The growth of a fatigue crack through a material is the result of a complex interaction between the applied loading, component geometry, three-dimensional constraint, load history, environment, material microstructure and several other factors. Previous studies have developed experimental and computational methods to relate the fatigue crack growth rate to many of the above conditions, with the intent of discovering some fundamental material response, i.e. crack growth rate as a function of something. Currently, the technical community uses the stress intensity factor solution as a simplistic means to relate fatigue crack growth rate to loading, geometry and all other variables. The stress intensity factor solution is a very simple linear-elastic representation of the continuum mechanics portion of crack growth. In this paper, the authors present fatigue crack growth rate data for two different high strength steel alloys generated using standard methods. The steels exhibit behaviour that appears unexplainable, compared to an aluminium alloy presented as a baseline for comparison, using the stress intensity factor solution.

Cracking in D6ac steel has played a key role in the development of thedamage tolerant design philosophy. Until recently it had been thought thatcracking in D6ac steel was well understood. However, Forth et. al. [1] haverecently reported that crack growth data obtained by NASA for D6ac steelrevealed that there was little R ratio dependence in the Paris Region and that therewas thus little, if any, crack closure. In this paper we reveal how the NASA dataconforms to the Generalised Frost-Dugdale crack growth law. We then show howthe average block variant of the Generalised Frost-Dugdale law can be used topredict the growth of small cracks in a USAF study into crack growth in D6acsteel specimens representative of the critical region of F-111 wing pivot fitting.

This paper examines the results of an extensive test program undertaken to study crack growth in D6ac steel and shows that in each case the increment in the crack length per cycle (da/dN) conforms to the Generalised Frost–Dugdale crack growth law. This is found to be true for both constant Kmax, constant R ratio load increasing, and compression–compression pre-cracking tests in the L-T, T-L and the S-T directions.

The ASTM standard for fatigue crack growth, E 647, was co-developed by industry and government in the late 1970s to define a standard procedure for generating material crack growth behavior under a variety of loading conditions. The standard specifies tolerances on laboratory procedures, specimen configurations to generate material characterization information of a high integrity. Recent research developing fatigue crack growth rate data has uncovered significant issues with the standard test method and specimen configurations. The data generated has been largely dependent on specimen configuration, i.e., M(T) and C(T) specimens produce very different crack growth rate curves; C(T) specimens of different sizes do not generate the same material response; and finally, the development of steady-state, as defined using constant ΔK testing, does not agree with the data generated with the standard ASTM test method. Therefore, in this paper the authors present data that does not lend confidence in the integrity of the standard C(T) specimen tested using the ASTM standard method and investigate causes for the observed anomalous crack growth rate behavior.

One of the responses of a material to extreme forces, such as stress, temperature, etc., is to crack. A crack appears when the material reaches a limit in its capability to absorb damage and fails. Sometimes, a crack will grow under a periodically applied condition, such as cyclic loading, that are well below the stresses required to fail the material, denoted fatigue crack growth. Over the past thirty years, laboratory experiments have been devised to develop fatigue crack growth rate data that is representative of the material response. The crack growth rate data generated in the laboratory is then used to predict the safe operating envelope of a structure. The ability to interrelate laboratory data and structural response is called similitude. In essence, a nondimensional term, called the stress intensity factor, was developed that includes the applied stresses, crack size and geometric configuration. The stress intensity factor is then directly related to the rate at which cracks propagate in a material, resulting in the material property of fatigue crack growth response.

Fatigue crack growth rate testing is performed using automated data collection systems that assume straight crack growth in the plane of symmetry and that use standard polynomial solutions to compute crack length and stress-intensity factors from compliance or potential drop measurements. Visual measurements used to correct the collected data typically include only the horizontal crack length, which underestimates the crack growth rates for cracks that propagate out-of-plane. The authors have devised an approach for correcting both the crack growth rates and stress-intensity factors based on two-dimensional mixed mode-I/II finite element analysis (FEA). The approach is used to correct out-of-plane data for 7050-T7451 and 2025-T6 aluminum alloys. Results indicate the correction process works well for high ΔK levels, but it fails to capture the mixed-mode effects at ΔK levels approaching threshold (da/dN ∼ 10−10 meter/cycle). Based on the results presented in this paper, the authors propose modifications to ASTM E 647: to be more restrictive on the limits for out-of-plane cracking (15°); to add a requirement for a minimum of two visual measurements (one at test start and one at test completion); and to include a note on crack twisting angles, with a limit of 10° being acceptable.

