Quantitative Image-Based Characterisation of Moisture-Induced Asphalt Stripping: Method Validation and Material Effects

Roads are built to last, but water is a relentless enemy. When rain or melting snow seeps into the pavement, it can quietly weaken the bond between the sticky asphalt binder and the hard stones, or aggregates, that hold the road together. Over time, this weakening causes the asphalt to peel away from the stones, a process engineers call stripping. Once the coating is gone, the road loses its strength, leading to potholes and cracks that shorten its life. For decades, checking how well a road mix resists this water damage has relied on human eyes. A technician would look at a sample of coated stones after soaking them and guess how much of the black asphalt had peeled off. This method is fast and cheap, but it is also subjective; two different people might see the same sample and disagree on how much damage occurred, influenced by lighting, fatigue, or personal judgment.

To solve this problem, researchers Yalcin Karakaya, Ayhan Yucel, and Dilay Uncu set out to replace the human eye with a camera and a computer. They wanted to create a system that could measure the exact amount of asphalt lost from stones with the same precision every time, removing the guesswork from the equation. Their work involved testing how different types of stones, different asphalt formulas, and various water conditions affect this peeling process. By turning a visual inspection into a precise digital measurement, they aimed to understand exactly which materials hold up best against moisture, providing a clearer path to building roads that last longer.

The team began by preparing hundreds of small samples, each consisting of stones coated in asphalt. They used two distinct types of asphalt: a standard, unmodified version and a more advanced type mixed with a polymer called SBS, which makes the binder more elastic and durable. They also tested two different kinds of stone: limestone, which is common and generally holds up well, and basalt, which is harder but can be more prone to moisture issues. To make the experiment thorough, they varied the size of the stones and even changed their shape. Some stones were left with sharp, jagged edges, while others were tumbled in a machine to smooth them into rounded forms. They also tested how the asphalt behaved after being subjected to heat, simulating the aging process that happens when roads sit in the sun for years.

Once the samples were ready, the researchers subjected them to two different water treatments to see how they would react. In one test, the samples sat in water at a warm temperature for up to a week. In the other, they were boiled in water for up to an hour. To mimic real-world winter conditions where roads are treated with salt, some of these water baths contained a four percent salt solution, while others used plain water. After each period of soaking or boiling, the samples were removed, dried for a day, and photographed. This drying step was crucial; it ensured that no water droplets on the surface would confuse the camera, allowing it to see only the actual asphalt and stone.

The core of their innovation lay in how they analyzed these photographs. Instead of a person squinting at a screen, a computer program took over. The software converted the color images into a specific color space that separates the brightness of an object from its actual color. This allowed the computer to distinguish between the dark asphalt and the lighter, exposed stone with high accuracy. Because limestone and basalt have different natural colors, the researchers first taught the computer how to recognize the exposed stone for each specific rock type using a set of calibration images. Once the computer learned these rules, it applied them automatically to every single photo, calculating the exact percentage of the stone surface that had been stripped of its asphalt coating.

To ensure this digital method was trustworthy, the researchers compared its results against a set of images that had been manually marked by a human expert. The computer's measurements matched the human expert's with remarkable precision, differing by less than five percentage points on average. This level of agreement proved that the digital system could reliably replace the subjective human eye. With this validated tool in hand, the team ran their full experiment, analyzing 96 different combinations of materials and conditions to see what factors mattered most.

The results revealed a clear story about what makes roads vulnerable to water. As expected, the longer the samples were exposed to water, the more asphalt peeled away. The presence of salt made the damage worse, and the effect grew stronger the longer the samples sat in the salty solution. However, the most significant factor was the type of asphalt used. The samples coated with the polymer-modified asphalt held up far better than those with the standard binder, showing significantly less stripping across all conditions. This advantage became even more pronounced as the samples aged and were subjected to longer periods of water exposure.

The type of stone also played a major role. The basalt stones consistently lost more of their asphalt coating than the limestone stones, regardless of the water treatment or the asphalt type used. This confirmed that the mineral composition of the stone itself is a critical factor in how well it bonds with asphalt. The size of the stone mattered as well; smaller stones tended to lose more coating than larger ones, likely because the thinner layer of asphalt on smaller stones is more easily washed away. Interestingly, the shape of the stone also influenced the outcome. Stones that had been smoothed into rounded shapes lost more asphalt than those with sharp, angular edges. This suggests that the rough, jagged surface of angular stones provides a better mechanical grip for the asphalt, helping it stay attached even when water tries to push it off.

The researchers also found that the aging of the asphalt binder made a difference. As the binder aged and became stiffer, it was more likely to peel away from the stones. This effect was particularly strong in the standard asphalt, whereas the polymer-modified binder resisted this aging-related damage much better. When comparing the two water tests, the team found that while the boiling water test caused more immediate damage, both methods ranked the materials in the same order. This means that even though the tests are different, they tell a consistent story about which materials are superior.

Ultimately, this study demonstrates that a digital camera and a smart algorithm can provide a precise, objective way to measure road durability. By removing the uncertainty of human judgment, engineers can now compare different road mixes with greater confidence. The findings confirm that using polymer-modified asphalt and choosing stones with the right mineral properties and shapes can significantly extend the life of a road. While the specific numbers apply to the materials tested in this study, the method itself offers a powerful new tool for understanding how water damages our infrastructure, paving the way for smarter, longer-lasting roads.