T. Zhai

A crystallographic mechanism for fatigue crack propagation through grain boundaries

A crystallographic model is proposed which takes into account both crack-plane twist and tilt effects on crack retardation at grain boundaries. The twist and tilt angles of the crack-plane deflection at a grain boundary are the key factors that control the path and growth rate of a short crack. Because of crack-plane twist, the area between the traces on the grain-boundary plane of the crack planes across the boundary has to be fractured in order for the crack to propagate through the boundary. This presents significant resistance to crack growth. As the area to be fractured increases with the extent of crack growth beneath the surface of observation, the grain boundary could still resist crack growth after the crack tip has passed the grain boundary on the surface, until the crack propagates through the whole boundary below the surface. A grain boundary with a large twist component could cause a short crack to arrest or branch. Studies of short fatigue crack growth in an Al-Li 8090 alloy plate provide evidence that supports the model.

A self-aligning four-point bend testing rig and sample geometry effect in four-point bend fatigue

Abstract A self-aligning four-point bend testing rig was designed and made which can minimise the possible misalignment associated with a four-point bend test and be used to study the fatigue of materials both at room and elevated temperatures. The stress distribution between the inner-rollers in a specimen under four-point bend, that is the nominal pure-bending section length, was analysed with respect to various load-span/specimen-thickness ratios ( t / h ) and support-span/load-span ratios ( L / t ) using a finite element method. It was found that the stress distribution could vary with both t / h and L / t . It was found that values of t / h and L / t between 1.2 and 1.5 and between 4 and 5, respectively, were the optimum testing geometry which led to a relatively uniform stress distribution consistent with the value calculated by beam theory. Fatigue tests ( R =0.1 and frequency=20 Hz) were carried out on samples with different thickness in a peak-aged 8090 Al–Li alloy using the rig. The results appear to support the finite element results. The S – N curve of the 8090 Al–Li alloy was measured using the optimum testing geometry in the four-point bend, and it was found to be consistent with that reported in the literature.

FATIGUE DAMAGE AT ROOM TEMPERATURE IN ALUMINIUM SINGLE CRYSTALS-III. LATTICE ROTATION

Abstract Lattice rotation was observed in different scales in aluminium single crystals fatigued in push-pull in air at room temperature, a constant shear stress amplitude 4 MPa, zero mean stress and frequency of 20 Hz. Using the channelling contrast technique in SEM, contrast of grey and dark bands consistent in dimensions with those of PSBs was observed on the surface sectioned parallel to the Burgers vector b in an Al single crystal after 1.2 × 10 6 cycles, suggesting that there was lattice misorientation (rotation or tilt) between PSBs and the matrix in the specimen. It might be caused by local net material movement due to irreversible slip or non-uniform deformation in PSBs. Lattice rotation (about 6°) always appeared between a macroband and the matrix relative to the normals of two perpendicular surfaces of the specimens, as a result of net irreversible slip in one direction of b in PSBs. Often more cracks were found in a positive macroband than a negative one. Deformation bands coarser than PSBs and smaller than macrobands were also found on the surface containing b by the scanning acoustic microscope. They deviated slightly from the direction of the PSBs and were probably formed to release the internal stresses between macrobands and the matrix due to macroband formation. The macroband effect is likely a general metallographic characteristic of unidirectional fatigue in planar-slip metal single crystals. The lattice rotation was probably one of the key factors controlling crack initiation and early propagation in the Al single crystals.

Fatigue damage in aluminum single crystals—I. On the surface containing the slip burgers vector

An aluminum single crystal with the axial direction of [25{bar 1}] was fatigued in push-pull at the constant resolved shear stress amplitude 4 MPa, frequency 20 Hz and room temperature. Microcracks, microvoids, macrobands, extrusions and intrusions were observed on the side-surface containing the Burgers vector b. Most microcracks were opened, and were within, but approximately perpendicular to, PSBs. Slip steps were found in the extrusions and intrusions. There was net irreversible slip in one direction in most PSBs. Some short cracks along the PSBs on the side-surface were also observed at 5 {times} 10{sup 6} cycles. These observations indicate that, without the aid of the surface roughness of PSBs, cracks can still be nucleated, and that, apart from the notch effect of a PSB, there are other factors controlling crack initiation in single crystal aluminum. There may be an internal tensile stress existing in a PSB in the direction of b, and a shear stress applied by the specimen grips in the specimens due to the irreversible slip in one direction in PSBs. These stresses and the applied stress are responsible for the formation of microcracks, microvoids, extrusions, intrusions and macrobands on the side-surface.

