Optical Sizing and Roundness Determination of Glass Beads Used in Traffic Markings

Abstract

This work presents an interlaboratory study (ILS) culminating in a draft test method in AASHTO format for optical sizing and roundness determination of glass beads utilized in traffic markings. The ILS was conducted to determine the precision estimates for computerized optical testing of the glass beads. The ILS also included testing of the glass beads according to the traditional ASTM sieve analysis and roundness measurement methods, ASTM D 1214 and D 1155, respectively. Three replicates of three types of glass beads were prepared and sent to participating laboratories for size and roundness measurements. The test specimens were blended according to the gradations of Type 1, Type 3, and Type 5 glass beads specified in AASHTO M 247. In addition to gradation, the level of roundness of the glass beads, as judged by ASTM D 1214, was controlled in such a way that Type 1, Type 3 and Type 5 samples contained 70%, 80%, and 90% round particles, respectively. The statistical analysis of the ILS results indicated that the computerized optical methods provided significantly better accuracy and precision than the traditional methods for the size and roundness measurements of Type 3 and Type 5 samples (larger glass beads) but not for those of Type 1 samples (smaller glass beads). To provide a baseline evaluation of the size and roundness parameters used with the computerized optical equipment, samples similar to those prepared for the ILS were tested using X-ray computed microtomography (X-ray CT). The mathematical analysis of X-ray CT data indicated that the parameter Xcmin, the shortest chord out of the measured set of maximum chords, accurately measures the size of the glass beads relative to traditional sieve analysis. Among the roundness parameters, the ratio of Xcmin to Xfe max (longest Feret diameter) in 2- D and the ratio of T to L (thickness to length ratio) in 3-D best estimated the intended roundness of the glass beads. It was also found that for the roundness determination, a single cutoff value, which separates round from non-round using optical scanning data, would not work for all glass bead types. A separate cutoff value for each glass bead size class would be more appropriate for classifying the roundness of the glass beads. In addition, it was determined that the existing cutoff values were overestimating the intended roundness of the glass beads, which allowed some of the non-round particles to be considered round. Based on the analysis of X-ray CT data, more accurate cutoff values for the roundness parameters were determined.