Application of LADAR in the Analysis of Aggregate Characteristics (2012)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

20 This chapter first verifies the accuracy and repeatability of the FTI results by comparing the dimension results using FTI results by different operators to manual measurements of the particles with standard shapes, such as triangular prism, circu- late plate, and lens. After that, there are two sets of coarse aggre- gates, including seven types of aggregates in Aggregate Set 1 (hereinafter referred to as Set 1) and four types of aggregates in Aggregate Set 2 (hereinafter referred to as Set 2), and some fine aggregates imaged and analyzed using the FTI system. The FTI results of shape, angularity, and texture of various aggregates are presented and discussed. Detailed data for the shape, angularity, and texture of coarse aggregates can be found in Appendix E; images of the aggregates analyzed are presented in Appendix F. 4.1 Verification of FTI Results to Manual Measurements of Particles with Standard Shapes To verify the accuracy of the FTI system, particles of stan- dard shapes, such as triangular prism, circular plate, and lens, were analyzed using the FTI system by two different operators, and also manually measured using a Vernier caliper. Each par- ticle was imaged by the first operator twice and by the second operator three times. The accuracy and repeatability of FTI system and analysis were verified by comparing FTI analysis results to manual measurements. Figure 4-1 shows the images of a triangular prism particle acquired with the FTI system. In the plan view color map, dark areas at the center represent larger z-values, while darker areas at both ends indicate low points on the prism surface. The object’s 2-D boundary is shown as a white line in Figure 4-1(c). Table 4-1 tabulates its dimensions by manual mea- surements and FTI analyses, and the differences between the FTI and manual measurements. All parameters are shown in Figure 4-1(d). The triangular prism has a height of 7.06 mm and a maximum length of 13.91 mm along the L4 direction with a vertex angle a of 91∞. In general, FTI analysis results are consistent with manual measurements with very small deviations, which verifies the accuracy of the FTI system. FTI results of length are more accurate than those of height. Deviation of height is 8.42%, while the maximum deviation of length is 2.56%. The calculated vertex angle a′ is 86° based on FTI analysis results, with a deviation of 5°. Figure 4-2 shows a circular plate particle imaged in the FTI system. In the plan view color map, the circled dark area repre- sents larger z-values, and the object’s 2-D boundary is shown as a white line in isometric shaded surface profile in Figure 4-2(c). Table 4-2 presents dimensions of the circular plate by manual measurement and FTI analysis, and deviations of FTI results from manual measurements. FTI analysis results show good agreement with manual measurements. The plate has a height of 6.01 mm with a diameter of 24.97 mm. Deviations of height and diameter are 1.95% and 0.12%, respectively. The repeatability of the FTI system is validated by the good consis- tency between FTI results operated by different operators and the very close FTI results of remeasuring the same particle. Figure 4-3 shows a lens imaged in the FTI system. In the plan view color map, the white area represents larger z-values, and the object’s 2-D boundary is shown as a white line in isometric shaded surface profile in Figure 4-3(c). Table 4-3 shows dimen- sions of the lens by manual measurement and FTI analysis, and deviations. Deviations of height and diameter are 5.12% and 1.47%, respectively. There is a larger deviation in diameter than that of the circular plate due to a smaller diameter (11.97 mm). Figure 4-4 shows the relationship between manual measure- ments and FTI analysis results. The manual measurement x is the average value of physical measurements of the parameters in all aggregates of standard shapes. The FTI analysis results of the corresponding parameters are labeled as y. As shown in this figure, the relationship between x and y can be regressed by a linear function. FTI analysis results show good agreement with the manual measurements as the correlation coefficient R2 is 0.9957. Possible reasons for the deviations are windowing out of error areas along the surface boundary during surface C h a p t e r 4 FTI Results

