On the Accuracy of Fracture Toughness Test Results for Concrete Using Different Size and Geometry Specimens and Data Reduction Methods

摘要

Much research has been performed on measuring the fracture toughness of concrete, but inconsistent toughness values in the literature leave some questions yet unanswered. This paper provides results of a broad-based experimental program designed to determine if certain tests produce an accurate measure of fracture toughness for concrete. The results of this study can be used to help make rational decisions when selecting a combination of specimen size, geometry and data reduction method to measure the fracture toughness of concrete. To be accurate, the fracture toughness value must be the same as would be obtained from an infinitely large test specimen. To show that a value of fracture toughness is accurate requires consistent values from tests using different size and geometry specimens and different data reduction methods. Therefore, this investigation uses three sizes of single edge, SE, and round double beam, RDB, specimens. More than one data reduction method was applied to the results of each size and geometry combination. Four different data reduction methods were used: linear elastic fracture mechanics, the two-parameter method, the size-effect method, and the Barker method. Results are presented from three batches of concrete, which represent two distinctively different mixes. The fracture toughness values obtained were not consistent within each batch; therefore, the most accurate value could not be shown conclusively. However, several significant conclusions were formed. The most common laboratory specimen size, no more than 310 mm deep/tall, is not sufficiently large to obtain an accurate measure of fracture toughness for concrete using either specimen geometry. Even the largest specimens, 1240 mm-tall RDB, experienced significant nonlinear fracture mechanics conditions for all of the concrete mixes. Combining the experimental results with numerical simulations could provide sufficient information to judge which of the fracture toughness values, if any, are close to the value that would be obtained from an infinitely large specimen.