Search

Natural fatigue crack formation and growth were studied in notched Al-Cu alloy coupons through high-resolution SEM fractography. The experiments were conducted under programmed loading conditions designed to induce microscopic marking of the crack formation and growth process under varying stress ratio and closure-free crack tip conditions. Control experiments were performed by switching between an air and vacuum environment. In air, varying the stress ratio from 0.74 down to 0.64 retards crack growth by up to a factor of five. This 'closure-free' stress ratio history effect totally disappears in vacuum, suggesting a significant environmental influence on stress ratio and its history. Crack-tip stress state appears to moderate environmental action, revealing a potential mechanism sensitive to residual stress. Consequently, crack closure, residual stress and crack front and plane orientation are identified as major load interaction mechanisms whose synergistic action controls fatigue under variable amplitude loading. The study also appears to suggest that as a consequence of the crack seeking the path of least resistance, load-sequence sensitive crack plane and front orientation may only induce retardation effects.

Fractures from tests on 2014-T6511 and 2024-T3 test coupons under specially designed programmed loading reveal voids with distinct fatigue markings. These ‘fatigue voids’ appear to form as a consequence of the separation of noncoherent secondary particulates from the matrix in early fatigue. The process of their formation is through the initiation, growth and coalescence of multiple interfacial cracks around the particulate. Such voids become visible on the fatigue fracture surface if and when the crack front advances through them. In vacuum, each fatigue void is the potential initiator of an embedded penny-shaped crack. The one closest to the specimen surface is likely to become the dominant crack, indicating that fatigue voids appear to be the likely origins of the dominant crack in vacuum. In air, the dominant crack forms at the notch surface and grows much faster, giving less opportunity for multiple internal cracks to spawn off from the innumerable internal fatigue-voids. Thus in air, fatigue voids do not appear to affect the fatigue process at low and intermediate growth rates. At high crack growth rates involving considerable crack tip shear, slip planes with particulate concentration offer the path of least resistance. This explains the increasing density of fatigue voids with growth rate. Very high growth rates signal the onset of a quasi-static crack growth component that manifests itself through growing clusters of microvoid coalescence associated with static fracture. Fatigue voids are likely to form in other Al-alloys with secondary noncoherent particulates. They have nothing in common with microvoids associated with ductile fracture.

Results of an experimental estimation of the crack closure stress, using electron fractography and laser interferometry for CT specimens of an aluminium-copper alloy are presented in this paper. Crack closure stress was estimated at stress ratios of 0.1, 0.3, 0.5 and 0.7 using programmed load sequences. In addition, the crack closure stress was estimated for a simple random load sequence. Scanning and transmission electron fractography provided a similar estimate of crack closure, with the later technique providing better quality fractographs. Laser-interferometry based estimates were closer to fractographic estimates when the crack-tip is very close to the micro-indent axis. In addition, laser-interferometry based estimates of the crack opening stress for a random load sequence, confirmed the possibility of a cycle-by-cycle variation of the crack opening stress.

A numerical investigation of the fatigue crack closure phenomenon has been performed by an elastic-plastic finite element analysis. Computer software was developed to consider many aspects affecting plasticity-induced crack closure. Linear and power-law hardening models are considered in the finite element analysis. The paper presents results from the study carried out on compact tension (CT) coupons at various crack lengths corresponding to different loading conditions. Finally the results of the analysis are compared with the experimental estimates of fatigue crack closure levels obtained from laser interferometry, scanning-electron and transmission-electron fractography presented in Part I of this paper on identical specimens.

A procedure was developed to determine the bridging fiber stress distribution from crack opening displacements measured during crack growth. The bridging fiber stress range is near-linear along the crack and non-zero stresses at the crack tip. The magnitude of the bridging fiber stress range is proportional to the applied stress range.

In using a representative volume element of a unidirectional composite, two out-of-plane boundary conditions represent edge and internal regions of a transversely loaded SCS-6/Timetal® 21S composite. A plane stress condition is proposed to represent a stress-free edge region of the composite, while a generalized plane (uniform longitudinal) strain condition is considered for internal regions. Numerical simulations with elastic-plastic matrix behavior and several different fiber-matrix interface strengths reveal a complex interaction of residual stress, fiber-matrix separation and matrix inelastic behavior which are all dependent on the out-of-plane boundary condition. Several permutations of plane stress and generalized plane strain solutions with several fiber-matrix interface strengths fail to accurately capture the non-linear behavior measured from experiments. The bounding of the experimentally determined transverse response between the plane stress and generalized plane strain solutions suggests that the transverse response may be a combination of both solutions. Strain measurements from the transversely loaded composite support speculation that the plane stress solution better represents edge regions of the composite than a generalized plane strain solution. Likewise, the generalized plane strain solution represents the internal strain state of the composite better than the plane stress case at low loads. Photomicrographs before and after transverse loading show fibers which have protruded from the edge of the specimen; thus, the strain state within the composite is transitioning to a less constrained condition.