21 reconstruction and computation errors during MATLAB program analysis of data acquired from the FTI system. 4.2 FTI Results of Coarse Aggregates in Set 1 In Set 1, seven types of coarse aggregate with sizes ranging from ¾ in. to #4 are imaged and analyzed. The seven types of aggregate are blast furnace slag (BFS), copper ore, dolomite, crushed glacial gravel, rounded glacial gravel, iron ore, and limestone. For each type of aggregate, the four size ranges are passing a 1-in. sieve and retaining on a ¾-in. sieve (hereinaf- ter referred to as ¾ in.), passing a ¾-in. sieve and retaining on a ½-in. sieve (hereinafter referred to as ½ in.), passing a ½-in. sieve and retaining on a 3⁄8-in. sieve (hereinafter referred to as 3⁄8 in.), and passing a 3⁄8-in. sieve and retaining on a #4 sieve (hereinafter referred to as #4). Table 4-4 presents the origins and physical properties of aggregates in Set 1. Table 4-5 pre- sents the total number of aggregates imaged in the FTI system. These aggregates were selected for the purpose of evaluating whether the FTI system is able to detect some special aggre- gates due to the aggregate colors and the surface porosity. Asymptotic analysis is performed to determine the required sample size for aggregate evaluations, at which constant mor- phological results can be achieved. Figure 4-5 through Fig- ure 4-8 present the mean value sample size relationship for the shape factor (i.e., sphericity, flatness ratio, and elonga- tion ratio), angularity, and texture for dolomite aggregates of sizes of ¾ in., ½ in., 3⁄8 in., and #4, respectively. As shown in these figures, 30 aggregate particles are sufficient for achiev- ing stable and constant values for each morphological char- acteristic. Asymptotic analysis results of the other types of aggregates also lead to the same sample size requirement of 30 particles. Therefore, with the consideration of statistical sta- bility, 30 aggregate particles are analyzed within each aggre- gate size range for every type of coarse aggregate. Table 4-6 shows the average values of the five morphologi- cal parameters for the seven types of aggregate in Set 1. The (a) Visual image (b) Plan view color map (c) Isometric shaded surface profile (d) Schematic configuration Height L3 L4 L1 L2α x (pixel) x (mm) y (mm) y (pixel) Height z (mm) Figure 4-1. Triangular prism particle, imaged in the FTI system.

22 morphological parameters presented are sphericity, flatness ratio, elongation ratio, angularity, and texture. Each mean value of the morphological parameter in this table represents the mean value of the 30 aggregate particles of each type of aggregate within a specific size range. Further illustrations are presented in this section with figures. 4.2.1 Shape Figure 4-9 presents the sphericity distribution of ¾-in. aggregate particles in Set 1. As we can see from the figure, both crushed glacial gravel and blast furnace slag have greater values of sphericity, whereas dolomite and iron ore have smaller sphe- ricity among the seven types of aggregate. Sphericity distribu- tions of rounded glacial gravel and limestone are very close. Figure 4-10 shows the sphericity distribution of ½-in. aggregates of the seven types of aggregate in Set 1. Compared with ¾-in. aggregates, crushed glacial gravel still remains the largest in sphericity of the seven types of aggregate; iron ore has the smallest sphericity. Figure 4-11 presents the sphericity distribution of 3⁄8-in. aggregates in Set 1. Both crushed glacial gravel and rounded glacial gravel aggregates have large sphericity values. Con- versely, iron ore has the smallest sphericity values. Figure 4-12 presents the sphericity distribution of #4 aggregates in Set 1. Both crushed glacial gravel and blast fur- nace slag have relatively larger values of sphericity; iron ore has smaller values of sphericity. Unlike the large dolomite aggregates (¾ in., ½ in., and 3⁄8 in.), dolomite #4 aggregates have the smallest sphericity values. Figure 4-13 shows the distribution of the flatness ratio of ¾-in. aggregates in Set 1. Dolomite and rounded glacial gravel have smaller values of flatness ratios, whereas iron ore has the largest values of flatness ratio. Figure 4-14 plots the distribution of the flatness ratio of ½-in. aggregates in Set 1. Aggregates of crushed glacial gravel and limestone have very close flatness ratio values. Different from the ¾-in. aggregates, dolomite has the greatest flatness ratio values, and iron ore has the smallest flatness ratio values. Figure 4-15 presents the flatness ratio distribution of 3⁄8-in. aggregates in Set 1. In this figure, crushed glacial gravel and rounded glacial gravel have similar distributions, with large values of flatness ratio. The flatness ratios of iron ore and copper ore are smaller than those of other aggregates. Blast furnace slag, dolomite, and limestone have flatness ratio dis- tributions very close to each other. Figure 4-16 presents the flatness ratio distribution of #4 aggregates in Set 1. Similar to 3⁄8-in. aggregates, both Triangular Prism FTI Analysis Manual Measurements Deviation Parameter AnalysisResults Average Measured Results Average (%) Height 6.5455 6.4376 6.4469 6.4476 6.4510 6.4657 7.06 7.06 7.05 7.06 8.42 L1 13.7895 13.8250 13.9313 13.9313 13.9313 13.8817 13.91 13.91 13.91 13.91 0.20 L2 9.7513 9.8011 9.7438 9.7815 9.7734 9.7702 9.92 9.93 9.92 9.92 1.51 L3 9.7952 9.7507 9.8015 9.7832 9.8034 9.7868 9.91 9.92 9.92 9.92 1.34 L4 9.8954 9.9018 9.8201 9.7756 9.8633 9.8512 10.12 10.10 10.11 10.11 2.56 Table 4-1. Parameters of triangular prism by FTI analysis and manual measurement (unit: mm).

