Breaking down new features of upcoming AFGROW release 5.2

In the last year, errors were found in the solutions for the unsymmetric double corner cracks at a hole subject to tension, bending, and bearing. The errors are related to very wide, thin plates with the bending case being most severe. Several investigations were conducted to understand the source of the error. The error is due to the geometric nonlinear behavior of very wide, thin plates to the applied loading. Fortunately, for most geometries of practical technical importance to the structural integrity community, the error is negligible.

Retardation effects due to spectrum loading are modeled with the Generalized Willenborg equation for the A-10 and T-38 weapon systems. The retardation parameter, SOLR (Shut off Overload Ratio), is determined from coupon test data for use in crack growth predictions. This correlation is accomplished with crack growth analysis that matches the test crack length and aspect ratio progression. Crack growth predictions; however, use a constant a/c assumption. This results in crack growth predictions that are unconservative relative to test results. Crack growth predictions using different crack aspect ratio assumptions are examined and compared to test results.

USAF (T-38) aircraft structural components that are fatigue and safety critical are typically managed using damage tolerance, including management of repairs. However, durability analysis may also be used for fleet management. An example will be presented with two repair options: 1) remove and replace the damaged component, or 2) repair, inspect, and reuse the component. Durability and damage tolerance analysis results will be presented to show a comparison between the two methodologies. This example demonstrates the pros and cons to inspecting and reusing a component versus replacing it with a new component. Items to consider when selecting an analysis methodology will be presented.

Cold working of holes has been shown to significantly improve the damage tolerance of materials and joints at high stress locations by engineering potentially high compressive residual stresses into aircraft structure; this has the effect of dramatically extending structural life from 2X to 10X the life were no residual stresses present. Cold working with resulting life extension will potentially allow aircraft maintainers to significantly extend inspection cycles over the life of the aircraft system, improving system availability of the aircraft and lowering life cycle costs. The cold working process is more than 40 years old and has been improved substantially in the interim. However, the designers and analysts who evaluate the damage tolerance of structures by using fracture mechanics-based analytical methods (primarily linear elastic fracture mechanics or LEFM) have not been able to take full advantage of life improvements because these methods have not thus far been able to accurately predict the extension of structural life possible by cold working. I present recent progress in the development and use of two new LEFM based tools that model crack growth in the cold worked induced residual stress fields, BAMF and CPAT, and show verification and validation (V&V) of the tools.

Cold working is a commonly used approach to increase the fatigue life of aircraft structural components. By inducing a circumferential compressive layer of residual stress around a cold worked hole, crack growth is slowed. Experiments show that primary cracks tend to preferentially evolve at the mandrel entrance side of the hole as a result of the variation in the residual stress field, then grow more quickly away from the hole, radially, than up the hole bore. Since the mandrel entrance face tends to be more difficult to inspect than the exit face, cracks may grow to become critical before they are detectable.

To support the development of inspection intervals that are not overly conservative a mathematical model for simulating crack growth at a cold worked hole was developed and implemented based on input geometry, material properties, and residual stresses from cold working. The model assumes that the crack grows in a single plane radially away from the hole, but makes no other assumptions about the shape of the crack front to allow complex crack shapes to develop. Validation experiments were performed to assess the predictive capability of the model and simulation predictions compare favorably to experimental data. The procedures of verification, validation and uncertainty quantification (VVUQ) were followed during the development of the model and its usage in prediction. Additionally, the development of such a mathematical model allows for low-cost sensitivity simulations for assessing the influence of input parameters on the fatigue life prediction. It is well known that experimental fatigue lives can vary significantly between test specimens. The model could be used to simulate the effect of variation in each input parameter to quantify expected variation in fatigue life. For example, the model was used in this project to quantify how small variations in residual stress can cause substantial variations in predicted fatigue life, demonstrating how an accurate determination of residual stress and its uncertainty is critical for prediction.

This work was performed in support of the project Integrating Residual Stress Analysis of Critical Fastener Holes into USAF Depot Maintenance (Rapid Innovation Fund - RIF - program AFRL-PK-11-0001).

Before the tabular K-solution for the oblique through crack at a hole was available in AFGROW, the classic, out-of-plane bending solution was limited to using 2/3 of the classic axial solution applied to the maximum bending stress as a simple (conservative) approximation. This was done because of the need to include the bending effect in the total life of cracked fastener holes under single shear loading. This conservative assumption has recently become an issue for several users who would like to use 1/2 of the classic axial solution, which is claimed to more closely match test data. This presentation will summarize the result of comparisons between the tabular oblique crack bending solution and the use of a fraction of the axial solution for a straight through-the-thickness crack for several initial crack lengths and stress levels.

