor (KI) for applied remote loading. As a result of these findings and subsequent discussions amongst the fatigue crack growth community, a follow-on collaborative round robin was established to investigate differences in stress intensity factors readily available in commercially available software like AFGROW and NASGRO. The primary objective of the Stress Intensity Factor (SIF) round robin was to evaluate differences between available SIF solutions for a single corner crack at a fastener hole with remote uniform tension loading. Evaluations included the root SIF solution and any corrections used to account for any additional corrections applied to the solution (e.g. single vs. multiple cracks, finite width, hole offset). Solutions compared to explicit Finite Element Analysis (FEA) results of each case. Findings intended to drive improvements to solutions available to the fracture mechanics community.

A building block test effort is a vital and necessary component to certifying aircraft structure. In this presentation, the application of coupon and element level building block test results to identifying analysis parameters, used as part of an air vehicles crack growth methodology, are examined. The applicability of key analysis parameters, such as available stress intensity solutions, angles at which stress intensities are evaluated (f-angles), closure correction factor (ßr), and load interaction models, to the analysis method in general and/or to specific usages/control points are discussed. Both quantitative and qualitative metrics used in assessing the validity of the parameter settings are discussed. Finally, the overall correlation of the final set of analysis parameters are discussed along with the validity of applying applicable parameters settings to analyses using BAMpF V7 is presented.

It was recently discovered that there are significant errors in the closed-form finite width correction commonly used for corner and surface cracks at a hole. This presentation will review the results of a study to assess the accuracy of K-solutions for corner cracked holes. In addition, the work being done to develop a new closed-form finite width correction will be discussed.

The A-10 ASIP team has performed a number of analytical studies recently to help in refinement of the DTA ground rules. Typically, these studies have involved reviewing the function in question, determining the impact of incorporating the function on A-10 analyses, and investigating the relative accuracy of the different options. This presentation summarizes the studies performed that investigated the use of varying vs constant corner crack aspect ratios, standard vs advanced models, net section yield as a failure criterion, and the Generalized Willenborg compressive cycle yield zone correction.

The A-10 program has recently discovered a trend in the fleet showing damage in the upper auxiliary longeron. Four aircraft have been discovered with severed longerons, and many with crack indications. The load path in the area is unique and poorly designed. The intended load path is shear transfer through a skin panel due to a step up from the forward fuselage longeron section to the center fuselage section. Due to the presence of a removable panel, this skin panel is attached with clearance fit holes and nut plates. The consequence of this is load gets transferred through the frame section common to the two longeron segments instead of the skin panel. The interest related to AFGROW is two fold. First, there is a large off-axis load component since the load has to transition vertically through the frame and skin panel. Secondly, there is significant variation in the TSF and BSF for the critical hole for different load levels within the spectrum. At about 25% max spectrum load, the TSF is approximately 0.78 and BSF is 13.5, and at max spectrum load the TSF is about 0.89 and BSF is 5.51. This presentation intends to drive discussion within the group on how best to handle these scenarios within AFGROW.

Many FEA programs offer Stress Intensity Factor (SIF or K) calculations for certain crack shapes. A notable exception is Siemens NASTRAN. MSC PATRAN requires additional solver add-ons to determine SIFs. ANSYS and StressCheck have the most robust SIF calculation tools but have limited capability with very large, multi-material assemblies. A general solution to determine SIFs based on nodal values of displacement and stress around the crack tip was desired. Using standard 2nd order elements (hexahedral, pentahedral, pyramid, and tetrahedral), models were built in Siemens Simcenter/NASTRAN. Both the Displacement Interpretation Method (DIM) and Contour Integral Method (CIM) were explored. A crack construction methodology, data extraction methodology, and exploration into the underlying formulas were also studied and will be presented.

Work has recently been completed to establish an integration of AFGROW with the BEASY Fatigue & Crack Growth software. Utilizing the AFGROW COM Interface, developments have been undertaken to support efficient communication between the two software platforms. This integration has significant value in that the BEASY software provides, highly accurate and structurally representative stress intensity factors, that can be used in conjunction with crack growth equations and fatigue crack growth material data available in AFGROW. By leveraging the unique capabilities of each software platform we can offer the engineering community, higher fidelity fatigue crack growth simulations based on more realistic structural models and loading conditions. Ultimately, the more accurate capture of crack shape evolution, coupled with vetted material data and crack growth relationships, will provide improved structural lifing prediction capability.

Fatigue crack growth round robin challenges can provide very useful verification and validation benchmarks for the damage tolerance design community. A corner crack at a fastener hole is a common round robin example due to its relevance in aircraft structural integrity assessment. Various tools are available to the analyst to perform fatigue crack growth life assessments that resemble specimen level testing conditions, compare against measurement data for reaching validation requirements and apply the same modeling technique with confidence at component or structure level. 3D finite element modeling is one of the techniques that can be employed in simulating incremental crack growth. Using SimModeler Crack, nominal experimental conditions from two fatigue crack growth test procedures (made available to the ERSI working group) were simulated using 3D FEA and, fatigue crack growth predictions were compared to the corresponding physical measurements. One advantage of employing readily available 3D modeling techniques over reduced order models is the ability to deterministically evaluate sources of variability related to modeling assumptions (one example is crack front shape), experimental procedure (specimen misalignment) or geometric dimensions (interference fit allowance). Off-nominal conditions were simulated to quantify the sensitivity of crack growth prediction (remaining useful life) to different sources of variability and therefore, provide a broader modeling validation assessment.

