抄録
The Load Resistance Factor Design (LRFD) was introduced in Korea about a decade ago. However, the resistance factors suggested by AASHTO (2007, 2010) represent the characteristics of bedrocks in the US that may differ from those in Korea. There is also a general belief that the calculated probabilities of failure from conventional reliability analyses are not realistic because of the conservative bias used when predicting resistance and tails of probability distributions for load and resistance. The existence of a lower bound of resistance affects the reliability and resistance factors even though the lower bound of resistance is small. Thus, it is important to determine accurate resistance factors for drilled shafts based on reliable load test results and to calibrate the resistance factors considering lower-bound resistance. In this paper, an accurate resistance factor for drilled shafts was determined based on 13 sets of drilled shaft bi-directional load test data with strain gauges. Using a new log-normal distribution with lower bound of resistance, reliability analysis and resistance factor was performed using the advanced first order second moment (AFOSM) method. The target reliability indices were determined as 2.5, 3.0, and 3.5. For the target reliability index of 3.0 (AASHTO recommended value), the resistance factors were determined in the range of 0.13-0.32 for shaft resistance, 0.19-0.29 for base resistance. The lower bounds of resistance for each resistance (shaft and base) were determined, to be applied in the various bearing capacity equations, based on the Hoek-Brown failure criteria (2002) and the downgraded GSI values. Considering the lower-bound resistance, resistance factors were increased by up to 8% for the shaft resistance factor and 13% for the base resistance factor.