Basic research laboratories typically perform a variety of material tests and obtain the associated data to model material behavior phenomena and develop life prediction methodologies. In this research environment, a mechanical test automation system must meet challenges that are not always present in an industrial testing setting. For example, real-time crack closure load analyses, at the present time, are not widely performed in industrial crack propagation testing. In the research environment, however, on-line crack closure studies are used to make decisions in real-time about changes in test conditions. A previous paper described the overall system strategy and hardware and one of the crack propagation software modules from the fourth generation of the material analysis and testing environment (MATE) automation system. The present paper discusses selected methodologies that the current (fifth) generation of the MATE system uses to meet the challenges posed while automating research style mechanical tests. The methodologies addressed in this paper include waveform generation and synchronization for cyclic, monotonic, and thermomechanical (TMF) testing as well as specimen damage computation for self-similar cracked geometries.

TIMETAL 21S matrix composite reinforced with silicon-carbide (SCS-6) fibers is a potential candidate material for the high temperature structural components of advanced aircraft. However, a considerable amount of mechanical property characterization is required before this material can be used for aerospace structures. This paper discusses the characterization of the sustained load behavior of SCS-6/TIMETAL 21S composite under simulated service conditions. TIMETAL 21S, reinforced with approximately 35% SCS-6 fibers (by volume) was used for this study. The tension creep specimens were fabricated from [0] 4 , [90] 4 , [0/90] s , and [0/′45/90] s laminates. Sustained load creep tests were conducted under cold grip conditions at 650, 760, and 815°C at various stress levels. The creep response for [90] 4 , [0/90] s , and [0/′45/90] s layups exhibited generic creep curves with three distinct stages: (1) an early region of primary creep, (2) a linear region similar to a steady stage, and (3) a relatively short region similar to a tertiary stage. In contrast, the creep response of the [0] 4 layup showed hardly any tertiary region and the failure was abrupt just after the secondary stage. The total strain to failure for all orientations, except the [90] 4 , was near the ultimate strain for fiber failure. Accelerated rate of damage accumulation was observed with increasing stress and temperature that resulted in higher creep strain under these conditions. A simple creep model based on the relaxation of matrix stress in the primary and secondary creep stages was developed to predict the creep response of the unidirectional SCS-6/ TIMETAL 21S composite at 650°C. The model predictions correlated well with the data for stresses less than 700 MPa. The optical and scanning electron microscopic (SEM) analyses of the failure surfaces showed multiple crack initiation sites near fiber/matrix interface. Two distinct zones of failures were observed in all cases. The flat zone, with little or no fiber pull-out, represented the stable crack growth region. The rugged zone, with significant fiber pull-out, was caused by overload failure. In all cases, the damage (crack) initiated from the edge. For [90] 4 and [0/90] s , the crack initiation sites were observed to be close to the 90° fiber/ matrix interface, while in the case of [0/′45/90] s ; the crack initiated predominantly from the 45° fiber matrix interface. The fibers that were damaged during the fabrication process provided the crack initiation site for the [0] 4 orientation. These edge cracks enhanced the environmental access to the fiber/matrix interface that resulted in progressive damage of the interface and the fiber. Data plotted on a Larson-Miller Parameter (LMP) chart enabled a direct comparison of the creep resistance of all the layups tested. The creep resistance of the [0] 4 layup was significantly greater than that of the other layups. The creep rupture life of specimens with 0° fibers was governed by the failure of unidirectional fibers. An empirical model was developed to predict the creep rupture life of all layups containing 0° fibers based on the LMP approach assuming that only the 0° fibers carry the load for the majority of the life.

Fracture tests were performed with carbon/polymer laminates and analyzed for the purpose of developing translaminar fracture toughness test and analysis procedures. Notched specimens were tested of two types of symmetrical layups--quasi-isotropic [0/45/90] and [0/90]; two carbon fiber/epoxy materials--a relatively brittle T300 fiber/976 epoxy and a tougher AS4 fiber/977-2 epoxy; two laminate thicknesses--2 mm and 4 mm; and three specimen configurations--the standard three-point bend and compact configurations, and an extended compact specimen with arm-height to specimen-width ratio of 1.9. Stress and displacement expressions were obtained for the extended compact specimen, including those for stress intensity factor, K, and crack mouth opening displacement, V, in terms of relative notch length, a/W, and for a/W in terms of V. Relationships for the bending stresses that control self-similar and off-axis cracking for the extended compact specimen were derived.

Fatigue tests of middle-hole tension specimens of SCS-6/TIMETAL 21S composite (silicon-carbide fibers reinforcing a matrix of Ti-15Mo-2.6Nb-3Al-0.2Si alloy) were performed on three laminate architectures: unidirectional, cross ply, and quasi-isotropic. Specimens were tested over a range of stress levels, and fatigue damage was documented in situ by macrophotography and direct-current electric potential drop measurements. Typically, failure evolved by the formation of a few dominant cracks at the notch that propagated into the composite matrix and, in many instances, were substantially affected by unbroken fibers bridging the cracks. Fractographic and failure mode characterization revealed key differences in the effectiveness of crack bridging in the three laminates. A shear-lag crack-bridging model was shown to correlate crack growth data in the laminates based on an empirical value of fiber/matrix interfacial shear stress. Crack-bridging stress distributions were predicted using the shear lag model and verified by comparing the predicted crack opening displacement profiles with measurements made using a laser interferometric displacement gage system. Implications of the results are discussed with respect to the potential use of these materials in practical structural applications.

