Welcome and Introductions (Continental Breakfast)

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

These new capabilities will be discussed in detail, and the timeline for the release will be provided.

Changes in constrain, particularly at/near surfaces, are known to change the stress intensity and crack shape. This behavior is often referred to as surface “pinning”. Observed differences between predicted and experimental behavior have resulted in various factors applied to the calculated stress intensity factor. Most notable is the crack closure factor developed by Newman that is included many of the NASGRO stress intensity solutions and was recently added as an option in AFGROW. With the advent of multi-point fracture mechanics, which allows crack shapes to evolve based on many discrete points along the crack front, mismatches between predicted and experimental crack shapes still persist. This presentation will discuss an investigation that was undertaken to understand the primary factors driving the observed differences, define an initial approach to incorporate these differences, and compare to available experimental data.

Break

Typical AFGROW predictions have been found to significantly overestimate the lives of Al7050 locations subject to highly compressive spectrum loading. This study aims at identifying potential analysis options for improving the life predicted for selected problematic locations. In particular, it looks at simplified representations of tensile residual stresses after compressive overloads, and at the effect of a neat-fit steel fastener. Material parameters related to the crack growth threshold and negative stress ratios are also studied with the goal to provide simple tuning options. Modelling results are compared with experimental data obtained from historical and ongoing experimental programs. Finally, recommendations for future work are made.

This presentation will review the history of the single corner cracked hole solution. This will include the 1986 Newman-Raju curve fit solution, the 2000 Fawaz-Andersson tabular database, and the 2018 updated Fawaz-Andersson database. Comparisons will be shown for several combinations of geometric parameters. A discussion of the two-point and multipoint life prediction options will also be included.

Lunch Break

This presentation will give an overview of Spectrum Manager v 1.2. Some of the features include Spectrum Generation from Exceedance data, ability to save individual sub spectra, a new preview selector, improved application speed, and COM support. There will also be a discussion of upcoming features that will be added to Spectrum Manager in future versions.

Break

There are subtle differences between a loading sequence and an analysis spectrum that is derived from the loading sequence. In some situations an incomplete understanding of these differences can lead to inaccurate crack growth predictions. Background information will be provided to differentiate between loading sequences and analysis spectra and the appropriate action that an analyst needs to take not just before conducting an analysis but also before executing fatigue tests.

This presentation will demonstrate the ability to use python to access AFGROW and Spectrum Manager's COM interface, run AFGROW predictions, and create spectra. The presentation will include examples of how to setup this connection, and show what functionality can be achieved through COM for both AFGROW and Spectrum Manager.

Design best practices avoid “knife edge” conditions in countersunk holes, typically defined as countersunk depth being greater than 2/3rds of the thickness. While well understood from a stress perspective, literature is lacking surrounding the effect of knife-edge on fatigue life. This analysis discretely modeled varying crack sizes in varying diameter and thickness geometry to characterize the Stress Intensity Factors (SIFs) life detriment caused by full countersink holes.

Situations may arise where a full countersink is needed from an unusual requirement, or from a poor repair choice implemented in the field. In these circumstances, it is important to understand how the countersink affects life. Depending on the expected life or need, the life detriment may be negligible. However, if the asset is to be retained for long periods of time, the countersink may be unacceptable.

This presentation will provide an overview to the online tools and resources available to AFGROW users, such as our website, YouTube channel, and other social media.

Adjourn

Continental Breakfast

An internationally coordinated analysis round robin was conducted where analysts provided blind predictions for three conditions: (1) open hole, (2) filled hole with 0.4% interference, (3) filled hole with 0.6% interference. All predictions were compared to test results from just two conditions: (1) open hole, (2) filled hole. In test the observed hole fill varied from 0.4% to 0.6% interference as a result of subtle fastener diameter variability up and down the fastener shank. Predictions are compared to test results to assess the accuracy of current analytical approaches to account for the life improvement from an interference fit fastener.

