Precision Estimates for AASHTO Test Method T308 and the Test Methods for Performance-Graded Asphalt Binder in AASHTO Specification M320 (2005)

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.

11 CHAPTER 3: RESULTS OF ANALYSIS AND ESTIMATES OF PRECISION 3.1 TEST DATA The individual results for each of the 91 proficiency data sets used to create precision estimates can be found in Appendices G to R. This chapter includes summaries of the data and the resulting precision estimates. 3.2 ANALYSIS OF THE DATA The following tables and, in some cases, graphs display the results of the analyses. Precision estimates are based, where appropriate, on either the coefficients of variation (CV%) or the pooled standard deviation (1s) values. In one instance, an equation is used to express precision. 3.2.1 Determining the Asphalt Binder Content of Hot-Mix Asphalt (HMA) by the Ignition Method, AASHTO T308 Results from analyzing the data for Asphalt Binder Content of HMA by the Ignition Method are found in Appendix G. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% IGN 3 & 4 353 PG 64-22 AC 20 4.049 4.256 0.064 1.57 1.49 0.107 2.63 0.108 2.55 IGN 5 & 6 399 PG 64-22 AC 20 4.802 5.098 0.072 1.49 1.41 0.119 2.47 0.116 2.28 IGN 7 & 8 461 PG 52-34 AC 10 4.480 4.745 0.072 1.60 1.51 0.124 2.77 0.121 2.55 Average Results Repeatability Reproducibility Reproducibility odd even samples samples Table 4 – Summary Table for T308, Percent Asphalt (%) A review of the data shown in Table 4 indicates that the form of the precision estimates should be based on the sample standard deviation. The pooled repeatability sample standard deviation for the three pairs of samples analyzed is 0.069 percent. The corresponding pooled reproducibility sample standard deviation is 0.117 percent. The pooled estimates are derived using the following equation from Ku [14]: ( ) ( ) ( ) knnn snsnsns k kk p −+++ −++−+−= ... 1...11 2 22 22 2 11 1 (Equation 9) Where: =ps pooled standard deviation =ks kth standard deviation =kn number of laboratories analyzed resulting in kth standard deviation

12 3.2.2 Flash and Fire Points by Cleveland Open Cup, T48 Results from analyzing the data for Flash Point by Cleveland Open Cup are found in Appendix H. There are no modified binders used in the analysis. Additionally, fire point data are not collected in the AMRL PSP. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% BAC 181 & 182 107 PG 64-16 AC 10 271.7 272.3 2.4 0.9 0.9 9.1 3.3 9.6 3.5 BAC 183 & 184 98 PG 70-22 -- 353.1 353.5 2.5 0.7 0.7 9.2 2.6 8.5 2.4 BAC 185 & 186 113 PG 64-22 AC 20 323.0 323.7 3.5 1.1 1.1 12.3 3.8 12.7 3.9 BAC 187 & 188 116 PG 64-22 AC 30 274.0 273.8 2.5 0.9 0.9 12.2 4.5 11.9 4.3 BAC 189 & 190 134 PG 64-22 AC 30 318.1 317.8 2.8 0.9 0.9 9.0 2.8 8.9 2.8 BAC 191 & 192 121 PG 52-34 AC 10 271.9 268.5 4.4 1.6 1.6 12.3 4.5 12.1 4.5 BAC 193 & 194 118 PG 64-22 AC 20 330.8 331.4 2.4 0.7 0.7 7.1 2.2 7.3 2.2 PGB 195 & 196 148 PG 70-22 -- 350.8 350.8 1.7 0.5 0.5 7.7 2.2 7.4 2.1 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 5 – Summary Table for T48, Flash Point (°C) A review of the data shown in Table 5 indicates that the form of the precision estimates should be based on the sample standard deviation. The pooled repeatability sample standard deviation for the eight pairs of samples analyzed is 3°C. The corresponding pooled reproducibility sample standard deviation is 10°C. The pooled estimates are derived using Equation 9. 3.2.3 Specific Gravity of Semi-Solid Bituminous Materials, T228 Results from analyzing the data for Specific Gravity can be found in Appendix I. There are no modified binders used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% BAC 181 & 182 104 PG 64-16 AC 10 1.0159 1.0157 0.0006 0.060 0.060 0.0012 0.119 0.0012 0.119 BAC 183 & 184 101 PG 70-22 -- 1.0425 1.0428 0.0010 0.100 0.100 0.0016 0.156 0.0017 0.161 BAC 185 & 186 104 PG 64-22 AC 20 1.0330 1.0329 0.0008 0.076 0.076 0.0013 0.129 0.0013 0.124 BAC 187 & 188 112 PG 64-22 AC 30 1.0345 1.0344 0.0006 0.062 0.062 0.0011 0.110 0.0013 0.126 BAC 189 & 190 112 PG 64-22 AC 30 1.0308 1.0308 0.0006 0.061 0.061 0.0010 0.095 0.0008 0.077 BAC 191 & 192 121 PG 52-34 AC 10 1.0273 1.0274 0.0007 0.071 0.071 0.0013 0.127 0.0012 0.118 BAC 193 & 194 110 PG 64-22 AC 20 1.0058 1.0058 0.0006 0.062 0.062 0.0010 0.098 0.0008 0.079 PGB 195 & 196 137 PG 70-22 -- 1.0404 1.0404 0.0007 0.067 0.067 0.0014 0.134 0.0013 0.125 Average Results Repeatability Reproducibility Reproducibility odd even samples samples Table 6 – Summary Table for T228, Specific Gravity A review of the data shown in Table 6 indicates that the form of the precision estimates should be based on the sample standard deviation. The pooled repeatability sample standard deviation for the eight pairs of samples analyzed is 0.0008. The corresponding pooled reproducibility sample standard deviation is 0.0013. The pooled estimates are derived using Equation 9.

