Materials Science Forum Vols. 571-572

Abstract: This paper summarizes the results of a neutron diffraction study of a single weld bead on a rectangular austenitic stainless steel plate. The measurement was carried out at SALSA, the engineering strain scanner at the ILL, Grenoble, France. The work has been carried out under the European NET project, and is a round robin exercise of residual stress simulation and validation benchmark in a stainless steel bead-on-plate (BOP) weldment. A monochromatic beam of wavelength 1.494 Å was used and the lattice spacing of {311} crystallographic plane was measured. The principal strain measured in the plate was corrected by measuring small cube sample of 3×3×3 mm3 as stress free reference. The measured strain was then used to calculate the principal stress distribution. Finally, the measured strain was compared with the strain measured in a similar specimen in a pulsed neutron source by the time-of-flight (TOF) technique.

Abstract: The effect of different welding sequences between a 4.5 mm thick AA 6156 T6 base plate and a 2 mm thick AA 6013 T6 clip – resembling a skin-clip joint of an airframe – using a 3.3 kW Nd:YAG laser is investigated. Under cyclic loading the breakdown of such T-joints happens at one end of the clip, which is due to local residual stress concentrations. Recent measurements indicated that tensile stresses could be lower at the run-in than at the run-out locations. For a deeper investigation of this effect sheets with different welding sequences were produced. One welding sequence was made with two starting points in the centre, and a second with starting points at the clip ends. Temperature measurements were made using thermocouples to verify the heat conditions for a finite element simulation of the welding process, which is used for predictions of the residual stress distribution. Actual values of the residual stress fields were determined by neutron diffraction. The influences of the welding sequence on the measured temperatures and the residual stresses are discussed.

Abstract: Neutron irradiation is known to have a considerable impact on the mechanical characteristics and the behaviour of materials and components. The distribution of residual stresses is one of the properties affected by irradiation. However, because of the difficulties in performing measurements in radioactive components, not many experimental data have to date been collected. At the High Flux Reactor (HFR) of the European Commission’s Joint Research Centre (JRC), a facility has been developed for residual stress measurements in steel specimens subjected to longterm irradiation. The objective of this development was to establish the neutron radiation induced changes in the residual stresses around welds in test pieces representative of the core shroud of boiling water reactors. Residual Stress measurements on such double-V butt welds in stainless steel plates after irradiation exposure have been performed by neutron diffraction using this facility. The comparison with measurements in non-irradiated companion specimens [1] showed that irradiation changed the distribution of residual stresses. The results suggested that the impact of irradiation varied with the distance of the test location from the specimen surface. On the basis of currently measured data we could not draw definitive conclusion about the influence of neutron dose and irradiation duration upon the stress modification.

Abstract: The conservative assumption of residual stress in highly restrained steel structures can lead to unnecessary repairs of defects in welded joints. This applies particularly with respect to high restraint, high yield strength thick sections welds because the assumption of yield strength residual stress is used in integrity assessments. By analysing the stress within components both before and after welding it is anticipated that a greater understanding of the residual stress field as a product of both the welding residual stress and the pre-welding residual stress can be made. This paper discusses a series of experiments designed to build up such a model of the changing stress field within a T-butt weld. Neutron strain scanning has been performed on unwelded flat and curved steel sections and a curved T-butt weld. To complement this, surface X-ray measurements have been carried out in order to gain a quantitative measure of the changing surface stress during welding and sectioning. The findings were that welding stresses dominate close to the weld, bending stresses dominate further from the weld.

Abstract: The level of residual stresses generated in fusion welds has been a major area of interest for many years. For steels, a major influence on the final state of stress is through martensitic transformation. This is because the martensitic transformation is accompanied by significant shear and volume strains. One way to mitigate the development of residual stress is by controlling the onset of the transformation such that the associated strain is able to compensate for thermal contraction all the way down to ambient temperatures. In the past it has only been possible to follow the evolution of the phase transformation during cooling of the weld metal using indirect methods such as dilatometry and differential scanning calorimetry. This paper describes the first work in which the phases present are characterized directly during the cooling of reheated weld metal at conditions typical of those encountered during welding by installing a thermomechanical simulator on a synchrotron diffraction beam line at ESRF.

