The movies presented below were recorded in 1979 at Imperial College, London, in cooperation with Sue Burrows and supervised by John Humphries. Results were published in Urai et al., (1980) and Urai and Humphries (1981). Most of the processes seen in the movies were described in detail in these two papers.

Camphor (C₁₀H₁₆O) occurs in several forms. It melts above 425 K, it is Cubic below 425 K, Rhombohedral I below 365 + 7 K, and Rhombohedral II below 243 K. The actual transformation temperatures are dependent both on the purity of the material and the proportions of d- and l-isomers present.

Camphor of commercial purity from two sources (Hopkin and Williams, and BDH) vas used. Although the deformation mechanisms were similar in both, the phase transition temperatures varied by about 10 K.

The granular material was cold-pressed into sheets ~ 0.15 mm thick between moist polished steel plates. The main "trick" of this technique is to get the camphor wafer to detach from both steel plates upon unloading and separating the plates. This was mostly a trial and error process, yielding a usable wafer after several tries. Specimens of approximately 7 x 7 mm were cut from these wafers and placed in the deformation cell. Silicon oil was used to minimize friction between specimen and glass, giving a more homogeneous deformation. First, the glass plates were pressed together very gently, taking care not to break the wafer which was usually slightly non-planar at this stage. Then the sample was heated into the cubic phase field where it became very soft and ductile, and the pressure on the glass plates slowly increased to iron out small irregularities in the sample surface, and slowly cooled. The phase transition produced large grains of the rhombohedral phase which were generally elongated but without preferred orientation.

This movie is presented in two parts. The initial microstructure is shown in the first frames of the movie, shot at room temperature with crossed polarizers and the gypsum plate. The movie was recorded in the part of the (larger) sample which was closest to the moving piston. Slip in this temperature range only occurs on the (001) plane, and thus only grains in an easy glide orientation (grain C) could undergo substantial glide. Therefore grains oriented with their (001) plane parallel or perpendicular to the specimen plane deformed in different ways. Grains in easy glide orientation first show fine linear features interpreted as slip lines, and later develop kink boundaries which slowly migrate through the grain; this can be compared with fig. 2 of Urai et al., 1980. Grains in a hard orientation (grain A and B with the slip plane parallel to the line of sight but perpendicular to the shortening direction, grain D with the slip plane parallel to the shortening direction) deformed by twinning. Already at low strains, small equiaxed grains formed by dynamic recrystallization at regions of strain heterogeneity such as kink boundaries, twins and grain boundaries.

With further deformation grain A starts to recrystallize, and finally a dextral shear zone (arrows) oriented at 45 degrees to the shortening direction is formed, cutting through the "hard" grains A and B.

This experiment is similar to the previous one in its initial microstructure. The difference lies in the slightly higher temperature of about 35 degrees C. This makes the material less anisotropic and deformation becomes more homogeneous. Also here the movie was taken from the part of the sample close to the moving piston coming from the right. This experiment can be best compared with fig. 6 of Urai et al., 1980.

Immediately after the start of the deformation recrystallization starts with the migration of existing grain boundaries. This produces an increasing amount of dynamically recrystallized material, in which a steady state grainsize is maintained.

Further deformation leads to a steady increase in the fraction of recrystallized material, until finally after about 50 % shortening the material is completely recrystallized. It is noteworthy that the recrystallized mass shows a spatially organized (domainal) crystallographic preferred orientation shown by the regions consisting of many grains with the same colour. Due to the small size of recrystallized grains (these are smaller than the sample’s thickness) details of this microstructure are not clear. The movie is presented in four different sections. These follow each other without a time lap, and show slightly different parts of the sample.

In some experiments, the heterogeneous distribution of the deformation resulted in the formation of shear zones. The initiation of these zones usually occurred in grains favourably oriented for slip: with the slip plane perpendicular to the glass plates and at high angles to the shortening direction. Such a grain constitutes a plane of weakness in a direction of high shear stress. In the experiment shown here, an easy slip grain is embedded in a large one in a hard orientation as shown by its colour. Stress concentration at the tip of this grain results in the initiation of a shear zone. In the conjugate direction, after some initial plastic deformation, another shearzone is initiated after a shortening of about 10 %. The initial rapid movement here could have been along an intragranular fracture. The shear zones rapidly recrystallize into a fine grained mass with a very good preferred orientation, one of easy glide. The deforming piston is seen to be moving in from the right side. This room temperature experiment was discussed in detail in fig. 2 of Urai and Humphreys 1981.

After shear zone initiation most of the deformation occurred by shear flow in the shear zones.

Mature shear zones consisted of fine grains (~10µm) due to the relatively high strain rate in the shear zone, with a strong preferred orientation. Often a darker contrast was observed along the centre.

In the second movie of this experiment the interaction between shear zones can be followed. Earlier shearzones become kinematically unfavourable for slip and new shear zones are initiated, passively carrying the earlier, now inactive shearzones along: these undergo metadynamic recrystallization in this stage. This process is strongly controlled by the geometry of the deformation which favours coaxial flow.

This movie is included to illustrate the effect of metadynamic recrystallization. The first part of the room temperature experiment is similar to the previous one: deformation is dominated by a long grain in soft orientation (A) which develops into a shearzone at high angle to the compression direction. Although deformation in this grain is strong, it recrystallizes very slowly, perhaps because of the low strain energy built up in grains oriented favourably for slip on one system.

After this deformation the loading piston was stopped and the sample allowed to statically recrystallize. The movie was recorded at a 10x slower rate, over a period of approximately 15 hours. The grain growth process strongly increases the size of the recrystallized grains and produces a good foam texture. It is noteworthy that some of the old grains which did not deform because they were located close to the soft grain A did not recrystallize either, presumably because they did not accumulate significant strain energy.

This movie shows a detail of one of the movies. It shows a nice example of the formation of kink boundaries in a grain (A) in easy glide orientation. With progressive deformation the grain is folded, and the kink boundaries migrate. After sufficient misorientation across the kink boundary, recrystallization starts with the growth of new grains in the kink boundary. The exact process of nucleation is unclear. This movie can be compared with fig. 2 of Urai et al., 1980.