23 Figure 4-2. Circular plate particle, imaged in the FTI system. (a) Visual image (b) Plan view color map (c) Isometric shaded surface profile (d) Schematic configuration Height Diameter y (mm) y (pixel) x (mm) x (pixel) Height z (mm) Circular Plate FTI Analysis Manual Measurements Deviation Parameter AnalysisResults Average Measured Results Average (%) Height 5.9030 5.8730 5.9167 5.8846 5.8946 6.01 6.01 6.01 6.01 1.91 5.8955 Diameter 24.7010 24.6443 24.6443 24.6797 24.6797 24.6698 24.97 24.97 24.97 24.97 0.12 Table 4-2. Parameters of circular plate by FTI analysis and manual measurement (unit: mm). crushed glacial gravel and rounded glacial gravel have large flatness ratio values, copper ore and dolomite have very close distributions, and iron ore has the smallest values of flatness ratio. Figure 4-17 plots the elongation ratio distribution of ¾-in. aggregates in Set 1. Of the seven types of aggregate, crushed glacial gravel has the largest elongation ratios, and iron ore has the smallest elongation ratios. All the other aggregates have distributions of elongation ratios that are very close to each other. Figure 4-18 shows the elongation ratio distribution of ½-in. aggregates in Set 1. In this figure, limestone has the (continued on page 27)

24 (a) Visual image (b) Plan view color map (c) Isometric shaded surface profile (d) Schematic configuration Height Diameter x (mm) x (pixel) y (pixel) y (mm) Height z (mm) Figure 4-3. Lens, imaged in the FTI system. Lens FTI Analysis Manual Measurements Deviation Parameter Analysis Results Average Measured Results Average (%) Height 5.4978 5.5033 5.5123 5.5031 5.4877 5.5031 5.80 5.80 5.81 5.80 5.12 Diameter 11.8012 11.7876 11.7795 11.7937 11.8052 11.7942 11.97 11.96 11.97 11.97 1.47 Table 4-3. Parameters of lens by FTI analysis and manual measurement (unit: mm).

25 Aggregate Type Origin Lithographic Description LAA Loss (%) Bulk Dry SpGr (g/cm3) 24-hr Soak Absorption (%) Blast furnace slag Wayne, MI Gray to black vesicular with poorly developed crystalline structure 43 2.27 3.18 Copper ore Keweenaw, MI Mixture of white, green, pink, and burgundy colors, fine to coarse grains 19 2.64 0.95 Houghton, MI 16 2.76 2.12 Dolomite Mackinac, MI Light tan to light gray, medium to coarse crystals 27 2.78 0.52 Monroe, MI Tan, well-defined, small euhedral dolomite crystals 45 2.45 4.16 Glacial gravel – crushed Kent, MI Various colors 17 2.73 0.71 Glacial gravel – rounded Kent, MI Various colors 19 2.68 1.10 Iron ore Marquette, MI Black with hints of brown, very fine grained and hard metamorphic rock 11 — — Limestone Schoolcraft, MI Light tan with very fine subcrystalline texture 25 2.65 0.64 Arenac, MI Light gray, very fine crystalline with abundant frosted quartz sand grains 42 2.56 2.13 Note: LAA = Los Angeles abrasion; Bulk dry SpGr = Bulk dry specific gravity. Table 4-4. Types and the physical properties of aggregates in Set 1. 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Triagular prism Height L1 L2 L3 L4 Circular plate Height Diameter Lens Height DiameterFT I a na ly sis re su lts y (m m) Manual measurement x (mm) y = 1.0833x-0.5473 R2= 0.9957 Figure 4-4. Comparison of dimensions by manual measurement to FTI analysis results for particles with standard shapes.

