Use of sampling techniques and reliability methods to assist in evaluation and repair of large-scale structures.

摘要

This dissertation evaluated the feasibility of using the theories of probability and statistics to assist in the data collection and management of nondestructive test results from large structures. More specifically, the research studied the applicability of using sampling methods to make predictions about damage levels and using reliability methods to evaluate repair options.; The first phase of the research involved applying sampling methods to actual structures. Case studies were used to evaluate the effectiveness of these techniques in making predictions about the entire state of the structure based on a limited number of tests. The first case study consisted of a post-tensioned bridge in which some of the tendons ducts were ungrouted. Sampling methods were used to determine the number and location of test points throughout the bridge. The information collected from these tests was then used to estimate the level of damage in the entire bridge within a given confidence, and these results were compared to the actual damage statistics. A variety of sampling methods were used in this case study including simple random sampling, stratified sampling, and adaptive sampling.; Once it was established that sampling methods may be useful in the nondestructive testing of structures, a case study with a larger structure was evaluated using similar techniques. This study involved a reinforced concrete retaining wall which was 7.5 miles in length. For this structure the methods which were employed included simple random, systematic, and adaptive sampling. As with the bridge, the sampling predictions were compared to the results from testing of the entire structure to assess the accuracy.; The retaining wall data was also used in the final stage of the research in which reliability methods were used to evaluate the repair options. The information from tests along the entire wall length were used to construct probabilistic models which revealed patterns in the damage. The various repair options were then incorporated into these models to determine their reliability. The results of these studies were stated in terms of the cost of repair versus the predicted cost of failure.; The research indicated that sampling techniques and reliability methods are useful in collecting and analyzing the data from the nondestructive tests on large structures. Certain sampling methods proved to be more effective than others at predicting the damage state of the structures in the case studies. Reliability methods provided a means to present the various repair options in terms of their failure probability and cost.