The accuracy of tests, performed for the purpose of determining compliance with specification requirements, is of primary concern to consumers and producers alike.
The California Division of Highways is the user of approximately 200,000 tons of various asphaltic products per year in the construction and maintenance of its state highways. These products are obtained from about twenty different oil refineries.
During the past ten years the Materials and Research Dept. has conducted, in cooperation with the producer laboratories, four rather extensive test series for the purpose of obtaining information on the precision and reproducibility of the various test methods. Different grades of asphaltic material have been used in the cooperative tests, and all normal specification tests have been included. A statistical analysis of each of the test series is presented, and a comparison is made between values obtained in these studies and ASTM accuracy requirements between laboratories. The Control Chart Method indicates that many laboratories are not within control limits in terms of reproducibility.
A complete study of penetration test results, obtained by sampling the transport truck at the producers loading rack and again at the job site, is also presented. During the past ten years, some 7 to 13 per cent of the job site samples have been found to be “off grade” and out of specification limits when tested for penetration at the Materials and Research Laboratory. The deviation determined from the normal routine testing procedure is considerably greater than that obtained in a carefully controlled cooperative test series.