Optimization of Long Range Major Rehabilitation of Airfield Pavements

The goal of this research has been to develop a methodology for managing pavement networks over prolonged analysis periods. Separate independent methods were devised for project and network level analysis, and the project level procedures were designed to provide inputs into the network level procedures. The project level analysis procedures optimally select and schedule major rehabilitation activities (routine maintenance, reconstruction, and overlays) over an extended analysis period. A computer code was written to use dynamic programming methods to optimally select schedule the activities (routine maintenance, reconstruction, and overlays) over the analysis period (20 years), by maximizing the structural performance {area under the utility weighted Pavement Condition Index (PCI) versus time curve}. These results can be used by decision makers and pavement engineers to help decide which rehabilitation activities should be considered and when to schedule them. The network level analysis procedures select, from the group of projects developed either with dynamic programming or by other methods of the engineers' choosing, those projects that maximized the sum of the user value weighted structural performance of each project. The mathematical representation of this selection process is a zero-one integer linear programming model. A heuristic developed by Toyoda provides extremely good solutions, resulting in a list of selected projects (each related to a specific feature) that maximizes the objective function with pre-established constraints (network funding limit, etc.). The network level analysis is run for several funding levels, and a series of management information reports are generated for each. With these reports, the consequences of selected network funding levels can quantitatively be compared. In addition, an estimate of an appropriate level of funding for the entire system can be made. A simple comparison of applying traditional manual techniques and those techniques developed in this research was made, and an application of the methods to an existing Air Force base was presented. The simple example shows a substantial difference between a manually developed network program and a program developed with the procedures developed in this research. The application to the existing Air Force base was done to show the feasibility of applying these techniques.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Civil Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2014

Author dc:creator

• Artman, David Henry, Jr.

• Engineering, Civil

Identifier

(UMI)AAI8309907

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/69920