X-ray Diffraction Microtomography And Its Three-dimensional Quantification Of Crystalline Grains In Alloys

Hard polycrystalline materials such as metals,alloys and ceramics form the basis of most modern industry.The physical,chemical and mechanical properties of these materials are,to a large extent,governed by the microstructures and the interactions between them at grain scales.Therefore,as a direct bridge to contact material design and manufacturing process,the characterization of microstructures in materials is at the heart of materials science.Unfortunately,the conventional characterization methods at grain scales have some limitations.For instance,the characterization methods based on electron microscopy can only be used for the two-dimensional study of grains and require the sample to be physically sliced before investigation.The diffraction methods based on laboratory X-ray sources are mainly used for the characterization of average structural properties.Although the tomography-based imaging techniques are three dimensional and non-destructive,it is difficult to obtain an effective contrast due to the small difference in electron density between single crystal grains.Besides,these methods are completely unable to obtain crystallographic information.Consequently,new methods must be developed for three-dimensional,non-destructive,quantitative characterization of microstructures.To answer the requirements,a three-dimensional X-ray diffraction(3DXRD)based method is established on the imaging beamline at Shanghai Synchrotron Radiation Facility.The method can obtain the grain information such as position,size,morphology,phase,crystallographic orientation and average strain tensor threedimensionally,non-destructively and in situ.The innovative achievements of this thesis are as follows:Based on the imaging beamline,the 3DXRD experimental platform was established.Considering the actual conditions of the beamline,both the software and the hardware were developed specifically.In order to verify the feasibility of the method,experiments were carried out by using cubic and hexagonal standard samples.Furthermore,the influence of several factors on the image quality was studied systematically,and the relevant optimizationmeasures were put forward.These studies and improvements include: a)The low density of photon flux was the main reason for the loss of grain number in the refinement operation.b)The median filter could explicitly suppress the effect of salt-pepper impulsive noise and improve the results of peak seach.c)The accuracy of the calibration can be greatly improved by using the figureof-merit method,of which the efficiency was enhanced by using parallel computing.d)The contrast enhancement method could improve the quality of the near field images as well as the results of peak search and reconstruction.Without prior knowledge,a method based on 3DXRD was developed for the non-destructive identification of unknown minor phases with the simultaneous acquisition of the spatial distribution,sizes,morphologies and crystallographic orientations of individual grains.The method was confirmed to be feasible by simulations and experiments,respectively.According to systematic simulations on six combinations of different crystal system,the method is practicable for an extensive sample set.Experiments for a bulk sample of aluminum alloy AA6061 show that the crystal grains of an unexpected ?-Fe(austenite)phase can be identified.This result was further validated by X-ray micro-computed tomography and energy dispersive spectrometry.The relationship between reheating temperature and grain quantitative information was investigated experimentally.The 3DXRD method can obtain the quantitative information of a large number of grains in the sample,including their positions,volumes,surface areas,sphericities,orientations,disorientations and textures.The results show that the optimized reheating temperature is 640? at which the grains have the optimum size,morphology and orientation,that is,the large quantity of grains,relatively small and uniform size,high sphericity and random distribution of orientations.