13 3.2.4 Effect of Heat and Air on a Moving Film of Asphalt (Rolling Thin-Film Oven Test), T240 Results from analyzing the data for change in mass using the RTFO can be found in Appendix J. One pair of modified binders, sample numbers 187 and 188, is used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 170 PG 64-16 AC 10 -0.2740 -0.2646 0.0160 5.8 6.0 0.0570 20.8 0.0568 21.4 PGB 183 & 184 172 PG 70-22 -- -0.0515 -0.0505 0.0087 16.9 17.2 0.0211 40.9 0.0203 40.3 PGB 185 & 186 166 PG 64-22 AC 20 -0.2658 -0.2630 0.0149 5.6 5.7 0.0433 16.3 0.0424 16.1 PGB 187 & 188 174 PG 76-22 -- -0.3435 -0.3363 0.0212 6.2 6.3 0.0722 21.0 0.0676 20.1 PGB 189 & 190 171 PG 64-22 AC 30 -0.0358 -0.0369 0.0076 21.3 20.7 0.0219 61.1 0.0218 59.0 PGB 191 & 192 191 PG 52-34 AC 10 -0.5133 -0.5107 0.0233 4.5 4.6 0.0827 16.1 0.0811 15.9 PGB 193 & 194 176 PG 64-22 AC 20 -0.0321 -0.0336 0.0063 19.7 18.9 0.0219 68.1 0.0225 66.9 PGB 195 & 196 191 PG 70-22 -- -0.0515 -0.0503 0.0074 14.3 14.6 0.0190 36.8 0.0184 36.6 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 7 – Summary Table for T240, Change in Mass (%) A review of the data shown in Table 7 indicates that the standard deviation can be expressed as a function of the mass change (x) by using the equations described below. The repeatability standard deviation, sr, for the eight pairs of samples analyzed is determined to be best described using the following equation: sr = 0.0061 + 0.0363(x) (Equation 10) The reproducibility standard deviation, sR, for the eight pairs of samples analyzed is determined to be best described using the following equation: sR = 0.0153 + 0.1365(x) (Equation 11) Where: sr = reproducibility standard deviation sR = reproducibility standard deviation x = mass loss in percent (a loss of mass is expressed as a negative number) y = -0.0363x + 0.0061 R2 = 0.9601 0.000 0.005 0.010 0.015 0.020 0.025 0.030 -0.600 -0.500 -0.400 -0.300 -0.200 -0.100 0.000 Average Loss Per Sample (%) S ta nd ar d D ev ia tio n Figure 4 – Repeatability Graph for T240 y = -0.1365x + 0.0153 R2 = 0.9541 0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 -0.600 -0.500 -0.400 -0.300 -0.200 -0.100 0.000 Average Loss Per Sample (%) S ta nd ar d D ev ia tio n Figure 5 – Reproducibility Graph for T240

14 3.2.5 Determining the Flexural Creep Stiffness of Asphalt Binder Using the Bending Beam Rheometer (BBR), T313 3.2.5.1 SLOPE Results from analyzing the data for BBR Slope can be found in Appendix K. One pair of modified binders, sample numbers 187 and 188, is used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 174 PG 64-16 AC 10 0.3686 0.3697 0.0041 1.10 1.10 0.0085 2.31 0.0090 2.44 PGB 183 & 184 178 PG 70-22 -- 0.3341 0.3339 0.0039 1.16 1.16 0.0067 2.01 0.0077 2.29 PGB 185 & 186 182 PG 64-22 AC 20 0.3303 0.3297 0.0037 1.12 1.13 0.0084 2.55 0.0090 2.72 PGB 187 & 188 189 PG 76-22 -- 0.3735 0.3730 0.0039 1.04 1.04 0.0098 2.61 0.0089 2.39 PGB 189 & 190 185 PG 64-22 AC 30 0.3135 0.3135 0.0032 1.02 1.02 0.0076 2.41 0.0081 2.58 PGB 191 & 192 189 PG 52-34 AC 10 0.3106 0.3101 0.0031 0.99 0.99 0.0083 2.66 0.0086 2.76 PGB 193 & 194 196 PG 64-22 AC 20 0.3090 0.3086 0.0025 0.80 0.80 0.0058 1.88 0.0060 1.94 PGB 195 & 196 196 PG 70-22 -- 0.3202 0.3197 0.0029 0.92 0.92 0.0083 2.58 0.0083 2.59 Average Results Repeatability Reproducibility Reproducibility odd even samples samples Table 8 – Summary Table for T313, Slope (m-value) A review of the data shown in Table 8 indicates that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed is 1.0 percent. The corresponding average reproducibility coefficient of variation is 2.4 percent. In each case, the average coefficient of variation is determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15]. 