load signal introduced relatively high inaccuracies for the decreasing part of the load curve, the localisation of deformation. Image Processing Image processing techniques have been applied to derive geometric data from the recorded images. As the low strain rate experiments have been monitored using digital cameras this resulted in good image quality, facilitating the processing techniques. The digital image was converted into black-white and detection of the specimen contours appeared to be rather simple. A sub-pixel technique was applied to increase the accuracy of the contour positions. Fitting of a circle at a small region of the neck of the specimen yielded an approximation of the radius of curvature, also the diameter has been measured from the images. At high strain rates the quality of the scanned images was less compared with the digital images and more complicated techniques have been applied for detection of the specimen contours. The main technique consists of taking the second order derivative of the light intensity along a horizontal scan line perpendicular to the specimen axis. The specimen contour is located at a zero value of this derivative. Several smoothing techniques (moving average) were applied to reduce scatter in the derivatives and contour positions. EXPERIMENTAL RESULTS Flow Curves at Low Strain Rates Mateman [5] performed tensile tests at strain rates up to 6.2 .10-2 s-1 (engineering strain rate) using the low rate tensile test equipment. Specimens have been taken in the plate rolling direction, a diameter of 5 mm and gauge length of 35 mm has been chosen. The true failure strain according to Eqn. 1 is smaller for specimen perpendicular to the rolling direction compared to specimen in the rolling direction, respectively 1.0 and 1.46. No remarkable differences have been found in flow stress between the rolling direction and perpendicular to the rolling direction, except for a slightly increased initial yield stress. This means that only the rolling direction has to be examined with respect to flow behaviour. Figure 2 shows a comparison between the flow stress for two material directions taken from experiments. 0 350 700 1050 1400 0.00 0.50 1.00 1.50 Strain [-] Stress [MPa] Bridgman Siebel Bakker Bakker // true stress Bridgman // Siebel // Bakker true stress // yield stress yield stress // Figure 2. Comparison between flow-stress in parallel (//) and perpendicular (⊥) plate rolling directions
RkJQdWJsaXNoZXIy MjM0NDE=