The results of strain measurements in this paper confirm the hypothesis of the propagation of Lüders bands across the gage length in the so called "yield point elongation." In this case, we find that the strain gages can accurately measure strains only up to the upper yield point in tensile test. The application of grid tab and image-handling software in strain measurement has been once again verified accurate and suitable for measuring the stress-strain curve in tensile test after the upper yield point. We also confirm that the accuracy of strains measured by using this method could be within±10%, if the measured strains are greater than 0.001 (0.1% or 1000με) in a tensile test without an upper yield point. Because the strain gages can measure only the small strains and the grid tabs can measure only the large strains, neither can stand along to be employed in the measurement of a whole stress-strain curve in tensile test. We first analyze the accuracy of the strains measured by using strain gages and those measured by using grid tabs and image-handling software, and propose the following complementary stress-strain curve measurements using a grid tab and a strain gage: (1) using the elastic strains measured by the strain gage up to the upper yield point, and followed by using the strains measured by the grid tab till the end of the test to construct a stress-strain curve when there is an upper yield point; and (2) using the strains measured by the strain gage up to the intersection point of the stress-strain curve by the 0.2% offset line, and followed by using the strains measured by the grid tab till the end of the test to construct a stress-strain curve when there is no upper yield point.