26 0.6 0.7 0.8 0.9 1.0 0 5 10 15 20 25 30 35 40 1E-7 1E-5 1E-3 Sphericity Flatness ratio Elongation ratio M ea n sh ap e fa ct or Angularity Texture M ea n an gu la rit y & te xt ur e Number of 3/4'' aggregate particles evaluated Figure 4-5. Asymptotic analysis to determine required sample size for ¾-in. dolomite aggregates. 1E-7 1E-5 1E-3 Sphericity Flatness ratio Elongation ratio Angularity Texture 0.4 0.6 0.8 1.0 0 5 10 15 20 25 30 35 40 M ea n sh ap e fa ct or M ea n an gu la rit y & te xt ur e Number of 1/2'' aggregate particles evaluated Figure 4-6. Asymptotic analysis to determine required sample size for ½-in. dolomite aggregates. Aggregate Retaining Sieve Size 3/4” 1/2” 3/8” #4 Aggregate Set 1 Blast furnace slag 30 30 30 30 Copper ore 30 30 30 30 Dolomite 30 30 30 30 Glacial gravel – crushed 30 30 30 30 Glacial gravel – rounded 30 30 30 30 Iron ore 30 30 30 30 Limestone 30 30 30 30 Total number of aggregates 840 Table 4-5. Number of coarse aggregate particles imaged and analyzed in Set 1.

27 Sphericity Flatness ratio Elongation ratio Angularity Texture 0.6 0.7 0.8 0.9 1.0 0 5 10 15 20 25 30 35 40 1E-7 1E-5 1E-3 M ea n sh ap e fa ct or M ea n an gu la rit y & te xt ur e Number of 3/8'' aggregate particles evaluated Figure 4-7. Asymptotic analysis to determine required sample size for 3⁄8-in. dolomite aggregates. smallest values of elongation ratio, with very close distribu- tions of all other types of aggregates. Figure 4-19 is the elongation ratio distribution of 3⁄8-in. aggregates in Set 1. In this figure, iron ore has much smaller elongation ratios than all the other types of aggregate. Both crushed glacial gravel and rounded glacial gravel have greater values of elongation ratio, and the elongation ratio distribu- tions of blast furnace slag and copper ore are very close. Figure 4-20 depicts the elongation ratio distribution of #4 aggregates. Blast furnace slag aggregates are the most elongated; dolomite aggregates are the least elongated. The elongation ratio distributions of all the other aggregates are very close. 4.2.2 Angularity Figure 4-21 plots the angularity distribution of all the ¾-in. aggregates in Set 1. It was found that crushed glacial gravel aggregates have the largest angularity values. Rounded glacial gravel and iron ore have smaller values of angularity compared to the other aggregates. Sphericity Flatness ratio Elongation ratio Angularity Texture 1E-7 1E-5 1E-3 0.4 0.6 0.8 1.0 0 5 10 15 20 25 30 35 40 M ea n sh ap e fa ct or M ea n an gu la rit y & te xt ur e Number of #4 aggregate particles evaluated Figure 4-8. Asymptotic analysis to determine required sample size for #4 dolomite aggregates.

28 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 3/4'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Sphericity Figure 4-9. FTI sphericity distribution of ¾-in. aggregates in Set 1. Aggregate Size Sphericity Flatness Ratio Elongation Ratio Angularity Texture 3/4” Aggregates Blast furnace slag 0.72 0.72 0.73 2.74E-04 6.38E-06 Copper ore 0.69 0.66 0.72 2.69E-04 4.97E-06 Dolomite 0.66 0.62 0.71 1.58E-04 3.90E-06 Glacial gravel – crushed 0.74 0.71 0.77 2.31E-04 3.86E-06 Glacial gravel – rounded 0.68 0.59 0.75 8.44E-05 1.60E-06 Iron ore 0.67 0.73 0.65 1.01E-04 2.13E-06 Limestone 0.69 0.71 0.70 9.85E-05 2.52E-06 1/2” Aggregates Blast furnace slag 0.72 0.76 0.72 2.44E-04 9.57E-06 Copper ore 0.69 0.69 0.71 9.76E-05 4.69E-06 Dolomite 0.70 0.80 0.68 5.99E-05 2.86E-06 Glacial gravel – crushed 0.74 0.77 0.74 1.46E-04 2.70E-06 Glacial gravel – rounded 0.70 0.71 0.71 9.71E-05 3.54E-06 Iron ore 0.67 0.68 0.68 9.93E-05 5.17E-06 Limestone 0.66 0.78 0.62 1.90E-04 4.60E-06 3/8” Aggregates Blast furnace slag 0.73 0.73 0.75 2.19E-04 9.48E-06 Copper ore 0.70 0.72 0.71 1.39E-04 4.39E-06 Dolomite 0.74 0.75 0.75 8.74E-05 3.59E-06 Glacial gravel – crushed 0.79 0.80 0.80 1.47E-04 6.94E-06 Glacial gravel – rounded 0.79 0.78 0.80 1.09E-04 4.81E-06 Iron ore 0.57 0.64 0.57 1.09E-03 7.75E-05 Limestone 0.69 0.72 0.68 3.32E-04 1.59E-05 #4 Aggregates Blast furnace slag 0.70 0.66 0.73 1.47E-03 6.00E-05 Copper ore 0.60 0.59 0.64 4.25E-03 1.90E-04 Dolomite 0.56 0.55 0.58 1.57E-03 8.23E-05 Glacial gravel – crushed 0.69 0.71 0.69 2.48E-04 9.33E-06 Glacial gravel – rounded 0.73 0.77 0.73 2.48E-04 1.44E-05 Iron ore 0.58 0.53 0.63 2.16E-03 7.42E-05 Limestone 0.65 0.63 0.67 1.25E-03 9.18E-05 Table 4-6. Average values of morphological parameters of aggregates in Set 1.

