Round-robin tests of the flexural toughness of fiber reinforced concrete were carried out using six different testing machines in five different laboratories. Six groups of beams, including a plain concrete control, two different volumes of polypropylene fibers, and three different volumes of steel fibers were tested in accordance with ASTM C 1018, with special care taken to exclude the "extraneous" deflections due to deformations at the specimen supports. The results from each laboratory were used to compute the ASTM C 1018 toughness indices I 5, I 10, I 20, I 30, and I 50 and the corresponding residual strength factors R 5,10, R 10,20, R 20,30, and R 30,50. In addition, the JSCE Toughness and Toughness Factor were also computed. It was found that, although the load vs. deflection curves were inherently quite variable, in most cases there was no significant difference among the participating laboratories, except for those mixes with a very low toughness. It was also found that the ASTM C 1018 toughness indices, particularly I 5 and I 10, did not discriminate very well between the different fiber contents or different fiber types; the JSCE parameters were rather more successful in this regard.

Procedures to obtain the experimental R-curves using a compliance calibration technique are revisited in this paper. R-curves provide a convenient means to study the process of fracture and the brittle-ductile transition in materials. Single edge notched beam specimens are tested under closed loop crack mouth opening control. The procedure to obtain the R-curves using loading/unloading compliance and the residual displacements are discussed. An elastically equivalent toughness K R as a function of crack extension is defined to compare the R-curves with the available data in the literature. The developed test method is applied to fiber reinforced concrete (FRC) composites with up to eight percent by volume of short, chopped alumina, carbon, and polypropylene (PP) fibers. Significant strengthening of the matrix due to the addition of short carbon and alumina fibers was observed. R-curves in these composites are characterized by an increase in the steady state fracture toughness. In PP-FRC composites, energy dissipation due to fiber pullout increases the ascending rate of the R-curve well after the main crack has formed. The work of fracture is computed from the cyclic loading/unloading tests and the results compared with the R-curves.

With the objective of understanding the reinforcing mechanisms of fibers in steel fiber reinforced concrete, the adherence between the fiber and the matrix was studied by conducting pullout tests of fibers from a cementitious matrix. In this paper, the effect of factors such as loading rate, inclination of fibers, and number of fibers have been investigated. An innovative measurement system was developed for high rates. It was experimentally obtained that by increasing the rate of loading, both pullout resistance and slip at peak were increased. Peak pullout force presents a higher rate sensitivity for a higher number of fibers. The lower the number of fibers, the higher the slip at peak rate sensitivity. Regardless of the number of fibers, a higher rate sensitivity for inclined fibers was observed.

The issue of how the method of determining midspan deflection in ASTM C 1018 toughness tests influences first-crack strength, first-crack deflection, toughness indices, and residual strength factors is addressed in this paper by comparing results obtained using the method now required in the current standard, which is based on net midspan deflection determined as the nominal midspan deflection minus the average of the deflections measured at the beam supports, with corresponding same specimen results based on nominal midspan deflection only which was not explicitly excluded in earlier versions of the standard. The problem of dealing with the portion of load-deflection relationship immediately after first crack when it is unstable is discussed. The range of test specimens for which comparative data are reported includes a series of third-point-loaded 500 x 150 x 150 mm beams with three different steel fibers ranging in length from 18 mm to 63 mm; a second, smaller series of 350 x 100 x 100 mm beams allowed for assessment of the effects of beam size and fiber alignment. Fiber contents varied from 20 to 75 kg/m 3 (0.25 to 0.94 percent by volume). Also included was a series of 350 x 100 x 100 mm beams with a single type of fibrillated polypropylene fiber of length 38 to 64 mm in amounts of 0.5 to 0.75 percent by volume. The results illustrate the extent to which the ASTM C 1018 parameters I 5, I 10, I 20, R 5,10, and R 10,20 are effective in distinguishing the performance of the various fiber reinforced concretes (FRC) mixtures in terms of fiber type, geometry, and amount. The index I 5 was found to be least effective. A case is made for greater emphasis on use of residual strength factors, especially R 10,20, when employing the test to specify and control the quality of FRC.

One of the possible factors inhibiting the wider application of fiber concretes is the fact that to characterize the engineering properties of fiber concrete, both cubes/cylinders and prisms have to be cast and tested, in addition to the determination of workability. This paper shows that with the use of superplasticizer, the slump test can be used to give guidance on the flowability characteristics of the fresh fiber concrete. The advantage of the slump test is that it is easy to carry it out in the field, apart from its simplicity and convenience. The paper further shows that the equivalent cube strength obtained from the broken pieces of a flexural test can adequately represent the compressive strength of fiber concretes. It is thus shown that it is possible to characterize the engineering properties of fiber concretes from only one set of prisms of about 100 x 100 x 500 mm size. Apart from first crack load, modulus of rupture, and fracture toughness properties, the prisms can be used to give additional information as appropriate, such as shrinkage and expansion, and through pulse velocity, internal microcracking. It is suggested that by rationalizing the approach to testing, it is possible to reduce not only the cost and inconvenience associated with different sizes of test specimens, but also to enhance the relevance of some of the information obtained from such testing.