P.J. Webster

Laser Surface-contouring and Spline Data-smoothing for Residual Stress Measurement

We describe non-contact scanning with a confocal laser probe to measure surface contours for application to residual stress measurement. (In the recently introduced contour method, a part is cut in two with a flat cut, and the part deforms by relaxation of the residual stresses. A cross-sectional map of residual stresses is then determined from measurement of the contours of the cut surfaces.) The contour method using laser scanning is validated by comparing measurements on a ferritic steel (BS 4360 grade 50D) weldment with neutron diffraction measurements on an identical specimen. Compared to lower resolution touch probe techniques, laser surface-contouring allows more accurate measurement of residual stresses and/or measurement of smaller parts or parts with lower stress levels. Furthermore, to take full advantage of improved spatial resolution of the laser measurements, a method to smooth the surface contour data using bivariate splines is developed. In contrast to previous methods, the spline method objectively selects the amount of smoothing and estimates the uncertainties in the calculated residual stress map.

Synchrotron X-ray residual strain scanning of a friction stir weld

Synchrotron X-ray strain scanning has been used to determine the residual stress distribution over the cross-sectional area of a friction stir weld across an I-section made of extruded aluminium alloy type AA 7108 in the T79 condition. Measurements were made using the 311 reflection and a wavelength of about 0.35 A. The dominant longitudinal residual stress field is tensile within and to just outside the friction band where the field drops steeply. Further out there are regions of balancing compression falling to near zero at the outer edges of the sample. The longitudinal residual stresses vary within the range from -60 to 140 MPa. The transverse residual stresses are generally weaker but show a more marked through-thickness variation below the friction band. The residual stress maps show detailed features which are related to the heat flow from crown to root faces and differences between the advancing and retreating sides. The results demonstrate that synchrotron strain scanning is a practicable technique for determining non-destructively internal and through-surface residual strains in light element components. The synchrotron technique complements and extends the range at present covered by neutron and X-ray diffraction methods and is capable of substantial development.

The Measurement of Residual Stress in Railway Rails by Diffraction and other Methods

Residual stresses have been measured in a new roller-straightened railway rail and a worn ex-service rail. Synchrotron {211} lattice strain measurements at ID11 (ESRF) were used to map in-plane components of the stress tensor acting in cross-sectional rail slices. Stress maps made using laboratory X-rays and the magnetic measurement system MAPS, although coarser in detail, show similar trends. The validity of the measured data was examined using a stress balance requirement. Whilst generally true (to ±15 MPa), stress balancing was worst (±50 MPa) in regions with significant plastic deformation, suggesting that the measured {211} lattice strain had become uncharacteristic of the bulk elastic strain. Attributable to plastic anisotropy, this is a well-established issue with diffraction-based stress determination. To complement the in-plane stress measurements, the contour method was used to map the longitudinal stress component in a similar new rail sample, this component being relieved in the slices. On the...

Impediments to efficient through-surface strain scanning

Abstract The efficient measurement of through-surface residual strains using multidetector neutron strain scanners is constrained by instrumental effects which arise when the ‘gauge volume’ is only partly immersed in the sample. The three principal causes, wavelength variation across the incident beam, lateral displacement of the ‘effective centre’ of the gauge volume relative to the detector, and asymmetric clipping of the diffracted peak profile are discussed and quantified. Results are described of a systematic quantitative investigation of the instrumental aberrations that affect the L-3 spectrometer operating in its strain scanning mode at the NRU reactor at AECL, Chalk River, Canada. Comparisons are made with theoretical models which accurately predict the observed aberrations. Techniques for reducing and correcting for the effects are discussed.

Residual Strain Measurement by Synchrotron Diffraction

Third generation synchrotron X-ray sources such as the European Synchrotron Radiation Facility and the Advanced Photon Source (USA) have made very intense beams of very high energy X-rays available for the first time. At energies in excess of 60 keV penetration lengths of the order of centimetres are possible in most engineering materials. The associated low scattering angles limit the strain measurement directions available at depth. Gauge dimensions as small as microns and sub-second measurement times give the technique unique characteristics, making 2 and 3 dimensional strain mapping economically feasible. The current state of the art is reviewed and the potential assessed, primarily using illustrative case studies made at the ESRF. These include the measurement of near surface strains caused by peening, TIG welding stresses for the development of finite element models, the mapping of crack bridging during fatigue crack growth in Ti/SiC fibre composites and crack field mapping in 3D.