3.2.5.2 STIFFNESS Results from analyzing the data for BBR Stiffness can be found in Appendix L. One pair of modified binders, sample numbers 187 and 188, is used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 179 PG 64-16 AC 10 182.8 179.2 6.1 3.32 3.39 14.7 8.03 14.5 8.09 PGB 183 & 184 188 PG 70-22 -- 179.0 179.6 3.9 2.19 2.18 11.7 6.54 12.0 6.68 PGB 185 & 186 181 PG 64-22 AC 20 197.3 196.2 5.6 2.83 2.85 11.4 5.78 12.1 6.16 PGB 187 & 188 184 PG 76-22 -- 125.4 125.5 3.2 2.59 2.59 7.8 6.22 8.4 6.73 PGB 189 & 190 192 PG 64-22 AC 30 216.7 216.1 5.8 2.68 2.69 12.8 5.89 13.2 6.09 PGB 191 & 192 181 PG 52-34 AC 10 225.9 225.4 4.9 2.18 2.18 14.0 6.21 14.5 6.45 PGB 193 & 194 193 PG 64-22 AC 20 158.7 158.9 3.6 2.28 2.28 8.8 5.55 8.2 5.19 PGB 195 & 196 187 PG 70-22 -- 235.7 236.8 5.0 2.13 2.12 12.6 5.37 13.2 5.56 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 9 – Summary Table for T313, Creep Stiffness (MPa) A review of the data shown in Table 9 indicates that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed is 2.5 percent. The corresponding average reproducibility coefficient of variation is 6.3 percent. In each case, the average coefficient of variation is determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15].

15 3.2.6 Determining the Fracture Properties of Asphalt Binder in Direct Tension (DT), T314 3.2.6.1 STRESS Results from analyzing the data for Direct Tension Stress can be found in Appendix M. One pair of modified binders, sample numbers 187 and 188, is used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 37 PG 64-16 AC 10 2.785 2.816 0.265 9.5 9.4 0.532 19.1 0.572 20.3 PGB 183 & 184 44 PG 70-22 -- 3.953 3.970 0.287 7.3 7.2 0.797 20.2 0.781 19.7 PGB 185 & 186 51 PG 64-22 AC 20 3.707 3.783 0.269 7.3 7.1 0.592 16.0 0.630 16.7 PGB 187 & 188 53 PG 76-22 -- 3.353 3.402 0.205 6.1 6.0 0.601 17.9 0.577 16.9 PGB 189 & 190 59 PG 64-22 AC 30 3.775 3.860 0.306 8.1 7.9 0.878 23.3 0.939 24.3 PGB 191 & 192 53 PG 52-34 AC 10 4.172 4.108 0.261 6.3 6.3 0.648 15.5 0.681 16.6 PGB 193 & 194 60 PG 64-22 AC 20 4.034 3.967 0.255 6.3 6.4 0.735 18.2 0.787 19.9 PGB 195 & 196 54 PG 70-22 -- 4.218 4.191 0.343 8.1 8.2 0.645 15.3 0.714 17.0 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 10 – Summary Table for T314, Stress (MPa) Following a review of the data in Table 10, coefficient of variation (CV%) was chosen as the estimate of precision. The average repeatability coefficient of variation for the eight pairs of samples analyzed was determined to be 7.4 percent. The corresponding average reproducibility coefficient of variation was determined to be 18.6 percent. In each case, the average coefficient of variation was determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15]. 3.2.6.2 STRAIN Results from analyzing the data for Direct Tension Strain can be found in Appendix N. One pair of modified binders, sample numbers 187 and 188, is used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 34 PG 64-16 AC 10 0.91 0.94 0.10 10.9 10.5 0.26 28.9 0.25 26.4 PGB 183 & 184 44 PG 70-22 -- 1.83 1.85 0.24 13.3 13.2 0.65 35.5 0.72 38.7 PGB 185 & 186 53 PG 64-22 AC 20 1.38 1.45 0.17 12.2 11.6 0.34 25.0 0.37 25.2 PGB 187 & 188 54 PG 76-22 -- 2.96 3.00 0.29 9.7 9.6 1.29 43.6 1.23 40.9 PGB 189 & 190 56 PG 64-22 AC 30 1.33 1.39 0.15 11.6 11.2 0.42 31.7 0.46 33.2 PGB 191 & 