Abstract
So far, the one bar method gave satisfactory stress-strain curves for "cylindrical" specimens, at the strain rate of 1x10^<-3>s^<-1>. On the other hand, for metallic "sheet" specimens only, the one bar method frequently detects an initial stress peak. The height of the peak is too high so that the extra height is apparent. One of the causes of this phenomenon is the deflection of the output bar end, as pointed out by Yoshida et al. in 2002. In order to get better alignment of the output bar, an output-bar supporting stand has been developed. Half-cut PTFE collars are installed in the stand. Via the collars, transverse compression to suppress the vibration can be applied to the output bar end. Comparing the quality of obtained dynamic tensile stress-strain curves, 50-190N is the suitable compressive load range. Even though the supporting stand is introduced, a small number of dynamic stress-strain curves still have the extraordinary stress peak. The additional cause of the peak is supposed that there are troublesome angles between the sheet specimen and the warp of the output bar. After measuring the angle between each specimen and the warp, thirteen dynamic tensile tests are carried out with variations of the angle and compressive load to the output bar. Probably, the angle is not the additional cause of the extraordinary stress peak.
