10. Report
10.1 Report to nearest whole unit the average of three penetrations whose values do not differ by more than the following:
11. Precision and Bias
11.1 Precision estimates for this test were developed using the AMRL proficiency sample database, which includes paired-sample data representing approximately 16 000 repetitions of the penetration test at 25°C, and approximately 4000 repetitions of the penetration test at 4°C. The materials for the database are conventional straight run and blended asphalts with penetration values ranging from 29 to 286 units measured at 25°C. Analysis of this data indicates that precision of the test can be described with the following equations:
11.2 The acceptable range of two results (95 % confidence) can be determined by multiplying the standard deviation estimates given in section 11.1 by a factor of 2.83.
11.3 This test method has no bias because the values determined are defined only in terms of the test method.
12. Keywords
12.1 asphalt; bitumen; penetration