Morphological properties of mineral aggregates are known to affect pavement and railroad track mechanistic behavior and performance significantly in relation to strength, modulus, and permanent deformation. With imaging technology, an objective and accurate measurement of aggregate particle size and shape properties can be obtained in a rapid, reliable, and automated fashion. But these advances in aggregate imaging must be brought to project sites and quarries in the field. This paper introduces field image-acquisition and image-processing techniques for extraction and analyses of size and shape properties of individual aggregate particles. Referred to as segmentation techniques, image-processing methods developed in this study analyzed two-dimensional field images of aggregates captured by a digital single-lens reflex camera. The segmented aggregate images were fed into the validated University of Illinois aggregate image analyzer to quantify particle size and shape properties by means of its image-processing algorithms for flat-and-elongated ratio, angularity index, and surface texture index. The developed method successfully determined the properties of coarse-aggregate samples collected from various depths in a layer of railroad track ballast. The promising preliminary results indicate that these segmentation techniques could be considered in the field for capturing several aggregate particles rapidly and reliably in a single image so that (a) size and shape properties of individual particles could be analyzed and (b) both spatial property variability and property changes with layer depth and usage (i.e., property degradation in time) could be evaluated under service loading.

Ballast fouling, often associated with deteriorating railroad track performance, refers to the condition in which the ballast layer changes its composition and develops a much finer grain size distribution. Fouling is commonly caused by degradation or breakage of ballast aggregates under traffic loading, although other fine materials including but not limited to coal dust, fine-grained subgrade soils, and sand can also contaminate a clean and uniformly graded ballast layer. An experimental approach is described to characterize stages of railroad ballast degradation studied through Los Angeles abrasion testing in the laboratory. An aggregate image analysis approach is used to investigate ballast particle abrasion and breakage trends at every stage through detailed quantifications of individual ballast particle size and shape properties. The experimental study indicated that the fouling index (FI) commonly used by practitioners was indeed a good indicator of fouling conditions, especially when all voids created by larger particles were filled by fine materials as FI values approached 40. Image analysis results of ballast particles larger than 9.5 mm (3/8 in.) scanned after a number of turns of the Los Angeles abrasion drum showed good correlations between percentage changes in aggregate shape properties, that is, imaging-based fatness and elongation, angularity and surface texture indexes, and the FI. The establishment of such relationships between in-service track fouling levels and ballast size and shape properties with similar field imaging techniques would help to understand field degradation trends better and as a result improve ballast serviceability and life-cycle performance.

The unbound aggregate ballast layer is a major structural and drainage component of railroad track that is known to degrade over time. Progressive degradation increases the fine-grained content of the ballast layer through particle breakage and abrasion or from external sources, such as subgrade or foreign material. The point at which ballast should be cleaned of these materials to avoid significant problems for drainage, track geometry, or ride quality is not well known. This paper attempts to ascertain the current state of the art on ballast permeability by reviewing previous studies, to fill any gaps by generating new laboratory test data, and to begin developing ballast cleaning considerations. A new and relatively simple test apparatus, the University of Illinois Constant Head Aggregate Permeameter, was used to study railroad ballast permeability as a function of degradation. Results of tests performed indicate that the cleaner the ballast, the more nonlinear the relationship between discharge velocity and hydraulic gradient, contrary to the findings of previous studies. In addition, flow decreased greatly after small increases in ballast degradation. Detailed findings related to the characteristics of flow—other than whether flow is impeded by in-service ballast condition—may not be extremely useful for rail practitioners because the amount of ballast degradation is difficult to determine in the field. However, the emerging ballast imaging technology described may be able to provide railroad personnel with a threshold for when ballast should be cleaned.

No abstract

Class I freight railroads nearly exclusively use premium, high-quality ballast aggregates such as granite, trap rock, and quartzite for their mainline ballast needs. These natural virgin aggregates are crushed stones with certain geologic origins, and their properties may vary significantly depending on quarry sources and production techniques. Findings are presented from a comprehensive laboratory study on ballast aggregates undertaken at the University of Illinois to investigate differences in engineering behavior of premium railroad ballast materials obtained from 13 quarries across the Union Pacific Railroad system. Laboratory research tasks consisted of gradation analysis; quantifications of shape, texture, and angularity properties with the University of Illinois aggregate image analyzer; shear strength evaluation by using large direct shear (shear box) tests; and ballast degradation and durability analyses with Los Angeles abrasion tests conducted at 400 and 1,000 turns and evaluations of subsequent ballast fouling. As the compacted ballast voids ratio increased, aggregate strength typically decreased because of lower solid densities obtained from shear box testing. Aggregate shape properties quantified through imaging-based shape indices were shown to be linked to the strength properties of different ballast density groups. One of the ballast aggregates with flat and elongated particles but very strong mineralogy had quite high strength properties despite the tendency for particles to break during shearing. Higher angularities that generally resulted in higher strength properties could also yield higher fouling and increased breakdown potential of the aggregates.

No abstract

Field and laboratory degradation trends of railroad ballast are investigated in this paper by using an aggregate image analysis approach. In addition to gradation changes, in-service ballast materials undergo degradation of particle shape properties (i.e., angularity, flatness and elongation, and surface texture), which can be quantified through imaging techniques. In the field, five ballast types were installed in eight steel boxes at a test section in the western heavy axle load revenue service test megasite near Ogallala, Nebraska. In-service performance evaluations consisted of sampling, sieving, and imaging-based shape property analyses of the field-collected ballast materials to monitor degradation twice a year. In the laboratory, following the Los Angeles (LA) abrasion test procedure, a sample of granite ballast was tested at increasing numbers of LA abrasion drum turns to assess changes in gradations and study degradation levels. All of the LA abrasion–tested granite samples and approximately 3,000 particles from the different sizes of the field-monitored ballast materials were collected and scanned with the Enhanced University of Illinois Aggregate Image Analyzer. Imaging-based particle shape indexes were determined at different degradation levels from both the LA abrasion and the field projects. The magnitudes and rates of measured degradation and breakage, as well as the captured changes in morphological properties of particles, were quite different between the LA abrasion test and the field degradation from repeated traffic loading. A comprehensive study with improved ballast sampling and testing techniques is recommended to further investigate mechanisms of ballast degradation.

No abstract