29 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 1/2'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Sphericity Figure 4-10. FTI sphericity distribution of ½-in. aggregates in Set 1. 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 3/8'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Sphericity Figure 4-11. FTI sphericity distribution of 3⁄8-in. aggregates in Set 1. 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 #4 Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Sphericity Figure 4-12. FTI sphericity distribution of #4 aggregates in Set 1.

30 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Flatness ratio 3/4'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-13. FTI flatness ratio distribution of ¾-in. aggregates in Set 1. 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Flatness ratio 1/2'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-14. FTI flatness ratio distribution of ½-in. aggregates in Set 1. 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Flatness ratio 3/8'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-15. FTI flatness ratio distribution of 3⁄8-in. aggregates in Set 1. 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Flatness ratio #4 Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-16. FTI flatness ratio distribution of #4 aggregates in Set 1.

31 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Elongation ratio 3/4'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-17. FTI elongation ratio distribution of ¾-in. aggregates in Set 1. 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Elongation ratio 1/2'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-18. FTI elongation ratio distribution of ½-in. aggregates in Set 1. 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Elongation ratio 3/8'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-19. FTI elongation ratio distribution of 3⁄8-in. aggregates in Set 1. 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 C um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Elongation ratio #4 Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore Limestone Figure 4-20. FTI elongation ratio distribution of #4 aggregates in Set 1.

32 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 1/2'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Angularity Figure 4-22. FTI angularity distribution of ½-in. aggregates in Set 1. Figure 4-22 depicts the angularity distribution of all the ½-in. aggregates in Set 1. All the aggregates have close distri- bution of angularity except dolomite aggregates, which have much smaller values. Figure 4-23 shows the angularity distribution of all the 3⁄8-in. aggregates. Copper ore has the smallest angularity val- ues; iron ore and limestone are the most angular aggregates of the seven types of aggregate. Rounded glacial gravel and dolomite have angularity distributions that are very close to each other. Figure 4-24 presents the angularity distribution of all the #4 aggregates. Iron ore remains as the most angular aggregate. Unlike 3⁄8-in. aggregates, both copper ore and iron ore aggre- gates are very angular, whereas blast furnace slag aggregates are much less angular. From the angularity distribution fig- ures, it can be inferred that angularity rankings of aggregates may be size dependent. 4.2.3 Texture Figure 4-25 presents the texture distribution of ¾-in. aggregates in Set 1. As expected, rounded glacial gravel aggre- gates have the smoothest surfaces, whereas blast furnace slag aggregates have the roughest surfaces. Copper ore and dolo- mite have similar values of surface texture; iron ore and lime- stone have very close surface texture values. Figure 4-26 plots the texture distribution of ½-in. aggre- gates in Set 1. As shown in this figure, blast furnace slag 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 3/4'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Angularity Figure 4-21. FTI angularity distribution of ¾-in. aggregates in Set 1.

33 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 3/8'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Angularity Figure 4-23. FTI angularity distribution of 3⁄8-in. aggregates in Set 1. 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 #4 Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Angularity Figure 4-24. FTI angularity distribution of #4 aggregates in Set 1. 1E-7 1E-61E-8 1E-5 1E-4 0 20 40 60 80 100 3/4'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Texture Figure 4-25. FTI texture distribution of ¾-in. aggregates in Set 1.