192 49 PG 52-34 AC 10 1.39 1.35 0.11 8.1 8.3 0.27 19.7 0.28 20.9 PGB 193 & 194 61 PG 64-22 AC 20 2.69 2.66 0.36 13.2 13.4 1.02 37.9 1.00 37.5 PGB 195 & 196 56 PG 70-22 -- 1.36 1.35 0.17 12.8 12.9 0.39 28.8 0.41 29.9 samples samples Repeatability Reproducibility Reproducibility odd even Average Results Table 11 – Summary Table for T314, Percent Strain A review of the data shown in Table 11 indicated that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed was determined to be 11.4 percent. The corresponding average reproducibility coefficient of variation was determined to be 31.5 percent. In each case, the average coefficient of variation was determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15].

16 3.2.7 Determining the Rheological Properties of Asphalt Binder Using a Dynamic Shear Rheometer (DSR), T315 3.2.7.1 Original Binder: G*/ sinδ Results from analyzing the data for DSR testing on original binder can be found in Appendix O. One pair of modified binders, sample numbers 187 and 188, was used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 185 PG 64-16 AC 10 1.109 1.067 0.028 2.55 2.65 0.061 5.54 0.061 5.69 PGB 183 & 184 192 PG 70-22 -- 1.345 1.348 0.032 2.40 2.39 0.083 6.14 0.078 5.80 PGB 185 & 186 189 PG 64-22 AC 20 1.287 1.280 0.026 2.00 2.01 0.070 5.46 0.073 5.71 PGB 187 & 188 189 PG 76-22 -- 1.397 1.409 0.034 2.44 2.42 0.073 5.21 0.073 5.19 PGB 189 & 190 205 PG 64-22 AC 30 1.576 1.572 0.034 2.17 2.17 0.092 5.82 0.086 5.46 PGB 191 & 192 208 PG 52-34 AC 10 2.338 2.342 0.059 2.50 2.50 0.182 7.79 0.176 7.53 PGB 193 & 194 208 PG 64-22 AC 20 1.376 1.370 0.026 1.89 1.90 0.080 5.80 0.078 5.66 PGB 195 & 196 208 PG 70-22 -- 1.444 1.448 0.030 2.08 2.08 0.095 6.61 0.095 6.53 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 12 – Summary Table for T315, Original G*/ sinδ A review of the data shown in Table 12 indicated that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed was determined to be 2.3 percent. The corresponding average reproducibility coefficient of variation was determined to be 6.0 percent. In each case, the average coefficient of variation was determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15]. 3.2.7.2 RTFO Residue: G*/ sinδ Results from analyzing the data for DSR testing on RTFO residue can be found in Appendix P. One pair of modified binders, sample numbers 187 and 188, was used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 186 PG 64-16 AC 10 2.374 2.274 0.079 3.34 3.49 0.182 7.67 0.174 7.63 PGB 183 & 184 184 PG 70-22 -- 2.626 2.645 0.102 3.88 3.85 0.196 7.45 0.201 7.60 PGB 185 & 186 188 PG 64-22 AC 20 3.062 3.063 0.087 2.83 2.83 0.227 7.40 0.220 7.18 PGB 187 & 188 195 PG 76-22 -- 2.972 2.975 0.110 3.70 3.69 0.264 8.88 0.236 7.94 PGB 189 & 190 198 PG 64-22 AC 30 3.641 3.646 0.107 2.93 2.93 0.268 7.36 0.254 6.96 PGB 191 & 192 199 PG 52-34 AC 10 7.683 7.733 0.230 2.99 2.97 0.780 10.15 0.732 9.46 PGB 193 & 194 204 PG 64-22 AC 20 3.185 3.191 0.097 3.04 3.03 0.241 7.57 0.243 7.62 PGB 195 & 196 205 PG 70-22 -- 2.539 2.557 0.069 2.72 2.70 0.176 6.94 0.195 7.62 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 13 – Summary Table for T315, RTFO G*/ sinδ A review of the data shown in Table 13 indicated that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed was determined to be 3.2 percent. The corresponding average reproducibility coefficient of variation was determined to be 7.8 percent. In each case, the average coefficient of variation was determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15].

