Moisture Damage Modeling in Lime and Chemically Modified Asphalt at Nanolevel Using Ensemble Computational...

Area of Science:

• Materials Science
• Nanotechnology
• Computational Intelligence

Background:

• Asphalt's adhesion and cohesion properties are critical for pavement durability.
• Moisture damage significantly degrades asphalt performance.
• Nanoscale characterization and advanced modeling are needed to understand these phenomena.

Purpose of the Study:

• To measure asphalt adhesion/cohesion forces at the nanoscale using Atomic Force Microscopy (AFM).
• To model asphalt moisture damage using Computational Intelligence (CI) techniques.
• To evaluate the effectiveness of different anti-stripping agents and environmental conditions on asphalt properties.

Main Methods:

• Utilized Atomic Force Microscopy (AFM) to measure nanoscale adhesion/cohesion forces in asphalt.
• Applied Computational Intelligence (CI) techniques, including Artificial Neural Network (ANN), Support Vector Regression (SVR), and Adaptive Neuro Fuzzy Inference System (ANFIS).
• Varied anti-stripping agent content (lime, Unichem), AFM tip properties, and environmental conditions (wet/dry) to create diverse asphalt samples.

Main Results:

• Artificial Neural Network (ANN) demonstrated superior performance in modeling moisture damage for modified asphalt compared to SVR and ANFIS.
• Ensemble methods combining CI techniques with statistical analysis yielded higher accuracy than individual CI models.
• The study successfully modeled adhesion/cohesion forces based on variations in anti-stripping agents, AFM tip parameters, and environmental conditions.

Conclusions:

• CI techniques, particularly ANN, are effective tools for modeling asphalt moisture damage.
• Ensemble CI approaches combined with statistical methods enhance predictive accuracy for asphalt performance.
• Understanding nanoscale forces is crucial for developing durable, moisture-resistant asphalt materials.