Research Roundup

Relationships between Laboratory Measured Characteristics of HMA and Field Compactability

(Fabricio Leiva and Randy West)

Achieving target density is vital to building long-lasting hot-mix asphalt (HMA) pavements that resist distresses such as rutting and moisture damage. However, meeting specified density levels can be challenging, as some mixes require greater compactive effort than others.

Compactability describes the relative ease of compacting an HMA mixture to reach acceptable density levels. Several laboratory-measured parameters have been suggested as indicators of HMA compactability, but most have not been correlated with actual pavement construction data. A practical approach is needed to evaluate lab compactability and use it to estimate the required field compactive effort for HMA mixtures. This would enable mix designers to make adjustments as needed during the mix design process to improve field compactability.

The Study

The primary objective of this study was to evaluate the following laboratory-measured mixture parameters and determine correlations with field compactability:

  1. Percentage of theoretical maximum specific gravity (Gmm) at the initial number of gyrations (Nini)
  2. Compaction slope, determined from the Superpave gyratory compactor (SGC)
  3. Number of gyrations required to achieve 92 percent Gmm (N92)
  4. Compaction energy index (CEI), determined from the SGC
  5. Number of gyrations required to reach the locking point of the mixture
  6. Coarse and fine aggregate ratios, determined by using the Bailey method
  7. Mix characteristics, including gradation, aggregate shape, binder grade and volumetric properties
  8. Primary control sieve index (PCSI), representing the relative coarseness or fineness of the gradation

A secondary objective was to evaluate mix characteristics, such as gradation and binder grade, as factors affecting compactability.

Monitoring field compactibility on the NCAT Pavement Test Track

Monitoring field compactability on the NCAT Pavement Test Track

Data from Superpave mixes placed on the NCAT Test Track during the 2000 and 2003 cycles were used to evaluate laboratory-measured compaction parameters, as well as field compaction. Both surface and intermediate mixes, representing a variety of aggregates, gradation types and binder grades, were included in the analysis. Accumulated compaction pressure (ACP), a concept used to quantify compactive effort applied during rolling operations, was then related to laboratory compaction parameters.

To refine the compaction models, several additional steps were taken. Laboratory specimens were compacted to field lift thickness to determine the number of gyrations to reach 92 percent Gmm and actual field density. A number of mixes placed on the track in 2003 and 2006 were also used to obtain ACP at 92 percent Gmm, and the results were correlated with laboratory compaction parameters using multiple regression analysis. Finally, one of the compaction models was validated using data from NCHRP 9-27, Relationships of HMA In-Place Air Voids, Lift Thickness and Permeability.

Conclusions and Recommendations

CEI, N92, compaction slope and locking point were found to be simple laboratory compactability parameters. PCSI and the fine aggregate ratio (FAC) determined by the Bailey method, both of which describe gradation properties, can also be used to indicate laboratory compactability.

A multiple analysis of variance (MANOVA) was used to determine the relative effect of physical properties on laboratory compaction parameters. This analysis revealed that laboratory compaction was significantly affected by gradation type, and to a lesser degree, aggregate type and aggregate size. As expected, the most easily compacted mixes in the gyratory compactor were fine-graded, while stone matrix asphalt (SMA) mixes tended to require the greatest laboratory compactive effort.

An analysis of variance (ANOVA) was also used to determine which mixture and construction variables significantly affected ACP in the field. Both the ratio of lift thickness to nominal maximum aggregate size (t/NMAS) and mix temperature had a significant effect on ACP. Much higher compaction energy was required for t/NMAS less than 3:1 and for mixes with temperatures less than 225°F at the first pass of the breakdown roller. Due to a more rapid loss of mix temperature, mixes with lift thicknesses less than 50 mm required more compactive effort to achieve target density. Although temperature substantially affected ACP for t/NMAS less than 3:1, the effect of temperature was minor for t/NMAS ratios greater than 4:1. Although binder grade was not significant at a five percent level of significance, it did influence ACP, with stiffer binders requiring greater compactive effort. Furthermore, mixes with higher binder contents and finer gradations required less compactive effort, as expected.

Due to the range of factors affecting field compaction, a single factor correlation could not be found to effectively describe the relationship between laboratory and field compaction. However, multiple regression analysis resulted in a model (R2 = 0.82) that includes four significant factors: FAC, PCSI, surface temperature at the start of field compaction and the number of gyrations to reach field density for specimens compacted to field lift thickness (NFTFD). Thus, an additional step in the mix design process—compacting specimens to field lift thickness and calculating the number of gyrations to reach target density—can be used to help predict required field compactive effort. This model was also verified using data from 16 surface mixes placed across the U.S. as part of the NCHRP 9-27 study.

Another model (R2 = 0.92) was developed to predict the ACP required to achieve a reference density level of 92 percent Gmm. This model, which was not independently validated, includes the following terms: FAC, PCSI, percent passing the 0.075 mm sieve, binder grade, compaction slope, locking point, lift thickness and surface temperature at the start of field compaction.

The models developed in this study can be used by mix designers to predict the compactive effort required to meet target density in the field. Thus, adjustments can be made to avoid designing mixes that require unnecessarily high field compaction energy.