|
“So AAPA had a series of meetings with the
Alabama Department of Transportation,” Monk says. The upshot of those
meetings was that the two parties came up with a “Locking Point” design
principle to govern N-design. “At a certain point the aggregate
structure in a mixture locks up,” says Monk. “The mix has been compacted
to an optimum point. We felt like once you go beyond that point you
degrade your aggregate and, secondly, you reduce the amount of space
available for liquid asphalt in your design.”
With the Locking Point design, a print-out shows
the lab technician the height of the compacted specimen vs. the number
of gyrations, as the gyratory compactor is working. The shorter the
height, of course, the more the specimen has been compacted. “In
Alabama, we said that when you have two consecutive gyrations that
produce no change in specimen height, the second gyration is the Locking
Point,” says Monk.
Monk says the DOT was still concerned about
setting N-design too low, because data showed that most mixes “locked
up” in the range of 45 to 55 gyrations. So the DOT set the minimum
N-design at 60 gyrations. “If the locking point is below 60, we use 60,”
says Monk. “But if it’s above 60, we use the higher number.”
Asphalt content? Compared to mixes based on an
N-design of 85, the binder content of the new N-60 mix has increased 0.2
to 0.4% as a general average.
The improvement has been marked, Monk says.
“We’re seeing a big difference in the field. The mixes are easier to
compact, which in turn makes them less permeable and improves our
durability,” he explains. “And we haven’t added so much liquid that we
would incur rutting problems. We’ve been very pleased, and so has the
DOT.”
Monk responds to Huber’s point about voids in
the mineral aggregate by saying that the Locking Point design does not
ignore VMA. “We’re getting the same or higher VMA with our new
procedure, but I’m a proponent of lowering N-design at the same time,”
says Monk.
Reduced air voids
In Colorado, the DOT now may decide, on a
case-by-case basis, to lower the air voids content from 4 to 3%. The
contractor still must optimize his mix at 4%. “Then we allow ourselves to
target up to 1% less air voids in production,” says Bill Schiebel, the DOT’s
Region 1 materials engineer.
“There’s more than one way to fill VMA,” says
Schiebel. “We want to ensure that decreases in air voids are due to an
increase in liquid binder content. If you tell somebody to design a mix at
3% air voids instead of 4%, they can plug up the voids with things like
rounded or dirty fines.
“So we adjust the air voids target after the mix
design is complete,” Schiebel says. “In some mixes, a change of 1% air voids
will result in an increase of 0.3% of binder, and in other mixes it might
change binder content by only 0.2 or 0.1%. Each mix reacts differently,
depending on the aggregate gradations and the aggregate source.”
Colorado made the change to lower air voids and
higher asphalt content as a result of a 2002 study that confirmed problems
with Superpave durability and fingered inadequate binder content as the
culprit. “The study showed that our lab-designed mixes were over-compacted
and they were too dry relative to the way they were supposed to perform,”
says Schiebel.
He says Superpave mixes designed at 4% air voids
were being laid down at 6% air. But those pavements were not seeing the
expected 2% of additional compaction under traffic. “We tested densities at
one-two-three-four-five years, and there was very little compaction in the
field,” says Schiebel. “That’s because the gyratory compactors are
compacting too much — more than rollers and traffic. We were at 5.4% voids
after three years, and we only got down to 5.2%. That was 1.2% higher than
it should have been, which was 4.0%. The mixes were too stiff to get any
more compaction.”
Improvement noticed
Has the additional asphalt helped? “Yes, we believe
so, this far,” says Schiebel. “The natural fear is that you’re moving toward
the rutting line, but we felt we were a long way from the rutting line. We
wanted to make a first step, a first iteration, and now we’re going to
repeat the same density study.
“We’re seeing improvements in terms of decreased
segregation, improved compaction during construction, improved density at
our longitudinal joints,” Schiebel says. “And the jury is still out on
durability, but I’m sure we’ll see an improvement in that.”
Like Colorado, Maryland has reduced design air voids
in its Superpave designs. Instead of designing mixes at 4%, the state
specifies 3%, says Brian Dolan, president, Maryland Asphalt Association.
The change permits an increase in binder of 0.2%. “With mutual agreement
between the industry and the State Highway Administration, this would be the
first step toward increasing asphalt content in order to improve the
durability of mixes,” says Dolan.
“Here in Maryland we’re encouraging county and local
governments to use Level 1 mixes — the 50 gyration mixes, instead of the 75
or 100 gyrations required in Level 2 or 3,” says Dolan. “It makes a fairly
significant difference in the binder content. Between Level 3 and Level 1
there’s probably an increase of 0.5 to 1.0% asphalt content.
