Statistical Distribution of Fiber-Reinforced Concrete Beam Test Data

Abstract

Experience in post-crack performance testing of fiber-reinforced concrete (FRC) and shotcrete (FRS) using beams has demonstrated that variability both at first crack and in the post-crack range is high. When variability in performance is high it is important to carefully consider the within-set standard deviation in order to distinguish significant from insignificant differences between sets of data. It is also important to consider the characteristic distribution of results and use an appropriate probability distribution function (PDF) to estimate the standard deviation. The present investigation has sought to identify whether the Normal probability distribution function is the most appropriate model describing variability in cracking and post-crack performance parameters derived using beam tests.

Statistical Process Control of Fiber-Reinforced Concrete Precast Tunnel Segments

“…If the p-value was lower than or equal to 0.05, the given distribution function was rejected from the null hypothesis, indicating that it was not suitable for describing the data. Table 3 shows that the normal distribution is acceptable for the flexural strength 25 The highest p-value of a normal distribution was in flexural strength at L/150, accounting for 0.509, and the lowest was 0.176 in first crack strength. The Weibull distribution is only applicable for residual strength at L/150 (f D 150 ), while the lognormal distribution function was rejected for all groups.…”