17 3.2.7.3 PAV Residue: G* sinδ Results from analyzing the data for the DSR testing on PAV residue can be found in Appendix Q. One pair of modified binders, sample numbers 187 and 188, was used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 181 PG 64-16 AC 10 4557 4489 249 5.5 5.6 695 15.3 656 14.6 PGB 183 & 184 178 PG 70-22 -- 2310 2334 117 5.1 5.0 293 12.7 313 13.4 PGB 185 & 186 178 PG 64-22 AC 20 3830 3818 223 5.8 5.8 526 13.7 486 12.7 PGB 187 & 188 185 PG 76-22 -- 1100 1102 61 5.6 5.6 167 15.1 157 14.3 PGB 189 & 190 182 PG 64-22 AC 30 4335 4340 143 3.3 3.3 597 13.8 603 13.9 PGB 191 & 192 185 PG 52-34 AC 10 3640 3673 171 4.7 4.7 660 18.1 660 18.0 PGB 193 & 194 188 PG 64-22 AC 20 2922 2937 137 4.7 4.7 364 12.5 359 12.2 PGB 195 & 196 199 PG 70-22 -- 3163 3171 137 4.3 4.3 432 13.7 424 13.4 odd even samples samples Average Results Repeatability Reproducibility Reproducibility Table 14 – Summary Table for T315, PAV G* sinδ A review of the data shown in Table 14 indicated that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed was determined to be 4.9 percent. The corresponding average reproducibility coefficient of variation was determined to be 14.2 percent. In each case, the average coefficient of variation was determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15]. 3.2.8 Viscosity Determination of Asphalt Binder Using a Rotational Viscometer, T316 Results from analyzing the data for Viscosity Determination can be found in Appendix R. One pair of modified binders, sample numbers 187 and 188, was used in the analysis. Sample Sample No. of PG AC odd even odd even Type Numbers Labs Grade Grade samples samples samples samples 1s CV% CV% 1s CV% 1s CV% PGB 181 & 182 142 PG 64-16 AC 10 0.277 0.272 0.004 1.27 1.30 0.015 5.40 0.015 5.52 PGB 183 & 184 176 PG 70-22 -- 0.715 0.719 0.008 1.12 1.11 0.028 3.96 0.029 4.08 PGB 185 & 186 172 PG 64-22 AC 20 0.414 0.414 0.005 1.27 1.27 0.017 4.02 0.015 3.71 PGB 187 & 188 180 PG 76-22 -- 1.621 1.638 0.020 1.25 1.23 0.070 4.34 0.069 4.19 PGB 189 & 190 179 PG 64-22 AC 30 0.439 0.439 0.005 1.18 1.18 0.016 3.59 0.016 3.55 PGB 191 & 192 192 PG 52-34 AC 10 0.290 0.291 0.005 1.69 1.68 0.012 4.06 0.013 4.31 PGB 193 & 194 202 PG 64-22 AC 20 0.445 0.445 0.005 1.22 1.22 0.020 4.54 0.020 4.48 PGB 195 & 196 195 PG 70-22 -- 0.685 0.688 0.006 0.84 0.84 0.031 4.47 0.031 4.44 samples samples Repeatability Reproducibility Reproducibility odd even Average Results Table 15 – Summary Table for T316, Viscosity (Pa·s) A review of the data shown in Table 15 indicated that the form of the precision estimates should be based on the coefficient of variation (CV%). The average repeatability coefficient of variation for the eight pairs of samples analyzed was determined to be 1.2 percent. The corresponding average reproducibility coefficient of variation was determined to be 4.3 percent. In each case, the average coefficient of variation was determined by calculating the “simple arithmetic average” as described in Section 8.4.2 of ASTM C802-96 [15].