Compression precracking (CPC) has seen renewed interest as a possible alternative procedure for generating fatigue crack growth threshold data with minimal load history effects, but recent testing confirms results from the literature that compression precracking does induce load history effects through residual stresses that influence subsequent fatigue crack growth test data. Using the CPC method, specimens are precracked with both maximum and minimum compressive loads. Compressive yielding occurs at the crack-starter notch, resulting in a local tensile residual stress field through which the fatigue crack must propagate. Although the tensile residual stress field contributes to the driving force for precracking, it also introduces the possibility of history effects that may affect subsequent fatigue crack growth. The tensile residual stress field elevates the local driving force at the crack tip, promoting higher crack growth rates than would be expected from the applied loading. This paper presents three-dimensional finite element results and experimental data for compact tension specimens that characterize the load history effects induced by compression precracking. The analysis results indicate that for low tensile loading levels near the threshold region, the residual stresses cause the calculated crack tip driving force to increase from the applied driving force by 25% or more. In addition, significant crack growth of about two times the estimated plastic zone size is needed to grow away from the residual stress field and reduce the calculated crack tip driving force to within 5% of the applied driving force. Experimental results show that growth of about two to three times the estimated plastic zone size is necessary to establish steady growth rates under constant ΔK loading for the materials and loading levels evaluated. Constant ΔK testing following compression precracking will demonstrate when residual stress effects are no longer significant and will ensure consistent growth rates.

The surface integral method, an indirect boundary element method that represents a crack as a distribution of force dipoles, has been developed to model three-dimensional non-planar crack growth in complex structures. The finite body was effectively modeled by superposition of stress influence functions for a half-space. As a result of this strategy, only the fracture has to be discretized. Crack propagation was modeled using the maximum circumferential stress theory to predict crack direction and the Forman fatigue equation, modified with an equivalent stress intensity solution for mixed-mode, to predict extension. Comparisons with benchmark solutions and field data verified the computational methodology and defined the limits of its applicability.

A new hybrid surface-integral-finite-element numerical scheme has been developed to model a three-dimensional crack propagating through a thin, multi-layered coating. The finite element method was used to model the physical state of the coating, and the surface integral method was used to model the fatigue crack growth. The two formulations are coupled through the need to satisfy boundary conditions on the crack and external surface. The coupling is sufficiently weak that the surface integral mesh of the crack surface and the finite element mesh of the uncracked volume can be set up independently. Thus, when modeling crack growth, the finite element mesh can remain fixed for the duration of the simulation as the crack mesh is advanced. This method was developed to evaluate the feasibility of fabricating a structural health monitoring system for real-time detection of surface cracks propagating in aircraft engine components. In this work, the authors formulate the hybrid surface-integral-finite-element method and discuss the mechanical issues of implementing a structural health monitoring system in an aircraft engine environment.

The accurate representation of fatigue crack threshold, the region defining crack growth as either very slow or nonexistent, is extremely important. If the experimentally measured threshold is unconservatively high, a structural component designed with this data may fail long before the fatigue analysis predicts. The fatigue crack growth threshold is experimentally defined using ASTM standard E 647, which has been shown to exhibit anomalies. Alternate test methods have been proposed, such as the constant K-max test procedure, to define the threshold regime without ambiguity. However, only the current test method defined by ASTM is designed to produce the range of fatigue crack thresholds (e.g., low and high R) needed to characterize an aerospace loading environment. It is the scope of this paper to determine the fatigue crack growth threshold of a well characterized aerospace alloy, 7075-T7351 aluminum, using different methods, to compare the results, and to draw conclusions.

The Federal Aviation Administration, the National Aeronautics and Space Administration and the aircraft industry have teamed together to develop methods and guidance for the safe life-cycle management of dynamic systems. Based on the success of the United States Air Force damage tolerance initiative for airframe structure, a crack growth based damage tolerance approach is being examined for implementation into the design and management of dynamic systems. However, dynamic systems accumulate millions of vibratory cycles per flight hour, more than 12,000 times faster than an airframe system. If a detectable crack develops in a dynamic system, the time to failure is extremely short, less than 100 flight hours in most cases, leaving little room for error in the material characterization, life cycle analysis, nondestructive inspection and maintenance processes. In this paper, the authors review the damage tolerant design process focusing on uncertainties that affect dynamic systems and evaluate the applicability of damage tolerance on dynamic systems.

Experimental test procedures for the development of fatigue crack growth rate data has been standardized by the American Society for Testing and Materials. Over the past 30 years several gradual changes have been made to the standard without rigorous assessment of the affect these changes have on the precision or variability of the data generated. Therefore, the ASTM committee on fatigue crack growth has initiated an international round robin test program to assess the precision and variability of test results generated using the standard E647-00. Crack growth rate data presented in this report, in support of the ASTM roundrobin, shows excellent precision and repeatability.

