NOT TOO SOFT, NOT TOO DENSE: RESEARCHERS SEEK STANDARDS FOR CONCRETE IN ROADS

FAYETTEVILLE, Ark. - Density measurements help determine the quality of a roadway, but such measurements vary widely between operators. Three University of Arkansas researchers are working on a new industry standard to reduce this variability and increase highway durability.

Kevin Hall, associate professor of civil engineering, graduate student Frances Griffith and research associate Stacy Williams presented the results of their tests recently at the Transportation Research Board Annual Meeting in Washington, D.C.

"Density of an asphalt mixture correlates with strength," Hall explained. "If it is too dense, it ruts prematurely, but if it is too soft it breaks down too soon. That’s why contractors get paid for keeping the density of the mixture within a specific range."

Because density is so important, the American Society for Testing and Materials (ASTM) is seeking methods to try to reduce variability between operators. ASTM is the primary developer and provider of voluntary consensus standards, related technical information and services both in the United States and internationally.

Saturated surface dry (SSD) currently is the most commonly used method for measuring density. While it was fairly accurate for the older, dense-graded asphalts, the reliability of the SSD method has been questioned for testing the density of Superpave, which has been in widespread use across the United States since 1993.

"With the SSD method, different operators often get different results although they are performing the test on the same material, using the same equipment and following the same procedures," said Griffith.

The researchers investigated 1,300 test results from hot-mix asphalt concrete sampled at six sites in Arkansas. Density was measured by three different testing methods: SSD, dimensional analysis (height/diameter) and vacuum sealing (Corelok).

An asphalt core of specified size is weighed and then soaked in water. The outside surface is dried and the sample is weighed again. A formula is used to determine the density.

"SSD worked very well with the denser Marshall mixes," said Hall. "The problem arises with the more open Superpave mixes. Some of the water runs out before it can be weighed the second time, but the quantity varies each time."

The vacuum sealing test was conducted by using a Instrotek Corelok vacuum seal machine. A dry core sample is sealed in a vacuum bag and the mass is determined. It is then submerged in water and that mass is recorded. A formula is used to correct for the bag that sealed the specimen to determine sample density

Nine trained operators tested 144 samples from six job sites in Arkansas. All of the samples were from mixes approved for use as a surface course by the Arkansas State Highway and Transportation Department. Mix design, aggregates and asphalt content varied according to job specifications. Samples were taken at random over a two-day period.

Operators and samples were assigned randomly and each sample was tested three times with each procedure. To test multi-operator repeatability, no operator tested the same sample twice using the same test.

"The Corelok method showed a lot less operator variability than the SSD method," said Hall. "It showed a lower variability in 82 percent of the 144 cores."