APAI Article Archive

(APAI Web Site, April 2006)

HMA NOTES: Lift Thickness - More Than Meets The Eye

By: Bob Nady, P.E.

Sure, you look at the typical cross section and you note the thickness of the lift you are abut to place. On the road, you set the correct blocks under the screed, bring the screed onto the blocks and take off. Using your mat depth gauge, you monitor and adjust the thickness as indicated, so that you get the plan thickness after rolling.

But there is a bit more to it than that. How was the lift thickness determined? How about the relationship between lift thickness and mat quality, including density, permeability and maybe smoothness? How was that magic number selected?

In the past, there was lots of discussion about what the lift thickness should be. But there was no hard evidence to support the selection process, except past practice - we always use 1-1/2 inches for surface course, etc. Now we have the results of an important study to support these decisions.

The National Center for Asphalt Technology (NCAT) recently completed research project NCHRP 0-27, which included both a laboratory phase and a field phase and answers many of these questions. There was both a laboratory phase and a field phase of this research.

The overall study included different mix gradations and nominal maximum aggregate sizes, different materials, and different mix types (Superpave and SMA). The gyratory compaction was used to compact the lab samples, all at the same compaction level (100 gyrations). The density of these samples was measured in the laboratory.

In the field phase of the study, mix was laid in several different sections, (one or more sections for each mix type and gradation) coarse graded or fine graded Superpave. The sections were laid with a paver, each starting as either a thin section start or a thick section start at each location. Using the thickness screws on the screed, the mat thickness was gradually changed, thickened or thinned, so that each section had a range of thicknesses. Compaction was by a selected number of passes on an 11 ton double drum steel roller. It could operate either in vibratory or static mode, giving each section the same number and type of compaction passes.

A large number of cores were drilled from these test sections. Each core was measured for thickness and density was determined in the laboratory. For each section we have a number of cores all compacted by the same effort, using the same mix and at a range of thicknesses.
For each lab sample and each field core a ratio was calculated of lift thickness over nominal maximum aggregate size (t/nmas). Thus for each type we have a range of ratios from about 2 - 5. For example, a 2 inch lift of a ½ - inch mix would be a ratio of 4. The results are significant. At a ratio of 2, the density is about 4 percent below the optimum value (at mix design of 4 percent air voids). The difference is about 1 percent when the ratio is increased to 3. Of course, there are some minor differences for different mix types, different mix sizes, and different materials in the mix. But the results are strikingly similar.

It should be noted that the results are for density. The inverse of density is air voids, and of course, at the higher ratios, the density goes up, the air voids come down. One would expect that as air voids go down, permeability also goes down. All making for a more durable pavement.

Important factors to consider:

Room: it is possible to put three trombone players, with their instruments, into a telephone booth. But the lack of room will certainly have a negative impact on the quality of music these musicians deliver. The same is true of HMA: adequate lift thickness not only makes possible the expulsion of air during compaction, it facilitates coated particle reorientation. The aggregates are free to assume optimum orientation and position. This has a positive impact on mat quality - density, voids, permeability, and likely smoothness.

Temperature: Thin HMA layers cool significantly faster than thicker lifts. Cooling results in an increase in binder viscosity (stiffness). The time to achieve density is greatly reduced. We now have the advantage of the NCAT research which documents these factors with numbers.

The following points were included in the results of the NCAT study:
+ at a ratio of 2, mat density is about 4% below optimum
+ at a ratio of 3, mat density is about 1% below optimum
+ at a ratio of 4, mat density is at optimum
+ no significant improvement was noted for ratios of 5 or more
+ a 1-inch mat cools twice as fast as a 1 ½ inch mat.

In the real world, you have a job that calls for a 2-inch mat using a ¾ -inch mix. You realize that something should be done, because the ratio is only 2.7. Things can be improved by using a 3-inch layer. But the owner expresses concern about the increased cost. You have another card to play: suggest using a ½ -inch mix, your ratio is still at 4 and a 2-inch lift. Everyone is happy. The quality of the mat will be assured with the ratio at 4 and the costs will be essentially the same using the ½ - inch mix. A win, win situation.

References:
Brown, Hainin, Cooley, AAPT Journal, Vol. 74, 2005.
NCAT, Asphalt Technology News, Vol. 17, No. 2
NAPA, HMAT Magazine, Vol. 11, No. 1