ABSTRACT

Automated truck platooning may cause up to 85% reduction in truck following distance compared to conventional vehicles and therefore reduces rest periods between loading cycles on pavement materials. Fatigue cracking is one of the major distresses in asphalt pavement structures, and pavement fatigue life is highly dependent on rest periods between loading cycles. In this paper, Artificial Neural Network (ANN) modeling was applied on the existing NCHRP 09-44A project’s database of extensive laboratory beam fatigue testing results on asphalt mixtures with various rest periods. The developed ANN model was used to predict number of cycles to failure as a function of rest periods and to estimate the impact of truck platooning on pavement fatigue life. In addition, a stand-alone linear multivariate regression equation of fatigue life was developed independently from the ANN model. Based on the results, an 85% reduction in the following distance of platooned trucks may lead to between 7% and 25% reduction in pavement fatigue life. The Platooning Fatigue Life Ratio (PFLR) was found to be dependent on temperature, applied strain level, and mixture parameters. Finally, the applied strain level was the most significant testing factor and binder grade was the most significant mixture parameter on PFLR.

摘要

与传统车辆相比，自动卡车编队行驶可使卡车跟车距离减少多达 85%，因此减少了路面材料装载循环之间的休息时间。疲劳开裂是沥青路面结构的主要破坏之一，路面疲劳寿命高度依赖于加载循环之间的休息时间。在本文中，人工神经网络 (ANN) 建模应用于现有的 NCHRP 09-44A 项目数据库，该数据库包含对具有不同休息期的沥青混合料进行广泛的实验室梁疲劳测试结果。开发的 ANN 模型用于预测失效周期数作为休息时间的函数，并估计卡车列队行驶对路面疲劳寿命的影响。此外，独立于 ANN 模型开发了疲劳寿命的独立线性多元回归方程。根据结果​​，成列卡车的跟随距离减少 85% 可能会导致路面疲劳寿命减少 7% 至 25%。队列行驶疲劳寿命比 (PFLR) 被发现取决于温度、施加的应变水平和混合参数。最后，施加的应变水平是最重要的测试因素，粘合剂等级是 PFLR 上最重要的混合物参数。