New emerging technologies in asphalt production and placement with the promise of saving fuel, reducing plant emissions and extending the paving season into colder weather have been gaining popularity in recent years.
These technologies, warm-mix asphalts (WMAs), are produced by incorporating additives that reduce the viscosity of the asphalt binders and allow aggregate coating at temperatures well below of those of the conventional hot-mix asphalt (HMA).
In need of more time
WMA technology was developed in Europe a dozen years ago in an effort to reduce greenhouse emissions. Higher energy costs and increased environmental awareness have brought attention to the potential benefits of WMA in the U.S. in general, and North Dakota in specific. Potential benefits such as reduced plant emissions, workability at lower temperatures, extension of the paving season into colder weather, decreased binder aging and reduced energy consumption at the plant may be realized with different WMA technology applications. The reported findings of previous WMA studies were not conclusive on issues such as rutting potential and moisture sensitivity. A lot of research relating to WMA mixtures is still needed, especially in cold-region climates.
The North Dakota Department of Transportation (NDDOT), which oversees one of the highest percentages of highway pavements per capita in the nation, has been mindful of the reported potential benefits of warm mixes. Due to the short paving season in North Dakota, NDDOT has embarked on WMA research in hopes that the use of WMA will lead to the extension of the highway-paving season into colder weather. Since previous WMA research was not conclusive on rutting and moisture issues, NDDOT constructed WMA and HMA overlays to evaluate the WMA resistance to rutting and moisture sensitivity. The WMA overlay sections were constructed using the Evotherm 3G chemical additive.
The primary purpose of the research is to evaluate the rut-resistance performance of the WMA overlays. Field core samples representing the WMA and the control HMA sections near Valley City, N.D., were tested for rut resistance under dry and wet conditions using the asphalt-pavement analyzer (APA). The field specimens’ rut-resistance evaluations may give insights into the utility of using warm mixes in North Dakota.
Sixteen 6-in.-diam. field specimens (cores) from a WMA project “H-MDF-2-011(025)035” near Valley City, N.D., were collected. Sixteen HMA cores from the control section of the same WMA project also were collected. The 32 cores were identified according to location, direction of traffic and mix type.
According to the project scope, 12 WMA core specimens were prepared for APA rut-resistance testing, six of which were tested under dry conditions and another six cores were tested under wet conditions. Twelve HMA core specimens also were prepared in a similar fashion to the WMA cases. The remaining eight specimens were kept as replacements for damaged or unusable test specimens. The core specimens were cut from the bottom (using a concrete saw) to a height of 3 in. measured from the top, which is the height needed for APA rut-resistance testing. The top surfaces of the specimens were preserved in their original condition (no cutting).
The bulk specific gravities and percent air voids of the core specimens were determined prior to APA testing based on density worksheets provided by the NDDOT. Most of the field air voids were determined to be within the range of 3-5%.
Prior to APA dry-condition testing, the specimens were heated to 58°C (matching the high temperature of the PG grade of the binder used in the project) for six hours. The 58°C also will be maintained during the actual APA dry-condition testing. The wet condition involves placing the specimens in a 58°C water bath for 24 hours prior to the rut-resistance testing. Also for the wet test, the specimens were tested while immersed in 58°C water tank of the APA. The test preparation procedure was maintained for the WMA and HMA cases.
A run of tests
The utilization of the APA to evaluate rutting resistance of asphalt mixtures has been fast, cost-effective and practical to use. In this study, testing with the APA was conducted according to TP 63-03 “Standard Method of Test for Determining Rutting Susceptibility of Asphalt Paving Mixtures,” a provisional AASHTO designation with modifications to accommodate NDDOT project requirements.
The 24 field core specimens that were prepared for APA testing are set to endure 8,000 loading cycles (for both dry and wet conditions) at 100-psi pressure. Each APA run consisted of four specimens (two HMA and two WMA). Figure 1 shows four specimens placed in the molds and ready for temperature or water conditioning before testing. Two WMA specimens and two HMA specimens were tested as one run in the APA. There were a total of six runs performed in the study.
A 3?4-in. deformation is considered the criterion of rutting failure for Class 29 or lower classification pavements. The relative performances of the mixes are examined based on comparing their APA rut values. Figure 2 displays the outcome of one APA test run that includes two WMA (right) and two HMA (left).
Staying with hot
The APA rut values for the dry and wet cases are shown in Figures 3 and 4.
The numbers between 1 and 32 in Figures 3 and 4 represent the specimen numbers cored from the field. In Figures 3 and 4, the numbers (such as 21,1) represent the specimen numbers for the WMA and HMA, respectively.
The APA results indicate that the WMA mixes generally exhibited higher rut values in comparison with the HMA control specimens. For the dry condition, the average WMA rut value was 13% higher than the average rut value of the HMA mixes. And for the wet condition, the average rut value for the WMA specimens was 29% higher than that of the HMA average value. The variations between rut values were lower for the warm mixes with standard deviations of 0.46 and 0.81 for the dry and wet cases, respectively. For the hot mixes, the standard deviations were 0.78 and 1.14 for the dry and wet cases, respectively.
Nineteen out of the total 24 specimens passed the 9-mm criterion. All of the five failed specimens were from WMA mixes. Three failed specimens were from the dry condition, and the remaining two failures were from the wet condition. Therefore, the failure rates for WMA specimens stand at 50% under the dry condition and 33% under the wet condition. Six out of the seven specimens that did not fail exhibited rut values above 8 mm or close to the 9-mm failure criterion.
No specific trend was noticed between air voids and their corresponding rut value. The air voids ranged between 2.71% and 6%, with the majority of the air-void values below 5%. A plot between air voids (in percent) and rut values (in mm) for all the data points is shown in Figure 5. The R2 value is 0.023, which indicates no significant trend between the data points exists.
The APA results indicate that the WMA mixes generally exhibited higher rut values in comparison with the HMA control specimens. As reported, the average WMA rut values were 13% and 29% higher than the average rut values of the HMA mixes under dry and wet conditions, respectively. Since the premise of wet testing using the APA is done to somewhat represent durability performance, research results indicate reduced durability when using the warm mixes.
The APA rut results show that five specimens out of the 24 have failed the 9-mm rut criterion. Six out of the seven WMA specimens that did not fail exhibited rut values close to the 9-mm criterion. Although no HMA specimens failed the rut test, nearly half of the HMA specimens had rut values between 8 and 9 mm. Even though the WMA rut values were consistently higher than those of the HMA specimens, one should be cautioned that those results are based on a small sample size. To be able to come up with definitive conclusions, more WMA samples should be tested.
While there was no specific trend between the calculated air voids and the rut values, the air-void percentages were mainly on the low side.
According to the NDDOT documents, both warm and hot mixes were placed as 1.5-in. overlays. The estimated thickness of the overlay based on the 32 specimens was 2 in. Since the overall thickness of the specimens is 3 in., the lower 1 in. of the study specimens consisted of unspecified old pavement. The results of this study should be valid assuming that the older pavement is consistent throughout the section where the cores were taken. A problem could arise if some of the cores were taken directly over a crack, a pothole or other form of old pavement distress. For this research study, the PI did not suspect inconsistency problems since the specimens’ results variability measured in standard deviations are not high.
At this juncture, the author is cautious about the use of WMA pavements in North Dakota on a large scale without further testing. Future tests may include additional APA rut testing, other strength tests or field monitoring to make a definitive decision on the utility of warm mixes in North Dakota. The potential to extend the paving season into cold weather, see savings in fuel cost and realize reductions in harmful emissions are very strong incentives to continue researching warm mixes in North Dakota. R&B