Binder bumping
Virginia’s experience has been different. The state
implemented the Superpave PG binder system in 1997, using PG grades with
their Marshall hammer mix design system, says Richard Schreck, Executive
Vice President of the Virginia Asphalt Association.
“Rut testing analysis by Brian Prowell, a senior
research scientist (at the time) at the Virginia Transportation Research
Council, found that the 50-blow Marshall mixes with the stiffer PG 70-22
binder had better rut resistance than the 75-blow Marshall with the same
binder,” says Schreck.
“Virginia started using ‘binder bumping’ and a
50-blow Marshall mix design for all traffic levels,” says Schreck. “Binder
bumping” means VDOT would use progressively stiffer binders as the design
traffic loading increased. “From a subdivision up to the heaviest truck
loading, our asphalt contents did not change. We just use a one-grade
stiffer binder for the heavier loadings, and stayed with the 50-blow
Marshall,” he recalls. “We had three years of that, starting in 1997, and it
worked beautifully.”
When Virginia implemented the Superpave mix design
in 2000, technicians found they had to go to coarse-graded mixes to maintain
VMA at the high gyration levels of design compaction. Those mixes didn’t
work as well. “When you placed them they ended up with low binder contents,
were more permeable, and they did not appear to be very durable,” says
Schreck. “We also did not accept the concept of the “Restricted Zone”
because all our best mixes from the Marshall era passed through it.”
 |
|
Many states have found that
Superpave designs need more asphalt content. |
By the third year of full Superpave implementation,
thanks to Prowell’s work at VTRC, Virginia was using 65 gyrations for
N-design on all traffic levels and continuing its “binder bumping” practice.
Schreck says recent VTRC studies conclude that 65 gyrations are still too
high compared to the Marshall mixes that performed so well. “For our
materials, 65 gyrations is close to a 75-blow Marshall and we know from past
experience that 75-blow Marshalls were not good durable mixes,” Schreck
says.
With the large-mix-design sample size, Virginia has
found it difficult to reduce gyrations further as a way to increase asphalt
content. “Instead we have increased the Voids-Filled-with-Asphalt Criteria,
and shifted the production mix Voids in Total Mix range to the low side,”
says Schreck. “It’s not the best fix, but with our materials and the
large-mix sample size we’re struggling with how to increase binder content
in a scientific way. We have also introduced permeability testing as part of
the mix design process to eliminate the permeable coarse graded mixes.”
So is Superpave a bust? No, but it’s evolving and
needs fine tuning. Some argue that no one design compaction level is right
for everybody’s aggregates across the country. N-design needs to change to
accommodate various aggregate types, they say. Jim Huddleston, executive
director, Asphalt Pavement Association of Oregon, is confident that
Superpave mixes can work. “With Superpave, we have the potential to get the
best of both worlds,” he says. “We can have durable, rut-resistant mixes
with lower air voids, sufficient asphalt content for durability, and an
aggregate gradation that will stand up under traffic.”
How Superpave Works
Superpave is a volumetric system of mix design. To
succeed, it requires careful control over air voids (typically specified at
4%), voids in the mineral aggregate, proper proportioning in the aggregate
blend, and binder content. VMA is the amount of interstitial space between
the stones in the aggregate blend for a given mixture.
Enter N-design, the level of design gyrations in the
laboratory’s gyratory compactor. For high-volume pavements, N-design is
typically set at 125 or 100 gyrations. Some in the industry argue that a
lower N-design, or fewer lab gyrations, will allow more space between
aggregates to boost the binder content.
That logic is flawed, says Gerald A. Huber,
associate director of research at the Heritage Research Group in
Indianapolis. He says that reducing N-design will do little or nothing to
change asphalt binder content, if VMA is left at the level required in the
original design.
“In actual practice, if design gyrations are
reduced, the mix will be redesigned to bring asphalt content back into
line,” says Huber. “As a result, design gyrations will not change asphalt
content because there is no requirement for the aggregate blend to be held
constant.”
Proponents of increased asphalt content also say
that the additional binder will make mixes more compactable and that film
thickness will increase on the aggregate. So less segregation is likely to
occur, especially with coarser mixes.
Granted, the solution may be to increase asphalt
content, Huber says. But specifiers cannot do it by lowering N-design, he
stresses. “The way to accomplish increased asphalt content is to increase
the VMA, and allow the aggregate to accommodate more asphalt binder.” |