The feasibility of fabricating an in-situ crack sensor for real-time detection of surface cracks propagating in engine components was evaluated using a computational fracture mechanics model. The in-situ sensor system would be required to: (1) be capable of sustaining normal function in a severe environment; (2) transmit a signal if a detected crack in the component was above a predetermined length, but below a critical length that would lead to failure; (3) act neutrally upon the overall performance of the engine system and not interfere with the engine maintenance operations. In this work, fracture mechanics methodologies are used to identify the requirements for an in-situ sensor system that could withstand the engine operating environment, foreign object damage, and minimally degrade engine performance. A computational fracture mechanics model was developed to evaluate the feasibility of fabricating an in-situ crack sensor for real-time damage propagation detection in engine components.

Finite-element fracture simulation methodology predicts the residual strength of damaged aircraft structures. The methodology uses the critical crack-tip-opening-angle (CTOA) fracture criterion to characterize the fracture behavior of the material. The CTOA fracture criterion assumes that stable crack growth occurs when the crack-tip angle reaches a constant critical value. The use of the CTOA criterion requires an elastic- plastic, finite-element analysis. The critical CTOA value is determined by simulating fracture behavior in laboratory specimens, such as a compact specimen, to obtain the angle that best fits the observed test behavior. The critical CTOA value appears to be independent of loading, crack length, and in-plane dimensions. However, it is a function of material thickness and local crack-front constraint. Modeling the local constraint requires either a three-dimensional analysis or a two-dimensional analysis with an approximation to account for the constraint effects. In recent times as the aircraft industry is leaning towards monolithic structures with the intention of reducing part count and manufacturing cost, there has been a consistent effort at NASA Langley to extend critical CTOA based numerical methodology in the analysis of integrally-stiffened panels.In this regard, a series of fracture tests were conducted on both flat and curved aluminum alloy integrally-stiffened panels. These flat panels were subjected to uniaxial tension and during the test, applied load-crack extension, out-of-plane displacements and local deformations around the crack tip region were measured. Compact and middle-crack tension specimens were tested to determine the critical angle (wc) using three-dimensional code (ZIP3D) and the plane-strain core height (hJ using two-dimensional code (STAGS). These values were then used in the STAGS analysis to predict the fracture behavior of the integrally-stiffened panels. The analyses modeled stable tearing, buckling, and crack branching at the integral stiffener using different values of critical CTOA for different material thicknesses and orientation. Comparisons were made between measured and predicted load-crack extension, out-of-plane displacements and local deformations around the crack tip region. Simultaneously, three-dimensional capabilities to model crack branching and to monitor stable crack growth of multiple cracks in a large thick integrally-stiffened flat panels were implemented in three-dimensional finite element code (ZIP3D) and tested by analyzing the integrally-stiffened panels tested at Alcoa. The residual strength of the panels predicted from STAGS and ZP3D code compared very well with experimental data. In recent times, STAGS software has been updated with new features and now one can have combinations of solid and shell elements in the residual strength analysis of integrally-stiffened panels.

The fatigue crack growth threshold, defining crack growth as either very slow or nonexistent, has been traditionally determined with standardized load reduction methodologies. These experimental procedures can induce load history effects that result in crack closure. This history can affect the crack driving force, i.e. during the unloading process the crack will close first at some point along the wake or blunt at the crack tip, reducing the effective load at the crack tip. One way to reduce the effects of load history is to propagate a crack under constant amplitude loading. As a crack propagates under constant amplitude loading, the stress intensity factor range, ΔK, will increase, as will the crack growth rate, da/dN. A fatigue crack growth threshold test procedure is experimentally validated that does not produce load history effects and can be conducted at a specified stress ratio, R. The authors have chosen to study a ductile aluminum alloy where the plastic deformations generated during testing may be of the magnitude to impact the crack opening.

The implementation of unitized structure in the aerospace industry has resulted in complex geometries and load paths. Hence, structural failure due to three-dimensional mixed-mode fatigue crack growth is a mounting concern. In addition, the development of functionally graded materials has further complicated structural integrity issues by intentionally introducing material variability to create desirable mechanical behavior. Ti–6Al–4V β-STOA (solution treated over-aged) titanium is a functionally graded metallic alloy that has been tailored for superior fatigue crack growth and fracture response compared with traditional titanium alloys. Specifically, the near-surface material of Ti β-STOA is resistant to fatigue crack incubation and the interior is more resistant to fatigue crack growth and fracture. Therefore, Ti β-STOA is well suited for applications where surface cracking is a known failure mode. Advances in experimental testing have shown that complex loading conditions and multi-faceted materials can be tested reliably. In this paper, the authors will experimentally generate three-dimensional mixed-mode surface crack data in functionally graded Ti–6Al–4V β-STOA and comment on the effect of the material tailoring.