The changes happening during in-situ isothermal heating of an experimental Al 0.1% Mg alloy can be observed using Oxford Instruments’ NordlysMax2 and AZtec® EBSD system, which is designed for rapid data acquisition at higher temperatures.
The results show good agreement with ex-situ isothermal heating analyses on folded Ni sheets, with the formation of larger grains in areas of least strain and finer grains in the compressed and tensile regions.
The sample preparation involved cold rolling of an experimental Al 0.1% Mg alloy to 2mm thickness, followed by annealing to form grains of about 100μm. The resulting sheet was then folded to form a 90° bend, at a right angle to the rolling direction.
The folded sheet was then cut into sections, which were mechanically polished to roughly 1mm thickness and then electrolytically polished. The specimen thickness was kept to a minimum to lower the heat load for the in-situ heating stage.
A Manchester University-developed in-situ heating stage was used in this analysis. Carbon cement was used to mount the sample in a FEG-SEM. The sample was then subjected to slow heating, to a holding temperature of 295°C. EBSD data were collected during the heating of the sample at intervals of roughly 15min. This was followed by the acquisition of a set of 100 maps one after the other over the next 15hr.
A step size of 5μm was used for EBSD mapping, covering an area of 2 x 1.8mm at an acquisition speed of 280 patterns per second. The data acquisition and processing was performed using the NordlysMax2 EBSD detector coupled with AZtec software. The SEM was run at 20kV acceleration voltage with a 12nA probe current. The heating cycle used during the analysis is depicted in Figure 1.
Figure 1. Heating Cycle during the experiment
Figures 2a-d show a chain of band contrast, local misorientation, IPF Z coloured, and grain boundary EBSD maps acquired at the onset of the heating cycle. These maps reveal the initial deformed microstructure 80μm in diameter, as well as a strain gradient that can be inferred from the bend curvature of the sample.
Figure 2. a-d) EBSD maps at the start of the heating cycle. Map a) pattern quality, b) local misorientation, c) IPF Z colored and d) grain boundary. Grain boundaries >1.5o aqua, >10o black and 3=red.
The initial deformed microstructure has been changed into a nearly strain-free microstructure with a mean grain size of about 160μm at the end of the annealing cycle, as shown in Figures 2e-h.
The final grains at the compressed side are smaller than the grains in the tensile region, with grain size of more than 300μm in the middle region. Some sigma-3 boundaries are also observed, as shown in red in Figures 2d and 2h, respectively.
Figure 2. e-h) EBSD maps at the end of the heating cycle. Map e) pattern quality, f) local misorientation, g) IPF Z coloured and h) grain boundary. Grain boundaries >1.5o aqua, >10o black and 3=red.
The microstructure grows in two distinct but overlapping events:
Nucleation and growth of several grains in the highly deformed compressive and tensile surface regions, with growth taking place in the deformed area. This represents a strain-induced boundary-type migration, which extends for about 90min and consumes the majority of the highly deformed regions.
Nucleation and growth of a few grains before the first set of recrystallized grains, which evolve into less deformed grains. These grains consume less deformed grains until they impinge and then grow towards the center of the sheet.
When the second event occurs, the initial set of grains remains constant, as shown in Figure 3, which plots the grain area of selected grains 1-7' undergoing growth against time. This plot reveals the nucleation and growth of the large grains in the middle of the sheet 5-7' after the size of the initial smaller grains 1-4' in the highly deformed areas reaches maximum.
Figure 3. Plot of grain area with time of selected grains in the compressed and tensile regions
Figure 4 shows the initial locations of the selected grains 1-7, with Figures 4 a-c revealing the respective 3D orientation views in EBSD maps. Figure 4d depicts the final locations of the new grains 1-7'.
All the initial nuclei labeled 1-4, in Figure 4a, are largely deformed and have a number of zero solutions. The arrowed location 4 in Figure 4a has no solution where a near (111) <110> oriented grain labeled as 4' in Figure 4d has nucleated and grown.
Figure 4. a, b and c) EBSD IPF Z maps with insets showing the 3D view of crystal orientations and location of original grains and d) final grain with their respective 3D views.
Nuclie in the other three grains 1-3' have no relationship with the original orientation 1-3 and are either near cube or (111) <110>orientation. Likewise, grains 5-7' from the later stage, illustrated in Figure 4d, are totally different in orientation from the parent orientations labeled as 5 and 6-7 in Figure 3b and c, respectively.
The results clearly demonstrate the ability of NordlysMax2 EBSD detector coupled with Aztec to analyze microstructures and better understand recovery and recrystallization under strain gradients.
This information has been sourced, reviewed and adapted from materials provided by Oxford Instruments NanoAnalysis.
For more information on this source, please visit Oxford Instruments NanoAnalysis.