Influence of grain orientations on the initiation of fatigue damage in an Al–Li alloy

The variation in microstructure and texture in a rectangular bar extruded from a billet of spray‐cast 8090 Al–Li alloy has been examined. The fine grain size of the as sprayed billet and the moderate extrusion ratio (≈ 25 : 1) were seen to cause geometric dynamic recrystallization (GDR) in regions of higher strain towards the edge of the bar. The grain morphology varied from the expected elongated grains at the centre of the bar to equiaxed grains where GDR occurred at the bar edges. A <111> + <100> double fibre texture, significantly distorted towards rolling components and varying through the bar thickness, was found using electron backscatter diffraction. Fatigue resulted in a high density of short secondary cracks, many of which had arrested at grain boundaries. The cracks preferentially nucleated in grains from the <100> fibre texture corresponding to high Schmid factors.

Measuring Short Cracks by Time-Resolved Acoustic Microscopy

Detecting defects, for example cracks, (1) is important in predicting the lifetime of a material. The growth behavior of short cracks(2) plays an essential role in the lifetime of a component, since the lifetime is mainly controlled by the time required for a crack to grow from a certain initial size to about 1 millimeter. Cracks are defined as short when for example the crack length is small compared with the microstructure of the specimen or when the crack is simply shorter than ≈0.5 mm.(3) The growth of short surface breaking cracks can be measured by light microscopy (LM) or scanning electron microscopy (SEM). A common method of studying short cracks is a replica technique based on taking several plastic replicas at various stages of crack growth and subsequently examining these replicas with LM or SEM.(4,5) The disadvantage of these techniques however, is that they give information about crack development only on the surface of the specimen, so that the depth of the crack has to be determined indirectly by assuming the shape of the crack. For long cracks this may be appropriate because a local change in propagation direction does not alter the overall crack geometry on which the driving force of the crack depends.(6–8) However for short cracks, a change in propagation direction can alter the crack geometry significantly and thus change the driving force for the crack propagation. If its size is comparable with the microstructure of the material, a deflection of the crack at a grain boundary can alter the overall crack geometry. To understand the behavior of short cracks, it is therefore necessary to measure their three-dimensional growth. This can be achieved by using acoustic waves, which can penetrate into the material. In this way the crack depth can be measured directly.

Mixed mode crack mouth reflection in time-resolved acoustic microscopy of short fatigue cracks in single crystal aluminium

An Al single crystal with the axial direction of (452) was fatigued in air, at a constant resolved shear stress amplitude (4 MPa), a frequency of 20 Hz and at room temperature. Time-resolved measurements were then carried out on short cracks in persistent slip bands (PSBS) on the top surface, which has the largest slip steps, using the acoustic signal from the crack, when the acoustic lens was scanned over the crack sending convergent acoustic beams down the specimen, was detected and interpreted quantitatively. The signal results from the reflections of a mixed mode surface wave (Rayleigh longitudinal lateral wave) from the crack mouth. At 1.2*106 cycles, two short cracks in the PSB were measured to be 17 and 27 mu m in depth. The angle with the specimen surface was observed to be about 54 degrees , compared with a value of 51 degrees determined by an X-ray method.

DETERMINATION OF SHORT CRACK DEPTH WITH AN ACOUSTIC MICROSCOPE

For the prediction of the lifetime of any component, subjected to alternating stresses, the knowledge of the growth behavior of defects is essential. Most methods of monitoring the propagation of short cracks are confined to measuring the length of the crack on the surface [1]. The depth of the crack must be determined indirectly, assuming the shape of the crack. Acoustic waves, on the other hand, offer the possibility of measuring the depth directly, since acoustic waves can penetrate into the material. This allows the measurement not only of the growth behavior of fatigue cracks on the surface, but also changes of the crack geometry inside the specimen. Current applications of direct acoustic monitoring of crack growth have been developed for cracks of the order of millimeters. One acoustic depth measurement technique is the Time-of-Flight-Diffraction (TOFD) technique [2–4], which is based on timing measurements of the scattered signals from the defect. Our investigations are concerned with the application of TOFD technique for the depth measurement of short cracks (70–200 μm in surface length) using a scanning acoustic microscope (SAM) [5–6]. Depth measurements were first carried out on cracks in the transparent material polystyrene. This allows a direct comparison between acoustic and optical depth measurements. Subsequently, the depth of fatigue cracks in an A1 alloy were measured, and the acoustic measurements were compared with direct measurements of the crack geometry by sectioning the crack.