Much of the industrial world’s transportation infrastructure has reached a critical state of deterioration at a time when revenues to fund its replacement are scarce.
Against that backdrop the National Lime Association (NLA) funded a comprehensive two-year hot-mix asphalt (HMA) study. The study used typical state department of transportation (DOT) mixtures to compare the performance of hydrated lime (HL) to liquid antistripping additives (LAS) and to mixtures containing no additives. The mixtures were subjected to tests that make up the backbone of the AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). The results were evaluated using the life-cycle analysis contained in the guide for a 20-year pavement design. The life-cycle cost analysis (LCCA) results demonstrated that HL performed best in all cases, improving some of the mixtures to a great extent.
The principal investigators for the study were Dr. Peter Sebaaly from the University of Nevada–Reno and Dr. Dallas Little from the Texas Transportation Institute. Five state DOTs from around the country agreed to participate in the study by providing aggregates, binders, liquid antistrips and mix designs that were commonly used in their states. HL, since it is manufactured to conform to national standards, was provided from a single source. The states also participated along with other asphalt pavement experts on a technical advisory committee to review and comment on the experimental design and results at critical junctures during the study. The states that participated were Alabama, California, Illinois, South Carolina and Texas.
The mixtures were all dense-graded and designed according to Superpave criteria for 3-10 million equivalent single axle loads (ESALs). The tests included dynamic modulus (E*) to assess moisture sensitivity and to contribute to the evaluations of fatigue and permanent deformation; flexural beam fatigue test and dynamic mechanical analysis (DMA) for fatigue cracking, repeated load triaxial test (RLT) for permanent deformation and thermal stress restrained specimen test (TSRST) for thermal cracking. In addition to being characterized to confirm their Superpave properties, portions of the binders were subjected to long-term aging to observe changes in their viscoelastic properties. Aged binders also were used in samples subjected to fatigue and thermal stress testing.
Finally, the dynamic modulus, flexural beam fatigue and repeated load triaxial test results were inputted into the MEPDG life-cycle cost analysis module to compare the costs and benefits of different additive selections. The results of those analyses demonstrated that the addition of hydrated lime improved the life-cycle costs of all five mixtures, some more dramatically than others.
The sensitive type
For decades, reducing moisture sensitivity (i.e., stripping) has been a major focus of attention to improve the performance of asphalt pavements. The reason for that attention lies in the fact that stripping contributes to all other pavement distresses, whether they are adhesive or cohesive in nature. Unfortunately, no completely satisfactory test exists to evaluate stripping. The most commonly used test, AASHTO T283, is empirical and can be subject to considerable variation among laboratories. However, for screening and comparing materials, the test can provide useful information. For this study AASHTO T283, incorporating one freeze-thaw cycle, was used to provide a ranking of the different aggregates. As recommended by Superpave, a tensile strength ratio (TSR) of 80% was adopted as the threshold for a satisfactory mixture. The results identified two poorly performing aggregates (South Carolina and Texas), one intermediate performer (California) and two good performers (Alabama and Illinois).
For the remaining tests, repeated freeze-thaw cycles were employed to accelerate the rate of damage to the samples. Repeated freeze-thaw cycles are not intended to simulate regional climatic cycles, but rather to accelerate the rate of damage, providing a picture of how pavements will perform over their design lives.
Dynamic modulus, which measures changes in pavement stiffness under different loading and temperature conditions, also provides a good measure of moisture sensitivity. To accelerate the damage, which leads to loss of strength, samples were vacuum-saturated with water to a 70% level. They were then frozen, thawed and tested at room temperature. The expansion of the water during freezing causes damage to the sample. Successive freeze-thaw cycles further increase the damage, which approximates the distress that pavements experience over their life cycles.
Samples from each state containing no additive, HL and LAS were subjected to 15 freeze-thaw cycles with test measurements taken every three cycles. Over the course of the tests the strength properties of the untreated and liquid-treated samples diminished at a more rapid rate than did samples containing lime. The data showed that the addition of lime significantly enhanced the pavements’ resistance to moisture damage over time.
Fighting off plastic Permanent deformation is a distress that occurs early in a pavement’s life, while its binder viscosity is relatively low, and before it ages and oxidizes. It also is most often seen during the summer when pavement temperatures are high.
The RLT applies stress to the samples at three increasing temperatures to measure plastic (nonrecoverable) deflections. The greater the plastic deflections, the more likely a pavement is to rut. Data was collected from each of the samples before freeze-thaw conditioning and after six freeze-thaw cycles.
