Specifying Shrinkage Limited Cement (Type SL)
Author: J. Mohammadi, W. South
Type: Journal Paper
Publisher: ASTM International
Source: Journal of Testing and Evaluation. 2016 Sep 09; 45(5)
Keywords: shrinkage limited cement, cement type SL, mortar drying shrinkage, proficiency assessment, repeatability, reproducibility
The Australian Standard AS 3972 specifies the maximum limit of 750microstrain for the 28-day mortar drying shrinkage of samples prepared with shrinkage limited cement (Type SL). Although the current shrinkage limit is 750microstrain, AS 3972 recommends a 600 microstrain for the maximum target or average of drying shrinkage of mortar samples. The lower recommended shrinkage level is justified in the Australian Standard by considering the inherent errors of the mortar drying shrinkage test method AS 2350.13, determined in 1997. The lower recommended shrinkage level is to provide a sufficient level of confidence that a single result obtained from different laboratories may not exceed the maximum limit of 750 microstrains. However, the “informative” recommendation of AS 3972 relating to the maximum average 600microstrain shrinkage is considered a “normative” term by other specifiers in Australia such as road authorities and ATIC-SPEC SP-43. The current study investigated the past ten-years mortar drying shrinkage test results from round-robin proficiency programs of the major Australian laboratories to calculate the current inherent repeatability and reproducibility error of performing AS 2350.13 test. The average ten-year reproducibility error at 95% confidence level was calculated 18.6%, and therefore, it is recommended to increase the recommendation for the maximum average drying shrinkage guidance in Australian Standard AS 3972 Section A3.3.5 from600 to 630microstrain. Moreover, repeatability of test was calculated to be 6.3%. This was equal to maximum of 44microstrains for Type SL cement with maximum drying shrinkage of 750microstrain. Results were compared with repeatability of other international test methods and some recommendations for improving the repeatability test procedure were provided for any future research in this area.