Accurate and reliable stress intensity factor K‐calculation for cracks in spectrum loaded joints is practically impossible due to the very complex contact, friction and stick‐slip conditions that take place between plates and at plate/fastener contact surfaces. An easy‐to‐use analysis tool is needed for safe design and maintenance of new and aging aircraft. In the lecture we describe how we determined the maximum K‐value that can exist at each point along a crack front. We also show, importantly, that the pin‐load model USAFA is using to derive hundreds of millions of K‐functions to be used in crack growth tools (e.g., AFGROW and NASGRO) provide K‐values that are within a few % of the highest K‐value possible, i.e., a value that will never be reached under practical circumstances. We consider the pin load model shown in the Figure 1 below. The loads 𝒒_𝟏 - 𝒒^𝟗 can model the contact/ friction conditions exactly at an arbitrary time during spectrum loading of the joint. We assume in the numerical analysis that 𝒒^𝟏 - 𝒒^𝟗 are polynomial functions, in r or z, and trigonometric functions in the circumferential direction. We make the following assumptions about friction: 𝒒_𝟕 = 𝜇₇𝒒_𝟖 and 𝒒_𝟗 = 𝜇₉𝒒_𝟖 where |𝜇₇| = |𝜇₉| (and similar assumptions for the conical and bottom surfaces). For 11 values of the friction coefficients in the range - 0.5 <= 𝜇 <= 0.5 (the sign determines the sliding direction) we solve the following optimization problem: Find functions 𝒒₁ - 𝒒_𝟗 that give the maximum value of 𝐾₁(φ*) at a point φ* at the crack front for the given set of friction coefficients {𝜇₁, 𝜇₃, 𝜇₄, 𝜇₆, 𝜇₇, 𝜇₉} . The highest value 𝐾₁(φ) found in these (11 ∗ 2)³ optimization studies we denote 𝐾_max(φ*). We repeat the analysis for 176 values of φ*. The Figure 2 below exemplifies 𝐾_max(φ*) near a crack front end (vertex) obtained from the 1.9 million optimumization runs. The figure also shows that the 2019 years pin‐load used by USAFA provide 𝐾₁‐values in almost perfect agreement with 𝐾_max(φ*) if the friction coefficient is 0.20 or lower. The 2019 years model underestimate the maximum value of K along the entire crack front with only 4% if the friction coefficient is 0.50 or lower. USAFA continues to use the 2019 year’s pin‐load model for the 2020 year’s work with K‐database generation.

DISTRIBUTION A. Approved for public release; unlimited distribution. USAFA‐DF‐2020‐298

In a recent study performed by the USAF, it appeared that if a neat fit fastener was installed in a hole that had a primary crack and a secondary crack, that the secondary crack was not influenced by the primary crack. To further investigate this finding a series of finite element models were created that varied crack size for the primary crack and secondary crack. Comparisons between models with and without secondary cracks are made to provide insight on how cracks, at a hole with a neat fit pin, may or may not influence each other.

Break

A new set of curve fit solutions for Multi-Site Damage modeling of through-the-thickness cracks in an infinite plate have been developed for implementation in AFGROW. These solutions include:

• Multiple through cracks
• Multiple holes with primary and secondary through cracks

These solutions will be compared to finite element results and solutions found in the open literature. There will also be a discussion on how we plan to implement the new solutions in AFGROW.

This presentation will go into a further discussion of the T-38 retardation parameter methodology. T-38 uses the Generalized Willenborg Retardation model, tuned by the parameter referred to as Shutoff Overload Ratio (SOLR). Last year Michael Wilcox presented updates to the T-38 SOLR Correlation for each FCL as part of the “Durability and Damage Tolerance Analysis Updates” (M. Wilcox 2019). Prior to that “The Generalized Willenborg shutoff overload ratio and its sensitivity to analytical parameters and techniques” (L. Smith 2012) evaluated the different stress intensity input techniques, and techniques for fitting crack growth curves during SOLR correlation. This presentation builds further from these studies. Using the automated SOLR correlation tool along with evaluation of various AFGROW stress intensity models, the goal is to determine optimum fits for the T-38 correlations.

Lunch Break

The current A-10 ground rules for determining the appropriate effective width for calculating the tension stress fraction results in total beta values vs crack lengths that lead to counter intuitive results, where an increase in bearing stress lead to an increase of life. This was presented and explained in AFGROW in 2019, "Relationship of tension (axial) and bearing beta solutions for pin loaded fastener holes.” If the effective width used is too small then what should the effective width be and does a larger effective width make sense? Additional comparisons were made between AFGROW and StressCheck, where the edge distance and number of holes were varied in order to match ground rule geometry configurations. The effective width was then adjusted until the two beta solutions matched. The intent of the comparison was to see if there was general pattern for determining what the effective width should be and how sensitive change in the effective width is to the resulting life.

Pipeline operators spend tens of millions of dollars annually on managing the crack threat on pipelines. A pipeline shutdown can result in up to $1M per day in lost revenue. Yet the state of crack growth analysis remains absurdly archaic. Introducing modern fracture mechanics to the industry has been met with roadblocks and resistance, but progress is nevertheless being made. The current uses of AFGROW in the industry, while limited, are already making a difference. The future is appearing brighter as time passes, and development initiatives which pave the way for integration of AFGROW crack life cycle analysis into assessment will be described. Maximizing incident learning, making pipelines more reliable, and reducing maintenance costs remain the main end game elements driving introduction of AFGROW and modern fracture mechanics to the pipeline industry.

Break

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

This session will discuss the current and upcoming releases of BAMF. The discussion will cover topics including:

• Features (users’ manual, installation packages)
• Major events
• Updates and bug fixes
• Utilization cases
• Future developments

Adjourn