34 1E-7 1E-61E-8 1E-5 1E-4 0 20 40 60 80 100 3/8'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Texture Figure 4-27. FTI texture distribution of 3⁄8-in. aggregates in Set 1. 1E-7 1E-61E-8 1E-5 1E-4 0 20 40 60 80 100 1/2'' Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Texture Figure 4-26. FTI texture distribution of ½-in. aggregates in Set 1. aggregates are the roughest, followed by iron ore. Copper ore and limestone have very close values of surface texture. Dolomite, crushed glacial gravel, and rounded glacial gravel have very similar surface textures, all of which are smoother than the other aggregates. Figure 4-27 shows the texture distribution of 3⁄8-in. aggre- gates. Iron ore aggregates are the roughest, followed by limestone and blast furnace slag aggregates; copper ore and rounded glacial gravel aggregates have the smoothest surface textures. Figure 4-28 depicts the texture distribution of #4 aggregates. Both crushed glacial gravel and rounded glacial gravel have very smooth surface textures. However, different from large aggre- gates (¾ in., ½ in., and 3⁄8 in.), copper ore #4 aggregates have the roughest surface texture, followed by limestone, dolomite, and blast furnace slag. Table 4-7 shows the relative rankings of sphericity, flat- ness ratio, elongation ratio, angularity, and texture for coarse aggregates in Set 1 based on the average values of each morphological parameter shown in Table 4-6. The ranking indicates that rankings of morphological properties are size dependent. 4.3 FTI Results of Coarse Aggregates in Set 2 In Set 2, four types of aggregate (Broadway, Maymead, Salem, and Strasburg) were analyzed. There are five aggre- gate particles passing the ¾-in. sieve and retaining on the ½-in. sieve for each type of the four aggregates that were imaged and analyzed, and there are 20 particles in total imaged and analyzed. The 20 aggregate particles were

35 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 #4 Aggregates Blast Furnace Slag Copper Ore Dolomite Glacial Gravel Crushed Glacial Gravel Rounded Iron Ore LimestoneC um ul at iv e pe rc en ta ge o f ag gr eg at es ( % ) FTI Texture Figure 4-28. FTI texture distribution of #4 aggregates in Set 1. scanned using the FTI system prior to the Micro-Deval tests. The FTI results of the original aggregates were referred to as “0-min. Micro-Deval FTI results.” Then the 20 aggregate particles were subjected to the Micro-Deval test for 15 min., and scanned again using the FTI system. The FTI results (shape, angularity, and texture) of the 20 aggregate particles after 15 min. of Micro-Deval testing were referred to as the “15-min. Micro-Deval FTI results.” The 20 aggregate par- ticles were subjected to the Micro-Deval test for another 30 min., with a total of 45 min. of Micro-Deval testing. At the end of 45 min., the aggregate particles were imaged and ana- lyzed using the FTI system again. The results will be herein- after referred to as “45-min. Micro-Deval FTI results.” The three sets of FTI results of the 20 aggregates were compared Aggregate Size Ranking 1 2 3 4 5 6 7 3/4” Sphericity GGC BFS CO LST GGR IO DLT Flatness ratio IO BFS GGC LST CO DLT GGR Elongation ratio GGC GGR BFS CO DLT LST IO Angularity BFS CO GGC DLT IO LST GGR Texture BFS CO DLT GGC LST IO GGR 1/2” Sphericity GGC BFS DLT GGR CO IO LST Flatness ratio DLT LST GGC BFS GGR CO IO Elongation ratio GGC BFS CO GGR DLT IO LST Angularity BFS LST GGC IO CO GGR DLT Texture BFS IO CO LST GGR DLT GGC 3/8” Sphericity GGC GGR DLT BFS CO LST IO Flatness ratio GGC GGR DLT BFS CO LST IO Elongation ratio GGC GGR BFS DLT CO LST IO Angularity IO LST BFS CO GGC GGR DLT Texture IO LST BFS GGC GGR CO DLT #4 Sphericity GGR BFS GGC LST CO IO DLT Flatness ratio GGR GGC BFS LST CO DLT IO Elongation ratio BFS GGR GGC LST CO IO DLT Angularity CO IO DLT BFS LST GGC GGR Texture CO LST DLT IO BFS GGR GGC Note: BFS = Blast furnace slag; CO = Copper ore; DLT = Dolomite; GGC = Glacial gravel – crushed; GGR = Glacial gravel – rounded; IO = Iron ore; LST = Limestone. Table 4-7. Ranking summary of shape, angularity, and texture for coarse aggregates in Set 1.

