Innovative Friction Surfaces at the NCAT Test Track

Since 2000, NCAT has partnered with the Alabama Department of Transportation (ALDOT) to evaluate asphalt surface friction at the NCAT Test Track using locked-wheel friction testing (LWFT, Figure 1) and texture testing using several methods. Over this 25-year span, hundreds of surfaces have been tested, yielding tens of thousands of data points related to friction performance. These results have been instrumental in reducing road departure accidents, identifying better aggregate classification methods, and advancing sustainability by optimizing the use of local aggregates for friction performance.

Figure 1: ALDOT Friction Testing Trailer at NCAT Test Track

This article summarizes key findings from the NCAT Test Track related to the impacts of material selection and mix design on friction. Key insights include the effectiveness of using "just enough" high-quality aggregates and special materials like calcined bauxite in asphalt mixes to achieve sustained friction results. While some findings have been adopted by sponsoring agencies, others require further investigation. Certain unexpected results have led to further research into friction mechanisms.

ALDOT has conducted LWFT testing with a ribbed tire (SN40R) on each section of the NCAT Test Track monthly since 2000. The original Track included numerous sections with both fine-graded and coarse-graded 9.5 mm and 12.5 mm mixes. Although the primary objective of these test sections was not friction-related, the differences in materials, mix type, and gradation provided an opportunity to investigate their effects on friction. The analysis revealed that the ribbed tire LWFT results for coarse and fine gradations were very similar for mixes with the same aggregate types. These findings helped provide confidence to agencies regarding the use of fine-graded mixes in highway applications.

In 2003, Mississippi DOT sponsored a test section with a thinlay containing 70% limestone from a known polish-susceptible source and 30% hard angular gravel and gravel sand, shown in Figure 2. This mix would not typically meet many state specifications due to its high carbonate aggregate content. Yet after 22 years and 80 million ESALs at the NCAT Test Track, the section still holds with friction numbers (SN40R) between 30 and 35. This showed that using just enough of the right-sized friction aggregate can deliver adequate performance, paving the way for further research on optimizing mix friction through targeted aggregate use.

Figure 2: 2003 70% Limestone / 30% Gravel Sand Surface

The concept of using "just enough" higher-quality friction aggregate in a blend with predominantly carbonate aggregates was reaffirmed in 2021.ALDOT sponsored two thinlay preservation sections: a dense-graded thinlay and a 4.75 mm SMA. The dense-graded thinlay contained 58% limestone, 22% gravel sand, and 20% RAP, while the SMA mix contained 62% limestone, 13% granite, 5% fly ash, and 20% RAP. The experiment was aimed at exploring pavement preservation options for high-traffic roads. Although the mixtures used mostly polish-susceptible carbonate aggregate, the two surfaces had SN40R results near or above 40 over six years and 20 million ESALs. This is clearly shown in Figure 3. These findings further supported the idea that using "just enough" good aggregate can provide good friction performance, even when most of the blend is polish-susceptible aggregate.

Figure 3: SN40R Performance for ALDOT Thin Lift Mixes with Predominately Carbonate Aggregates

Seeking to develop asphalt mixes with sustainable, long-term enhanced friction performance, NCAT placed a 4.75 mm NMAS SMA using 40% calcined bauxite and 59% granite in 2017, pictured in Figure 4. Calcined bauxite, shown in Figure 5, is predominantly used as a high friction surface treatment (HFST) in highway applications. Contemporary research investigating alternatives to HFST indicated that other aggregates and surface types could not consistently produce long-term friction results comparable to HFST. However, those experiments demonstrated the ability to design “enhanced” friction asphalt surfaces to compete with HFST. Due to their lower macrotexture, smaller NMAS thin lift mixes generally have lower SN40 results with a smooth tire. Increased macrotexture with SMA gradations can mitigate this issue. Blending 40% calcined bauxite into the SMA resulted in SN40R values that increased consistently over one year and approximately 3 million ESALs from the mid-40s to the low 60s, an uncommon behavior for typical asphalt surfaces.

Figure 4: Blend of 40% bauxite in SMA with granite aggregates.

Figure 5: Close up of Calcined Bauxite used in WV Test Section

Over six years and 20 million ESALs, friction performance held steady around 60, highlighting the strong impact of the calcined bauxite aggregate. Similar sections using granite did not show comparable results.

In 2024, the concept of using “just enough” friction aggregate was combined with premium friction aggregate to optimize enhanced friction results. The West Virginia Department of Highways (WVDOH) explored using calcined bauxite in a High-Performance Thin Overlay (HPTO) to provide additional options for friction surfaces to improve safety on WV highways. Blends developed for this experiment added up to 40% calcined bauxite, targeting laboratory dynamic friction testing (DFT) results of at least 0.50. The optimized HPTO design considered friction, rutting resistance, volumetric properties, and cost, included 25% calcined bauxite. Through six months of trafficking on the NCAT Test Track, the WV section has shown increasing friction exhibiting early life performance similar to the 2017 SMA blend.

In conclusion, these examples show that innovations in material selection and mix design can lead to effective, economical improvements in pavement friction. Continued research is needed to better understand friction behavior and evaluate alternative materials. Ongoing NCAT work with the Kentucky Transportation Cabinet, concluding in 2026, aims to strengthen lab-to-field friction correlations and support more reliable friction thresholds in mix design.

Contact Nathan Moore for more information about this article.