Abstract: Diamond-cobalt composites are used for cutting tools. Residual stress after manufacture can reduce the lifetime of such composite cutting tools and, hence, the stress state needs to be well understood. Within this study, stress measurements on a cobalt diamond composite were made by SXRD to deduce stress states in the cobalt matrix using the (222) reflection. The application of different apertures allowed the investigation of stress in small areas of cobalt surrounding a diamond and in areas at different distances to a diamond. In the areas adjacent to the diamond increased residual stresses were found in the cobalt matrix. Furthermore, approximations for radial and tangential residual stresses have been derived which show to be different.

Abstract: Friction welding processes, such as friction stir welding (FSW) and inertia friction welding (IFW) are popular candidate procedures for joining engineering materials (including dissimilar pairs) for advanced applications. The advantages of friction welding include lack of large scale material melting, ability to join dissimilar materials, and relatively low propensity to introduce defects into the weld joint. For these reasons FSW and IFW have become the subjects of a number of studies aimed at optimising the joining operations to obtain improved joint strength and reduce distortion and residual stress. In the present study we used the diffraction of high energy polychromatic synchrotron X-rays to measure interplanar lattice spacings and deduce nominal elastic strains in friction stir welds between dissimilar aluminium alloys AA5083 and AA6082, and in coupons from inertia friction welds between dissimilar nickel-base superalloys IN718 and RR1000. Energy-dispersive diffraction profiles were collected by two detectors mounted in the horizontal and vertical diffraction planes, providing information about lattice strains in two nearly perpendicular directions lying almost in the plane of the plate samples mounted perpendicularly to the incident beam. Two-dimensional maps of residual stresses in friction-welded joints were constructed. Apart from the 2D mapping technique, the sin2ψ method (transmission) was also used in the case of inertia friction-welded joint between nickel alloys. Comparison between the two results allowed the variation of the lattice parameter with the distance from the bond line to be deduced. It was found that friction welding of two dissimilar materials with significant strength mismatch may lead to the creation of a region of compressive stress in the vicinity of the bond line, in contrast with the behaviour observed for joints between similar materials.

Abstract: AlSi7Mg/SiC/70p (AlSiC) is used for heat sinks because of its good thermal conductivity combined with a low coefficient of thermal expansion (CTE). These properties are important for power electronic devices where heat sinks have to provide efficient heat transfer to a cooling device. A low CTE is essential for a good surface bonding of the heat sink material to the insulating ceramics. Otherwise mismatch in thermal expansion would lead to damage of the bonding degrading the thermal contact within the electronic package. Therefore AlSiC replaces increasingly copper heat sinks. The CTE mismatch between insulation and a conventional metallic heat sink is transferred into the MMC heat sink. The stability of the interface bonding within a MMC is critical for its thermal properties. In situ thermal cycling measurements of an AlSi7Mg/SiC/70p MMC are reported yielding the void volume fraction and internal stresses between the matrix and the reinforcements in function of temperature. The changes in void volume fractions are determined simultaneously by synchrotron tomography and residual stresses by synchrotron diffraction at ESRF-ID-15. The measurements show a relationship between thermal expansion, residual stresses and void formation in the MMC. The results obtained from the in situ measurements reveal a thermoelastic range up to 200 °C followed by plastic matrix deformation reducing the volume of voids during heating. A reverse process takes place during cooling. Thus the CTE becomes smaller than according to thermoelastic calculations. Damage could be observed after multiple heating cycles, which increase the volume fraction and the size of the voids. The consequence is local debonding of the matrix from the reinforcement particles, which leads to an irreversible reduction of the thermal conductivity after multiple heating cycles.

Abstract: Alumina-zirconia multilayered ceramics have been proposed as an alternative for the design of structural ceramics with improved fracture toughness and strength reliability. During the processing of these laminates, significant residual stresses may arise due to the thermal expansion mismatch between adjacent layers. The correct evaluation of such stress distribution in the laminate may determine its range of application. In this work, the residual stress state in a layered material designed with five thick alumina layers of approximately 650 microns alternated with four thin alumina-zirconia layers of approximately 140 microns was estimated using different methods. A finite element analysis (FEM) was performed for stress evaluation in the bulk and an indentation method and X-Ray diffraction to account for stresses at the surface. Experimental findings show a constant stress distribution within the bulk for each layer, while at the surface stress position dependence is observed in the alumina layers, being the maximum tensile stresses near the layer interface. The accuracy of the results provided by each technique is discussed.