Abstract
The best way to investigate the tension softening process is by applying a uniaxial tension force directly on a concrete specimen, because it can measure both tensile strength and the tension softening curve from an identical specimen. However, no standard tests have been adopted to provide a direct measurement of the tension softening curves of concrete, though many inadequate test procedures were proposed until recently. There are four misunderstandings in investigating the tension softening process, concerning the effects of notches, secondary flexure, boundary conditions and specimen geometry. In this paper, the misunderstandings are discussed and clarified in detail with theoretical and experimental considerations. A test procedure free from these misunderstandings is proposed for the uniaxial tension test of concrete and some successful test results are presented.
