A Systematic Method to Determine Optimal Specimen Geometry for Measuring Concrete Fracture Properties by the Peak-Load Method

The peak-load method is a simple method for determining the fracture properties $KIcS$ and CTOD_c of the two-parameter fracture model of concrete. With this method, the fracture properties can be determined from the measured peak loads of several different (or distinct) specimens. Specimens can be made distinct in specimen shape, size, notch length, and loading configuration. To quantitatively identify the degree of specimen distinction, a variable named specimen distinction number is proposed. When test specimens are considerably distinct, random experimental errors have negligible effect, and determination $KIcS$ and CTOD_c is of high accuracy. Thus test specimens can be optimized to obtain the most accurate values of the fracture properties. A previous experimental program found that change in specimen shape (e.g., a hole drilled at the axis of splitting tension cylinder) can largely expand the specimen distinction. An experimental program reported in this paper shows that changing loading configurations of beams (or prisms) from bending to eccentric compression can also largely enhance the specimen distinction and therefore provide $KIcS$ and CTOD_c determination of high accuracy.