Footprints of deformation mechanisms during in-situ x-ray diffraction

Nanocrystalline metals have been an area of great interest in recent years due to their enhanced characteristics. One of the most striking is implied by the Hall-Petch relation: with decreasing grain size the material becomes stronger. This promise is fulfilled in nanocrystalline metals which display strengths up to 10 times higher than their coarse grain counter parts. With such high strengths they show potential significance for engineering and structural applications. However, for these materials to be of use in structural applications it is paramount to improve their low ductility. This in turn requires a thorough understanding of the fundamental principles governing plastic deformation at the nanograin length scales. In coarse grain metals, with grain sizes in the micrometer range, plastic deformation is governed by dislocations that are generated by sources within the grains, propagate and interact with pre-existing structures and also with each other. Upon unloading at a given level of deformation all the dislocation segments that have not yet annihilated make up the final microstructure of the deformed state. As the operation of a traditional dislocation source is grain size dependent there would be a critical length scale below which such a source can no longer operate. For FCC metals such as Ni this grain size is between 20-40 nm and thus the mechanisms governing classical plasticity in coarse grain materials would appear to break down in the nanocrystalline regime. Whether plasticity in nc materials with average grain sizes below 100 nm and an intrinsic grain size distribution is still governed by dislocation mediated processes is still an open question. Post-mortem TEM observations have not found dislocation debris and in-situ TEM methods have noted some dislocation activity during deformation. Molecular dynamics studies have, however suggested that in this nanograin regime and with the absence of internal dislocation sources, GBs of nc FCC metals can act not only as a source but also as a sink for lattice dislocations. These suggestions motivated us to investigate x-ray diffraction (XRD) signature through an in-situ experiment designed specifically for providing insight into plastic deformation behaviour of bulk nanocrystalline materials during deformation. Thanks to the high intensity of the radiation at the Swiss Light Source and the development of the Microstrip detector allowing time-resolved measurements and covering an angle of 60° we have built-up an experiment which allows us to perform tensile deformation in-situ. The uniqueness of the experiment is that it allows both a continuous and simultaneous monitoring of peak position and peak broadening during tensile testing. The Microtensile Machine mounted at the SLS had been specifically designed for this experiment and allows a variety of different sample shapes and sizes to be mounted and observed in the beam. Measurements can be conducted using a range of different strain rates and modes of testing: continuous tensile tests, stress relaxation experiments, strain rate jump tests, creep tests, as well as the experiment to be presented: load/unload (L/U). The elongation is measured using a CCD camera with image recognition software and allowing free access to the central part of the sample gauge. The X-ray beam used for the presented experiments was a monochromatic 17.5 keV beam focused to illuminate about 0.5 mm of the central part of the 1.7mm gauge length of the sample. Both the nanocrystalline materials chosen for this experiment are fully dense and chosen with an eye for their typical grain size. Electrodeposited (ED) nanocrystalline (nc) Ni purchased from Goodfellow has an average grain size of ∼30 nm as determined by XRD using the Williamson-Hall method. This grain size is confirmed in TEM observations that also indicate a narrow grain size distribution of equiaxed grains with some grown in twins in the as-prepared material. The ultra-fine grained (ufg) high pressure torsion (HPT) Ni was produced by Prof. Zehetbauer at the University of Vienna and has an average grain size of ∼300 nm with evident sub-grain boundaries visible in the TEM. During the in-situ measurements the XRD data is related to the tensile data through the individual peak positions and the respective peak broadening. Peak broadening originates from limitations in the spatial extent of the coherent scattering zones and from the presence of inhomogeneous strains. The first type of broadening is diffraction order independent whereas the second is diffraction order dependent allowing a separation of the size and strain effects on the broadening. An analysis of 6 different diffraction peaks allows a continuous monitoring of the structural changes during deformation. The results of L/U experiments conducted on these materials show obvious and striking differences. By far the most interesting result is the full reversibility of the peak broadening evident in L/U experiments conducted on ED-nc-Ni. The L/U experiments on ED-Ni and HPT-Ni where performed using a strain rate of 6 × 10-5 s-1 and had all the unload cycles conducted in the fully plastic regions of the respective stress – strain curves. The ED-Ni sample failed in the third loading cycle and a visual inspection of the gauge exhibited no pronounced necking in the course of L/U experiment. The peak broadening analysed in terms of the full width at half maximum (FWHM) shows total reversibility of the broadening upon unloading unlike that of cg materials. The full reversibility of the broadening for all measured peaks implies that no permanent residual strain is accumulated during deformation indicating a lack of dislocation debris. These observations are supported by suggestions from MD simulations where dislocations nucleate, propagate and are absorbed at grain boundaries. Equivalent measurements conducted on HPT-Ni indicate that the yield stress for HPT is lower than for ED-Ni. The sample broke during the 4th loading cycle exhibiting substantial softening and pronounced necking. Upon unloading, the peak broadening for HPT-Ni is not reversible as is the case for ED-Ni, suggesting a permanent build up of a dislocation network in the UFG regime and thus a classically based dislocation based plasticity. In summary, during this thesis a new in-situ technique was developed allowing greater insight into deformation mechanisms of bulk nc and ufg materials. In its application to investigations on nc and ufg Ni, this thesis has provided evidence of a reversible peak broadening for nanocrystalline materials confirming that no residual inhomogeneous strain is accumulated during deformation and that upon unloading a dislocation based plasticity leaves no footprint within the nanosized grains. Moreover, for the first time this work shows that the flow softening behaviour has different origins in the nc and ufg regimes. The experiments conducted point to future areas of investigation, in particular, the apparently special behaviour during the first loading cycle where the material yields for the first time, and the order dependent peak position and width behaviour during loading and unloading.