This presentation documents work done to curve fit crack growth rate data for several materials that have been identified as a high priority for AFGROW Users. The presentation will explain the process used to fit the data. Curve fits have been completed for nine materials.

The objective of this work is to develop a database where temperature, PH2O, and loading parameters vary for legacy (AA7075-T651) and modern (AA2199-T86) aerospace aluminum alloys, then gain mechanistic understanding of the environmental fatigue process, and inform and support the development of a fracture mechanics based life prediction algorithm for variable environments. This work will result in (1) a broad characterization of the fatigue properties of two aerospace aluminum alloys in environments relevant to airframe application, (2) guidance on the relevance of the effect of low-temperature crack growth rates on inspection intervals and residual strength, (3) mechanistic understanding of the governing mechanisms that will inform integration of such environmental effects into fracture and mechanics based life prediction model. Furthermore, understanding and properly integrating the effects of low-temperature fatigue into airframe management tools could result in a significant reduction in the airframe inspection burden.

This presentation gives an overview of the work carried out at NRC on hole cold expansion simulation, interference fit fastener modeling, and riveting simulation. Using 3D multi-step nonlinear finite element analysis that replicates as closely as possible the actual processes, residual stress fields at fastener holes resulting from complex manufacturing and repair processes can be estimated and used for crack growth modeling. The example of an analysis of a countersunk hole containing riveting residual stresses is presented. For this case, the fatigue life distribution to first link-up is calculated by the use of finite element analysis, AFGROW analysis, and CanGROW Monte Carlo simulation. Key results from previous simulation studies are presented and future work is suggested.

Since the first K-solution for bearing loading was included in AFGROW, continued improvements have been made to address a number of issues that were discovered as a result of comparisons to more detailed FEMs. It was initially assumed that the bearing solution for a wide plate could be compounded with the finite width correction for an open hole. This was very quickly shown to be incorrect based on a comparison to the open hole solution for a narrow plate with an equivalent remote gross section stress. Later, the offset correction for the bearing load case was also shown to be significantly different from the open hole solution. The latest issue that has been discovered for the bearing load case is the effect of an in-plane bending constraint on the solution for a single thru-crack at a hole in the center of a plate. This presentation will discuss the influence of the in-plane bending constraint on the K-solution for a cracked hole under bearing loading and the implications for the current AFGROW bearing solutions.

The presentation describes recent measurements of residual stress in a large population of aluminum coupons with cold expanded holes, as well as follow-on analyses that quantify the stochastic character of the residual stress data and their influence on crack growth behavior.

The T-38 and A-10 analysis groups have developed an AFGROW plug-in that allows the user to ascertain a more physics based life prediction by coupling Finite Element Modeling (FEM) and AFGROW's crack growth analysis capability. .

This presentation will be a follow-on discussion from the 2012 and 2013 AFGROW Workshop presentations and will describe the recent developments within AFGROW and StressCheck to allow for residual stresses, the increased stability of the code and features for analyzing the results. Discussion will also include validation of code versus test results for models with and without residual stresses. The presentation will conclude with a demonstration on the macro that was developed to take a user defined un-cracked model and prepare it for BAMF simulation.

Currently the A-10 and T-38 ASIP Analysis Groups are validating the predictive capability of BAMF through deep residual stress fields. Multiple fatigue tests have been performed and many contour measurements have been provided and currently the group is determining the proper protocol for utilizing this new capability. The presentation will outline a series of critical parameters that are being investigated within AFGROW and StressCheck, through BAMF, to determine their influence on the crack growth prediction accuracy when compared to test data. These factors include material files, propagation limits and residual stress field measurements.

Work on new K solutions performed over the past year will be summarized. Included is work on solutions for single and multiple cracks at straight and countersunk holes. Error checking schemes will be described. The results of the evaluation of a new High-Performance Computing (HPC) asset (a CRAY XC30) to perform finite element analyses of large structural problems will be presented. Other CAStLE work of interest to the crack growth modeling community will also be discussed. Finally, inputs will be solicited from attendees regarding the direction of future K solution efforts.

Information on the latest research and development efforts and plans beyound AFGROW Release 5.2.