Fatigue failure is a major problem in pavement layers. When using non-conventional aggregate as a replacement for natural stone aggregate, the probability of pavement distress is high. The fatigue behaviour of asphalt concrete with crushed brick aggregate (CBA) reinforced with polypropylene fibre (PPF) was analysed in this study. Two different aggregate types were used with four different PPF contents. The CBA asphalt mixes performed better against fatigue failure when reinforced with PPF. An effort to predict the fatigue life was also carried out. The prediction of the fatigue life of asphalt concrete is a challenging task due to the complexity of variables involved. Two different fatigue prediction models for PPF-reinforced asphalt concrete with CBA were compared. Regression-based predictions and artificial neural network (ANN) model predictions were studied, with models based on the initial strain and the initial dissipated energy. To compare the models, 12 fatigue beam replicas with four different PPF percentages and two different aggregate groups were tested. The ANN models showed better performance when compared with the regression models, where the initial strain and dissipated energy were found to have a strong influence on the fatigue life properties.

中文翻译：

---

聚丙烯改性碎砖沥青混合料的疲劳寿命：分析与预测

疲劳失效是路面层的主要问题。当使用非常规骨料代替天然石材骨料时，路面损坏的可能性很高。本研究分析了聚丙烯纤维（PPF）增强的碎石骨料（CBA）沥青混凝土的疲劳性能。使用了两种不同的聚合类型以及四种不同的PPF含量。当使用PPF增强时，CBA沥青混合料在抗疲劳破坏方面表现更好。还进行了预测疲劳寿命的努力。由于所涉及变量的复杂性，预测沥青混凝土的疲劳寿命是一项艰巨的任务。比较了使用CBA的PPF增强沥青混凝土的两种不同的疲劳预测模型。利用基于初始应变和初始耗散能的模型，研究了基于回归的预测和人工神经网络（ANN）模型的预测。为了比较模型，测试了具有四个不同PPF百分比和两个不同骨料组的12个疲劳梁副本。与回归模型相比，ANN模型表现出更好的性能，在回归模型中，发现初始应变和耗散能量对疲劳寿命特性具有很强的影响。