Preliminary Report Summary
Investigation of Correction Factors in Back-Calculation
of Density
Using the Superpave Gyratory Compactor
The Bituminous Section of the Texas Department of
Transportation, Materials and Tests Division, investigated some
of the effects of using corrected back-calculated laboratory molded
densities at Ndes, as used in the Superpave design
procedure. This is a brief summary of some of the interim findings
of this investigation.
Experiment Design
Four mixes were used in this investigation: three
plant mixes (Chico-Limestone, Austin-Limestone and Pharr-Gravel)
and the 19.0 mm Superpave Laboratory Standard used by the South
Central Superpave Center. The plant mixes were not designed as
Superpave mixes but were within gradation control points for 12.5
mm nominal size Superpave mixes.
For each mixture, a set of two specimens was molded
on a Superpave gyratory press at each of the following numbers
of gyration: 80, 90, 100, 110, 140, 150, 170, 190, 220 and 288.
Correction Factor as a Function of Nmax
Corrected densities were back-calculated to 80, 90
and 100 gyrations for specimens molded as described above. These
corrected densities were then compared to the measured densities
for the specimens actually molded to 80, 90 and 100 gyrations.
For two of the four mixes (Pharr-Gravel and Lab Standard), the
largest discrepancies between measured density and corrected density
were at the extreme maximum number of gyrations (220 and 190).
The maximum differences between measured and corrected densities
averaged to about 1.1%. The Austin-Limestone mixture had a very
consistent difference between measured and calculated densities,
averaging about 0.4%. The Chico-Limestone mixture had differences
averaging about 0.4%, but with much more variation. The graphs
on the following page provide a clearer representation of the
results described above.
As can be seen by the graphs, there is little correlation
between mixes in terms of the amount of error for varying maximum
number of gyrations or the design level to which they are back-calculated.
However, for these four mixes, maximum gyration numbers between
150 and 170 produced back-calculated density errors of less than
0.5%. This would result in asphalt contents that are in the range
of 0.1% to 0.2% away from the actual optimum asphalt content.
Back-Calculated Densities using Actual Height
Versus Height at Nmax
Corrected back-calculated densities were determined
using the height recorded at the maximum number of gyrations and
using the actual height of the extruded specimen. At a maximum
number of gyrations of 150, the densities calculated using the
extruded height were a consistent 0.26% less than the densities
calculated using the height recorded at Nmax.
Effect of Aggregate Degradation on Results
Sieve analyses were performed on the softest aggregate
to determine if the mixture experienced any significant aggregate
crushing during the compaction process that might impact the densities
of the molded specimens. No significant degradation was experienced.
Additional Work in Progress
Effects of the cooling rate of the mixture as it
is being compacted and its effect on densities at the higher number
of gyrations has yet to be completed. Additionally, to aid in
the control of mixtures while being produced at the hot mix plant,
a brief study on the relationship of slightly varying asphalt
contents and correction factors will be performed.
Interim Conclusions
Using the current corrected density back-calculation
procedure seems to yield design target asphalt contents within
reasonable expectations, although better design accuracy could
be obtained. However, for production control, correlation to
a measured bulk specific gravity at Ndes should be
required as a minimum; it may also be determined that production
density correction factors are sensitive to slight changes in
asphalt content, which would require multiple correction factors
when used during plant production, where asphalt contents can
vary in the range of several tenths of a percent.
It does not appear that aggregate degradation plays
a role in the difference between measured densities and corrected
back-calculated densities. Also, the use of the recorded height
at Nmax versus the height of the extruded specimen
has no significant impact on the mixture design process.
Return to Articles about Superpave
Return to Superpave Home Page
Project Manager: Dr. Yetkin Yildirim
Website Administrator: Clair LaVaye
© The Superpave Asphalt Research Program,
The University of Texas at Austin.