The Use of Combs for Evaluation of Strain-free References for Residual Strain Measurements by Neutron and Synchrotron X-ray Diffraction

The measurement of residual strain using diffraction techniques relies on the determination of a change in lattice parameter relative to a reference or “strain-free” lattice parameter. Therefore, o...

Residual stresses in alumino-thermic welded rails

Abstract Neutron strain scanning has been used to map the residual stress field that is generated in a railway rail by a standard gap alumino-thermic weld made using routinely specified procedures. The longitudinal and vertical residual stress fields in the sections well away from the weld are characteristic of many unwelded rails, being generally tensile in the head and foot with balancing compression in the web. In the vicinity of the weld the residual stress patterns are very different. At the top and bottom surfaces of the rail the longitudinal residual stress field is strongly compressive, which is generally beneficial in that it would tend to inhibit the initiation and propagation of fatigue cracks from surface defects. Just at the surface the vertical residual stress attenuates to zero but internally, in the web region, both longitudinal and vertical components of the residual stress field are strongly tensile, which increases the susceptibility of that region to crack initiation and propagation from internal material defects. This pattern is consistent with practical operating experience, which is that most of the small proportion of alumino-thermic welds that do fail do so as a result of porosity or inclusions in the weld. It is found that the ‘boundaries’ of the ‘weld type’ residual stress fields do not coincide with the boundary of the weld, nor of the heat-affected zone, but correlate reasonably well with the positions of the extremities of the mould assembly and with the location of the steepest longitudinal temperature gradients.

Neutron strain scanning of a small welded austenitic stainless steel plate

Abstract Neutron strain scanning has been used to determine the residual stress distribution over the area of the mid-thickness plane of a small austenitic steel piate across which a weld had been laid. The longitudinal and transverse stresses in the centre of the plate and weld are strongly tensile. Stresses normal to the plane of the plate are low everywhere except near the ends of the weld where tensile ‘hot-spots’ are observed. Near the edges of the plate the effects of boundary and balancing conditions are evident. Longitudinal and transverse stresses tend to zero along the edges of the plate to which they are perpendicular but become strongly compressive near the middle regions of the edges to which they are parallel. The results are compared with published data from other welds that were made and constrained differently. The accuracy of the technique is discussed and the results obtained by measuring with different reflections are compared. It is shown that neutron strain scanning is an acceptable, and unique, method for determining non-destructively the residual stress distribution throughout small austenitic steel components with sufficient accuracy to provide data for design purposes and to validate model calculations.

Residual Stress Measurements in Rails by Neutron Diffraction

The relatively new technique of neutron strain scanning has several unique advantages over most traditional methods of measuring internal strains. It is non-destructive and, in principle, measurements can be made and repeated at any point in any direction within a sample. In practice its range of its application is limited in thick section samples, such as long lengths of rail, by neutron beam attenuation. In such cases it is often necessary to reach a compromise between what is ideally preferred by engineers for use in stress analysis calculations and what it is practical or economic to measure using the technique. Examples are given of how neutron strain scanning may be applied, in a cost effective manner, to the problem of the measurement of residual stresses in railway rails. Results are presented of representative longitudinal, transverse and vertical stresses measured down the centre-lines of rail sections and of stress contours measured in rail heads. The data are of sufficient quality and quantity to be used to validate theoretical calculations and reveal details of the residual stress distributions in rails not obtainable by other methods.

Problems with Railway Rails

Neutron diffraction is now being applied to measure macrostrain distributions inside components of engineering interest. The technique has several unique advantages but there are practical constraints which limit its widespread adoption. Compromises are often necessary to bridge the gap between what is required by engineers for use in stress analysis calculations and what it is practical to measure using the technique, especially when large sized components are involved.