Experimental crack growth rates were determined in Inconel 718 middle cracked tension [M(T)] specimens at 649°C under conditons of high frequency (10 to 100 Hz) and high load ratio, R. Under these conditions, the materialexperiences cycle-dependent crack growth as well as time-dependent crack growth. At very high R values approaching unity, the observed growth rates are lower than those obtained under sustained load at the same mean load in the absence of the superimposed cyclic loading. Tests on compact tension [C(T)] specimens at lower frequencies were used to demonstrate the existence of three regions of behavior-cycle-dependent, mixed mode, and time-dependent. A linear cumulative damage model was used to predict the growth rates due to combined cycle-dependent and time-dependent mechanisms. The model was developed from 427°C data for the cyclic term and sustained load crack growth data for the time-dependent term. Although the model cannot predict the synergistic effect at high R, it provides a reasonable representation of much of the data. The authors conclude that the use of low-temperature data for the cyclic term is inadequate for representing the threshold values and growth rates at low AK values at the higher temperature.

The FIDEP2 (Finite-Difference code for Elastic-viscoplastic analysis) is a PC-compatible, user-friendly computer code developed in-house at the Materials Directorate, Air Force Research Laboratory, Wright-Patterson AFB, Ohio. The program is capable of predicting micromechanical stresses in metal matrix composites under complex thermal and mechanical loading histories. The FIDEP2 is a generalized version of the FIDEP program, which was limited by the concentric cylindrical geometry and elastic-plastic constitutive model. The FIDEP2 program incorporates different loading histories, micromechanical models, and constitutive models in a modular form to allow for easy implementation of new requirements. Operation of the FIDEP2 program is straightforward, requiring a loading history file and a material properties data file. Input data are in free format, and some descriptive titles are allowed.

Relaxation of shot peen induced compressive residual stresses due to thermal exposure was measured using X-ray diffraction. The material used in this study was a hot isostatically pressed (HIP) powder metal (PM) IN100 nickel base superalloy. A total of 14 IN100 samples were shot peened to an Almen intensity of 6A using MI-170-R shot with 125 % coverage. The sample dimensions were nominally 16×13×4-mm thick with an irradiated X-ray region of 8×5 mm. Residual stress measurements were made at the surface and at nominal depths of 12, 25, 50, 75, 125, 175, 250, and 350 microns. The shot peened samples were thermal exposed at two temperatures (650, 704°C) and a range of exposure times (0.5–300 h). Residual stress measurements on shot peened samples without thermal exposure were used as a basis for comparison. The relaxation of shot peened compressive residual stresses under purely thermal loading was examined. The residual stresses exhibited an initial rapid decrease on the surface and in the depth at both temperatures. However, continued thermal exposure produced little or no change in surface residual stresses while peak compressive stresses in the depth continued to relax with time at both temperatures. In all cases of this study the retained peak compressive residual stress after thermal exposure was greater than 50 % of the baseline value.

Fractographic studies on Al alloys and a nickel-base super-alloy point to the combined action of three major load-interaction mechanisms in controlling crack growth under variable-amplitude loading. These are crack closure, residual stress, and crack front incompatibility. Crack front incompatibility attenuates crack tip response to applied load and can also increase crack closure stress. Notch root residual stress affects crack closure, while crack tip residual stress moderates environment-enhanced fatigue crack extension. The latter effect disappears in vacuum. Experiments designed to isolate the effect of individual mechanisms provide a framework to model their synergistic action, consistent with microscopic observations of crack growth bands in variable amplitude fatigue.

Mechanical stress wave propagation in a long, thin, isotropic, elastic rod containing a single transverse edge crack is studied theoretically and experimentally. It is found that one-dimensional wave theories, coupled with an effective “compliance” of the cracked region, predict reasonably well the observed dynamic strains induced by a longitudinal impact. Remote sensing of relatively deep cracks appears feasible and consequently the procedure may be useful clinically for determining the rate of healing of partially cracked bone.

Two empirical equations were developed that express displacement across the crack in a compact type specimen as a function of crack length and distance behind the crack tip. The coefficients in the equations were generally determined by least square error fit to crack surface displacements obtained from finite element analysis. In one expression, simple types of functions which depend on either crack length or distance from the crack tip were combined to form an equation that is similar to a series expansion in two separate variables. The second equation was developed from a combination of the far field behavior of the displacements and a near field fit which incorporates the near tip analytical displacement behavior. Both equations have application to a broad range of values of crack length and distance from the crack tip. Either equation can be used in conjunction with measurements of displacement across the crack surface to indirectly determine the crack length. Keywords: Fracture mechanics; Tables(Data).