36 Aggregate Source Aggregate Description Aggregate Size Maymead Granite Non-carbonate 3/4”–1/2” Salem Quartzite Broadway Dolomite Carbonate Strasburg Limestone Table 4-8. Types of aggregates in Set 2. with each other to detect the changes of both angularity and texture due to different durations of the Micro-Deval test. Table 4-8 shows the aggregate types in Set 2. The tests and results presented in this section are to assess whether the FTI system is sensitive to detecting the angularity and texture changes due to the abrasion in the Micro-Deval test. The same particles are imaged before and after Micro- Deval testing of different durations. We assume that shape will not change significantly during Micro-Deval testing; there- fore, only angularity and texture are analyzed in this section. Table 4-9 shows the average values of angularity and texture for each type of aggregate in Set 2. Regardless of the duration of the Micro-Deval test, Broadway and Strasburg have relatively smaller angularity, and Maymead has the largest angularity of the four types of aggregate. An interesting observation is that the Micro-Deval test does not necessarily help to decrease the angularity of aggregates. Among the four types of aggregate, the angularities of Strasburg aggregates decreased from 1.02 × 10-4 for the unpolished to 9.26 × 10-5 after 15 min. of the Micro- Deval test, and finally to 9.73 × 10-5 after 45 min. of Micro- Deval. As the images were acquired from the same locations of the same aggregate particles, no statistical analysis is needed. Figure 4-29 plots the mean value of FTI angularity and texture for Broadway in the 65% confidence interval of FTI angularity and texture for dolomite ½-in. aggregates in Set 1. Figure 4-30 plots the mean value of FTI angularity and texture for Strasburg in the 65% confidence interval of FTI angularity and texture for limestone ½-in. aggregates in Set 1. Mean val- ues of FTI angularity and texture for Broadway and Strasburg aggregates show good correlations with the FTI angularity and Micro-Deval Duration 0 Min. 15 Min. 45 Min. Morphological parameters Angularity Texture Angularity Texture Angularity Texture Broadway 8.64E-05 3.30E-06 8.94E-05 3.62E-06 1.27E-04 1.05E-05 Maymead 1.58E-04 8.01E-06 1.17E-04 9.38E-06 2.86E-04 1.95E-05 Salem 1.22E-04 5.66E-06 1.03E-04 6.50E-06 2.06E-04 1.50E-05 Strasburg 1.02E-04 5.01E-06 9.26E-05 6.56E-06 9.73E-05 3.56E-06 Table 4-9. Average values of angularity and texture of aggregate in Set 2. Figure 4-29. The mean value of FTI angularity and texture of Broadway in the 65% confidence interval of FTI angularity and texture for dolomite ½-in. aggregates in Set 1.

37 Figure 4-30. The mean value of FTI angularity and texture of Strasburg in the 65% confidence interval of FTI angularity and texture for limestone ½-in. aggregates in Set 1. 0 min of Micro-Deval Broadway Maymead Salem Strasburg 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 Cu m ul at iv e pe rc en ta ge o f a gg re ga te s ( %) FTI Angularity Figure 4-31. FTI angularity distribution before test. 15 min of Micro-Deval Broadway Maymead Salem Strasburg 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 Cu m ul at iv e pe rc en ta ge o f a gg re ga te s ( %) FTI Angularity Figure 4-32. FTI angularity distribution after 15 min. of the Micro-Deval test. texture of dolomite and limestone ½-in. aggregates in Set 1, as the majority of mean values are reasonably distributed within the 65% confidence interval of FTI angularity and texture for dolomite and limestone ½-in. aggregates. 4.3.1 Graphical Presentation of Angularity Due to the relatively small values of angularity and texture, it is hard to appreciate the sense of these values in the absolute scale. This subsection presents these values graphically so that one can make a better judgment about the relative ranking of these aggregates. Figure 4-31 plots the angularity distribution of the aggre- gates before they are subjected to Micro-Deval testing, denoted as “after 0 min. of Micro-Deval.” It can be seen that Broadway aggregates are the most angular; Salem and Strasburg have smaller angularity values. Figure 4-32 shows the angularity distribution of the aggre- gates after being subjected to the Micro-Deval test for 15 min. Broadway and Strasburg have similar angularity distributions and smaller values, whereas Maymead and Salem have similar distributions with greater values. Figure 4-33 shows the angularity distribution of aggre- gates after being subjected to the Micro-Deval test for 45 min. Both Maymead and Strasburg have larger angular- ity values. The FTI angularity ranking is consistent with the AIMS II result for aggregates after being subjected to 45 min. of the Micro-Deval test, shown in Appendix F. By compar-

