Proposed evaluation method for three-point bending beam tests of fiber reinforced concrete (2023)

Fiber reinforced concrete (FRC) is a composite material whose post-crack response is highly dependent on the properties of the mixed fibers, as well as their position and orientation in the matrix and on the fracture surface. To determine the material parameters of fiber reinforced concrete, a relatively small, notched beam with a cross-section of 125 × 150 mm is used in the European test standard. Although a uniform distribution is assumed, this is not the case in the small beam test. The coefficient of variation of the residual test results is usually high, which can be attributed to the relatively small size of the specimen and the random location of the fibers. This large scatter of the results leads to a low characteristic value during the traditional statistical evaluation, which results in an uneconomic design. Furthermore, the fibers were aligned to the longitudinal axis of the beam during production, which also led to an unrepresentative value.
When evaluating material parameters, ignoring the number, distribution, and location of fibers intersecting the fracture cross-section can lead to uneconomical, ineffectual, or even exaggerated material parameters. It is therefore necessary to modify and supplement the actual test standards.
In this paper, a mixing model is presented to determine the number of fibers intersecting the cross-section at a certain fiber geometry and dosage. The effect of mixing is demonstrated using measures of fiber-moment and uniformity. Different methods for the determination of the fiber-moment are presented, and the accuracy and sensitivity of these methods are also investigated. The values obtained are compared with the results of laboratory tests on steel and synthetic FRC. At the end of the experiment, the correlation between fiber-moment and residual strength is presented, along with the correlation coefficients.
A novel extended test method for the evaluation of beam test results using these methods is presented. The method allows more accurate mean, characteristic, and design material parameters to be determined.