A key challenge generally for predicting fatigue crack growth is the use of baseline constant amplitude (CA) data to predict the behaviour under spectrum loading. Predicting CA behaviour for a given geometry and loading should be relatively simple and accurate. In the case discussed here, a compact tension specimen manufactured from 7075-T6 aluminium alloy was subjected to tension-tension R=0.1 loading. After a period of sustained and stable growth, a single factor 2.0 spike overload was applied and then the CA loading was resumed. The crack did continue to grow, and after another period of sustained stable growth, another single factor 2.0 overload was applied. The CA loading was again resumed and growth continued. This scenario was recently offered up to the fatigue and fracture community as a blind prediction challenge. Several submissions were received. The results from this challenge case are presented and discussed here.

Obtaining and assessing the utility of crack propagation data, specifically with limited documentation for how the data were generated, is a challenge. One option is to disregard the data altogether and develop your own based on ASTM E647 in combination with test expertise. This can prove prohibitive for many analysts due to lack of funding, lack of experience, or both. Another is to analyze data available and see how various data sets compare to each other and as a group. In this example, a couple options were evaluated. First, da/dN distributions were determined incrementally at fixed ranges of delta K. The second approach was to develop delta K distributions based on fixed ranges of da/dN. It was determined that the latter approach was impractical because a fixed range of da/dN, i.e. a scatter factor, is not consistent across the range of delta K. In contrast, choosing a fixed range of delta K, for conversation sake, could be justified by an acceptable load uncertainty. Although a 5% variation, positive and negative, was chosen, the concept could be adapted based on unique analysis criteria. A two parameter Weibull analysis process was applied to each of the various data groups for a collection of 7050 crack propagation data. In addition to highlighting how the data varied through a range of delta K, other observations will be highlighted for further discussion.

A comparison study is performed to substantiate a material change for a fracture critical part from a forging (7175-T736) to a machined plate (7050-T7451). Minimal spectrum development for the component in question exists. Similarly, there is minimal information on appropriate retardation model parameters. To justify the material change and avoid extensive testing, thorough detailed analytical comparisons are made to demonstrate comparable or superior performance of the redesigned machining across the potential range of retardation parameters. The study with associated fleet history / past experience provided enough confidence to justify the material substitution.

Durability analysis is in work on A-10 control points with crack indications from fatigue tests. SwRI began with both a strain-life approach and a crack growth life approach. For the crack growth calculation, each CP was assigned an initial flaw size of 0.01” and the COM was used following the A-10 DTA ground rules. For the strain-life analyses, a new COM was developed. The strain-life approach saw significant increase in durability life compared to the damage tolerance analysis. In order to understand and verify the results, SwRI performed a sensitivity study on the parameters required for the strain-life method.

A fatigue test program is being conducted for the continuing damage scenario (two cracks growing on opposite sides of a hole). The test matrix includes some coupons that are pre-cracked and some that are not. Various analytical models will be used to predict the crack growth and comparison will be made to the measured cracking in the coupons. It is desired to gain insight into several questions related to the analytical models (such as: what is an appropriate secondary flaw size?) and to compare the multi-phase approach vs the simultaneous crack growth approach.

Steep residual stress gradients at a cold expanded hole can complicate the accuracy of fatigue predictions. The interaction of that residual stress with a steep stress concentration gradient from the hole exacerbates prediction accuracy and deciphering the source of inaccuracies. A test program was executed where test specimens were manufactured with the test piece machined off at the tangency of a cold expanded hole to a specimen similar to a single edge notch tension, SEN(T), specimen. The resulting specimen is one with residual stresses from cold expansion without the geometric complications from the presence of a hole. Crack growth measurements are compared with fatigue predictions for increased understanding of how to accurately model crack growth at cold expanded holes.

Fatigue predictions that include residual stresses have historically required iterative type analyses with FEA. A complete fatigue prediction would often take many hours. Alternatively, AFGROW has a built in capability to include residual stresses on a traditional analysis, completed in seconds. An overview of recent comparisons using the AFGROW built in capability to test data, and FEA based predictions is shown with lessons learned and recommendations for applicability of each approach.

This presentation summarizes a cross-committee effort within the ERSI Working Group to help identify the driving factors in the Cx process that contribute to extended damage tolerance life. FTI modeled 29 coupons with varying parameters using their ABAQUS process simulation model and the resulting residual stress fields were extracted. AFGROW-BAMpF models were then created for each individual geometry and stress field. A surrogate model was created by fitting the input variables to the output (life). Global sensitivity analysis revealed that sleeve thickness was dominant; however, results should be interpreted with caution since the input data was from observed measurements and not a DOE.

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.

Information on the latest research and development efforts and plans beyond AFGROW Release 5.4