The results were mixed, with lime generally improving the rutting resistance of the samples made from the worst aggregates. The balance of the results showed similar behavior for the remaining three test cases. The ambiguous results were clarified later in the study when dynamic modulus was included with the RLT results in the life-cycle analysis.
It’s in the DMA
Unlike rutting, fatigue cracking is a problem that commonly occurs later in the life of a pavement, when it becomes oxidized and brittle. This testing employed the flexural beam fatigue test on samples prepared with aged binders. As with the permanent deformation tests, fatigue was evaluated on unconditioned mixes and samples conditioned for six freeze-thaw cycles. Overall the results from the flexural beam fatigue test were ambiguous and required an additional mechanistic analysis to clarify the fatigue performance of the pavements. That additional analysis was included as part of the life-cycle calculations.
The significant changes in the slopes of the liquid-treated mixtures between the unconditioned and conditioned states show that the pavements respond very differently to low and high strain levels. The explanation of that behavior is not straightforward and was revisited in the life-cycle analysis. Fortunately, the study included a second fatigue test, DMA, which helped to clarify the flexural beam fatigue results. The DMA is a new test that looks specifically at the moisture sensitivity and fatigue properties of the fine aggregate and mastic when samples are subjected to torsion. Many pavement engineers believe that the impact of fatigue (as well as moisture damage) is most pronounced in the fine aggregate matrix.
In the DMA test, small cylindrical samples (1?2 x 2 in.) are subjected to a damaging level of strain, and the shear modulus (G*) and dissipated pseudo strain energy (DPSE) are measured. Tests are conducted using aged binder, both unconditioned and moisture conditioned (with no freeze-thaw). The shear modulus indicates the sensitivity of mixtures to moisture damage. When used in combination with DPSE, the shear modulus provides an excellent picture of a mixture’s fatigue resistance. A higher shear modulus lead to lower strains in the asphalt layer, while higher DSPEs indicate a greater capacity to resist cumulative damage.
HL-treated samples generally performed better than the untreated or liquid-treated samples both when unconditioned and after moisture saturation.
Can’t stress enough
The TSRST was used to evaluate the stress that a pavement can withstand before cracking due to shrinkage. Samples are restrained at both ends and placed in a thermal chamber where the temperature is reduced at a controlled rate until the samples break. The stress induced in the sample is monitored throughout the test.
In all but one instance, the HL-treated samples cracked at higher stresses than the untreated or liquid-treated samples. That ability to withstand higher stress is believed to indicate a longer spacing between transverse cracks in the pavement. The longer spacing of cracks results in lower maintenance costs to repair the pavement.
Looking at 20 years
A proven method to compare the long-term performance of design alternatives is to use life-cycle cost analysis. The MEPDG includes a life-cycle cost analysis module that was used to consolidate the results of the key mechanistic-empirical tests so that the three additive choices could be compared. A 20-year life cycle was selected, and the LCCA was restricted to consider only changes in the hot-mix layer. Each state provided information for two highway projects including traffic, environmental and subgrade strength.
The unbound aggregate base layer properties were held constant among the various alternatives, allowing only the asphalt to change. The HMA pavements for the three cases were modeled using the rutting and fatigue results from dynamic modulus, RLT and flexural beam fatigue tests. HL and liquid additives were compared with the untreated mixtures from each state to evaluate their performance. In addition to comparing the thickness of the HMA layers, financial savings were compared based on the following costs:
Unit cost of untreated HMA mix: $5.12/sq yd-in.;
Unit cost of liquid-treated HMA mix: $5.16/sq yd-in.; and
Unit cost of lime-treated HMA mix: $5.39/sq yd-in.
The unit costs for lime treatment included capital costs for investing in silos and metering systems.
The HL and liquid additives were compared with the untreated mixtures to determine the percent saving or percent additional cost over the 20-year life cycle. The costs were for comparison only, since the model only considered rutting and fatigue and was confined to the HMA layer.
Lime aid
It is clear from the combined results of this suite of mechanistic-empirical tests and the analysis included in this comprehensive study that the addition of HL significantly improves pavement performance. The life-cycle cost analyses demonstrate that although lime’s initial cost is somewhat greater than the costs of other additives, the return on that investment is substantial. HL is an important component of high-performance asphalt pavements and is particularly valuable when transportation revenues are inadequate to meet system needs.
