Building better confidence intervals for the resistance factor in LRFD
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Engineers employ the LRFD model system in the planning and development of
structures (e.g. bridges, highways, buildings). The system takes into account the load
the structure will bear and the resistance of the structure against that load. Of interest
in the LRFD system is the resistance factor (φ) that affects the resistance. Samples are
collected to estimate φ, but the estimate (φ ̂ ) has typically been treated as a
deterministic number; when instead, it should be viewed as a random variable.
This thesis examines the performance of one method of proposed confidence
intervals for φ ̂ , investigates the sampling distribution of φ ̂, and proposes
improvements upon the existing method. A bootstrap simulation process is employed
to evaluate the performance of the proposed method (nonparametric and parametric)
through the metrics of coverage and relative interval length. Additionally, another
simulation is used to determine the distribution of the estimator (φ ̂). Finally, remedial
measures for the original interval method are proposed, implemented, and investigated
via simulations.