Fractal Analysis of Track Geometry Data

Abstract

An ongoing FRA-sponsored research project at the University of Massachusetts has been exploring the use of fractal analysis of track geometry data for indicating track geometry condition, planning maintenance, and evaluating the cause of substructure-related problems. Results indicate that fractal analysis is able to provide unique numerical values (fractal dimensions) that characterize railway track geometry patterns, can discern different orders of roughness within track geometry data, and is effective for maintenance management by providing independent fractal parameters for trend analysis and geometry deterioration assessment. Fractal theory is discussed, and the usefulness of fractal analysis for quantifying railroad geometry data is demonstrated by highlighting key aspects of the research results. The relationship between track structure conditions and fractal dimensions for use in maintenance planning and condition evaluation is also discussed.