Incorporating Size Effects into Multiscale Adhesion Modelling of Bitumen-Mineral Interfaces

英国的公路网总长度超过25万英里，为货物和服务的有效分配、经济安全和社会繁荣提供了一种手段。整个公路网的资产价值为7500亿英镑，作为英国的主要交通基础设施，为道路使用者、商业和工业提供重要服务。然而，该网络需要不断升级，维护和修复，预计未来20年需要花费1810亿英镑。这些铺设道路中超过95%是由沥青混合物建造的，其中包括三种主要成分，即矿物骨料（微米至厘米），天然或添加填料（< 63微米）和沥青（膜厚度10-20微米）。然而，尽管其重要性，沥青混合料的劣化从来没有被完全理解或准确预测。主要原因是，目前评估和预测沥青和矿料之间的粘附行为的方法没有考虑到不同材料尺寸尺度下的尺寸效应。这些尺寸效应主要来自沥青膜厚度、矿物表面粗糙度、空隙半径、沥青极性分布（分子尺寸）和矿物组成分布的变化。由于这些尺寸效应的存在，使得沥青混合料的破坏（如材料断裂）无法准确预测（交通荷载引起的疲劳开裂、与非荷载相关的热开裂和与老化相关的开裂）、湿损敏感性（材料崩解和软化、剥离和微动磨损）、坑洞和其他形式的严重表面劣化，本项目旨在开发一个全面的“粘附分析框架（AAF）”，重点是测量和预测沥青（粘结剂）与矿物之间的界面粘附性能。和矿物骨料在沥青混合料中使用的尺寸影响的多尺度实验和建模方法。结合使用实验技术，粘附理论和材料建模方法，将确定尺寸相关和尺寸无关的材料特性，并从纳米到微观到宏观尺度按比例放大，以预测一系列沥青混合料类型的沥青-矿物界面粘合剂剥离特性。该研究将使用显微镜和光谱成像以及纳米级分子动力学（MD）建模的组合来预测沥青-矿物界面粘附力和一系列与尺寸无关的材料特性。粘弹性格里菲斯能量平衡原理，然后将在微观尺度上产生一个基于力学的剥离引发标准，将临界材料的尺寸效应和尺寸无关的材料性能从纳米级MD模拟。然后，通过结合不同材料和尺寸效应以及荷载和环境条件的拉拔附着力和内聚力测试，验证理论沥青-矿物剥离标准。最终的按比例放大效应将处理通过巴黎定律传播模型开发的裂纹（脱粘）传播，该模型将在纳米和微米尺度下确定的尺寸相关和尺寸无关的材料参数结合起来。这些理论预测，然后将实验验证一种新的预制的初始开裂尺寸与材料和条件变量的“裂纹”测试。最后，所有这些不同的多尺度效应将被纳入一个多尺度建模层次结构，用于预测粘合剂失效和整体材料反应，并作为一个基于网络的开源软件和数据库提供。这款用户友好型软件将用于设计和生产更好、更持久的沥青材料，以确保这一关键国家资产的长期可持续性。

The UK's road network totals over 250,000 miles of paved roads providing a means for efficient distribution of goods and services, economic security and social prosperity. The entire road network has an asset value of £750 billion and as the UK's main transport infrastructure provides a vital service to road users, commerce and industry. However, the network requires constant upgrading, maintenance and rehabilitation with a predicted spend of £181bn required over the next 20 years.Over 95% of these paved roads are constructed from asphalt mixtures which comprise three principal components, namely, mineral aggregates (microns to centimetres), natural or added filler (< 63 microns) and bitumen (film thickness 10-20 microns). However, in spite of their importance, the deterioration of asphalt mixtures has never been fully understood or accurately predicted. The key reason is that the current means of assessing and predicting adhesive behaviour between the bitumen and the mineral aggregates does not account for size effects at the different material dimensional scales. These size effects result mainly from the variations of the bitumen film thickness, mineral surface roughness, air void radius, bitumen polarity distribution (molecular sizing) and mineral compositional distribution. Neglect of these size effects makes it impossible to accurately predict the asphalt mixture's distresses such as material fracture (traffic load induced fatigue cracking, non-load associated thermal cracking and age related cracking), moisture damage susceptibility (material disintegration and softening, stripping and fretting), potholes and other forms of severe surface deterioration, all of which are directly affected by the bitumen-mineral interfacial adhesive properties.The project aims to develop an overall 'adhesion analysis framework (AAF)' focusing on the measurement and prediction of interfacial adhesive properties between bitumen (binder) and mineral aggregates in asphalt mixtures using a size-affected multiscale experimental and modelling approach. Using a combination of experimental techniques, adhesion theories and material modelling approaches, size-dependent and size-independent material properties will be determined and scaled up from nano to micro to macroscale to predict the bitumen-mineral interface adhesive debonding properties of a range of asphalt mixture types. The research will use a combination of microscopy and spectroscopy imaging and molecular dynamics (MD) modelling at the nanoscale to predict bitumen-mineral interface adhesion and a range of size-independent material properties. The viscoelastic Griffith energy equilibrium principle will then be used at the microscale to produce a mechanics-based debonding initiation criterion incorporating the critical material size effects and the size-independent material properties obtained from the nanoscale MD simulations. The theoretical bitumen-mineral debonding criterion will then be verified by means of pull-off adhesion and cohesion testing incorporating different materials and size effects as well as loading and environmental conditions. The final scaling up effect will deal with crack (debonding) propagation developed through a Paris' law propagation model incorporating both size-dependent and size-independent materials parameters determined at the nano and microscales. These theoretical predictions will then be experimentally verified by a novel 'sandwich-cracking' test with prefabricated initial cracking dimensions together with material and conditioning variables. Finally, all these different multiscale effects will be incorporated into a multiscale modelling hierarchy for predicting adhesive failure and overall material response and delivered as a web-based opensource software and database. This user-friendly software will be used to design and produce better and long-lasting asphalt materials to ensure long-term sustainability of this key national asset.