38 45 min of Micro-Deval Broadway Maymead Salem Strasburg 1E-7 1E-6 1E-5 1E-31E-4 0.01 0 20 40 60 80 100 Cu m ul at iv e pe rc en ta ge o f a gg re ga te s ( %) FTI Angularity Figure 4-33. FTI angularity distribution of aggregates after 45 min. of the Micro-Deval test. 0 min of Micro-Deval Broadway Maymead Salem Strasburg 1E-7 1E-6 1E-5 0 20 40 60 80 100 Cu m ul at iv e pe rc en ta ge o f a gg re ga te s ( %) FTI Texture Figure 4-34. FTI texture distribution before test. 15 min of Micro-Deval Broadway Maymead Salem Strasburg 1E-7 1E-6 1E-5 0 20 40 60 80 100 Cu m ul at iv e pe rc en ta ge o f a gg re ga te s ( %) FTI Texture Figure 4-35. FTI texture distribution after 15 min. of the Micro-Deval test. 45 min of Micro-Deval Broadway Maymead Salem Strasburg 1E-7 1E-6 1E-5 0.0 0.2 0.4 0.6 0.8 1.0 Cu m ul at iv e pe rc en ta ge o f a gg re ga te s ( %) FTI Texture Figure 4-36. FTI texture distribution after 45 min. of the Micro-Deval test. ing the angularity distributions in Figure 4-31, Figure 4-32, and Figure 4-33, it can be seen that aggregates having larger angularity tend to remain relatively more angular after the Micro-Deval test. 4.3.2 Graphical Representation of Texture Figure 4-34, Figure 4-35, and Figure 4-36 plot the texture distributions for aggregates in Set 2 after Micro-Deval testing for 0 min., 15 min., and 45 min., respectively. Of the four types of aggregates, Maymead has the largest texture val- ues, whereas Broadway and Strasburg have smaller texture values, regardless of the duration of the Micro-Deval test. 4.4 FTI Analysis of Fine Aggregates Images in Figure 4-37 through Figure 4-39 show that the FTI system is capable of capturing high-resolution images of fine aggregates. Image processing and morphological analy- sis can also be performed on these images. The aggregate particles studied were #30 to #50 crushed limestone. For each particle, a set of images were taken: the left was taken under ambient light, and the right was taken under fringe illumination. The angular particle outlines and rough sur- face texture can be clearly seen from these images. Several tests were conducted to illustrate the system’s abil- ity to image fine particles; examples of these images are shown

39 Figure 4-37. Crushed gravel (Montgomery, AL), #4–#8, magnification = 0.15. Figure 4-38. Crushed limestone (Brownwood, TX), #8–#16, magnification = 1.5. Figure 4-39. Crushed limestone (Brownwood, TX), #30–#50, magnification = 2.0.

40 Figure 4-40. Glacial gravel sand #50, imaged in the FTI system. in Figure 4-40 and Figure 4-41. In these figures, a small cap was filled nearly level with #50 glacial gravel sand, and a 3-D image was generated. The fine texture of the sand sample sur- face is clearly visible. No row-to-row errors were generated on this highly rough surface, so no error correction techniques are used here. It should be noted that the cup containing the aggregates in this test has a lip, which caused artifacts in the boundary calculation. Using a cup with straight sides should eliminate the boundary errors. Qualitative comparison with the commercial systems illustrates the fundamental resolu- tion of the systems. In the case of the FTI system, the row-by- row processing technique means that there is slightly higher resolution of fine detail in the y direction than in the x direc- tion. In both directions, the fine detail is better than the line- scan system, and comparable to the XCT system. The #8 and #16 glacial gravel—rounded (GGR) and gla- cial gravel—crushed (GGC) aggregates were analyzed using the FTI system. The FTI analysis results of GGR and GGC fine aggregate and the other fine aggregates are discussed in Chapter 6.

41 (a) Visual image (b) XCT image (c) in laser line-scan system Figure 4-41. Glacial gravel sand #50, imaged in other commercial systems.