Towards a new generation of smart long-span bridges with auto-detection systems for nonlinear dynamic response and local deterioration under wind effect

加拿大的主要悬索桥正在接近其使用寿命的终点，因为这些桥建于1920-1940年。桥梁的使用可以通过维护、修理和更换损坏的结构部件来延长。然而，这些结构的复杂性使得很难决定哪些元素必须修复或替换。因此，目前的项目将开发一种结构智能（SI）工具，该工具将在连接到每座桥梁的结构健康监测（SHM）系统时实时检测损伤，并指示可能受影响的结构元件。创新的SI工具将基于HQP生成的用于训练机器学习（ML）算法的巨大损伤和风致疲劳数据库开发。通过调整不同长度悬索桥的多个校准步骤，确保SI Tool的成功应用。校准所需的设计和监测数据由加拿大公共服务和采购部门（PSPC）提供，该部门负责监督加拿大大多数悬索桥的服务。一旦与悬索桥的SHM现场数据进行验证，当与任何类似的悬索桥SHM系统相连接时，SI工具将能够预测未来的损伤和风疲劳。一项新的知识进步是为桥梁结构单元开发一种新的非线性损伤模型，该模型将修改频率传递函数，以考虑每个桥梁结构单元的阻尼和刚度项的非线性损伤减少。采用集成有限条法（Integrated Finite Strip Method， IFSM）， PI具有专业知识，而不是依靠传统几何特性还原的FEM，因此无法真实地模拟损伤。拟议研究计划的短期目标是检测连接土著社区的利亚德河大桥老化的严重损害，并协助PSPC监测和修复/更换决策。长期目标是实现一个集中的自动检测损坏系统，该系统将允许PSPC及其合作伙伴远程访问任何仪表和监控的悬索桥的结构状况，并决定是否修复或更换信号损坏的桥梁结构元件。由19名HQP （6PhD， 7MASc和6UG）组成的多元化和包容性团队将通过与渥太华大学土著事务办公室、黑人健康跨学科中心和科学与工程委员会妇女办公室合作制定的选择标准报名参加该计划，从而尊重HQP队列的隐私和多样性。该项目将受益于机器学习算法专家Dr. Inkpen和SHM桥梁数据解释专家Dr. Lau的跨学科共同监督。因此，HQP将通过参与有计划的活动获得跨学科的技术技能和软技能。

The main suspension bridges in Canada are approaching the end of their life span, as these were built during 1920-1940. The service of a bridge can be extended through maintenance, repairs and replacement of damaged structural elements. However, the complexity of these structures makes it difficult to decide which element must be repaired or replaced. Thus, the current Program will develop a Structural Intelligent (SI) Tool which will detect the damage in real-time when connected to the Structural Health Monitoring (SHM) system of each bridge, also indicating the potentially affected structural elements. The innovative SI Tool will be developed based on a vast damage and wind-induced fatigue database generated by HQP for training Machine Learning (ML) algorithms. The successful application of the SI Tool will be ensured through multiple calibration steps adjusted for suspension bridges of different lengths. Design and monitoring data necessary for this calibration are made available by Public Services and Procurement Canada (PSPC), who oversees the service of most of the suspension bridges in Canada. Once validated with SHM field data of a suspension bridge, the SI Tool will be able to predict future damage and wind-fatigue, when linked to any similar suspension bridge SHM system. A novel advance in knowledge is developing a new non-linear damage model for bridge structural elements, which will modify the frequency transfer functions to account for the non-linear damage reduction in damping and stiffness terms for each structural bridge element. The implementation will be done in Integrated Finite Strip Method (IFSM), where the PI has expertise, instead of FEM which relies on conventional geometric properties reduction, thus cannot realistically simulate damage. The short-term objectives of the proposed Research Program are to detect the critical damage of the aging Liard River Bridge, which connects indigenous communities, and to assist PSPC in monitoring and repairing/replacement decisions. The long-term objective is to implement a centralized auto-detection damage system which will allow PSPC and their partners to access remotely the structural condition of any instrumented and monitored suspension bridge and to decide towards repairing or replacement of the signaled damaged bridge structural elements. A diverse and inclusive team of 19 HQP (6PhD, 7MASc and 6UG) will enroll in this Program through selection criteria developed in collaboration with uOttawa offices of: Indigenous Affairs Office, Interdisciplinary Centre for Black Health and Women in Science and Engineering Committee, such that the privacy and the diversity of the HQP cohort are respected. The proposed Program will benefit of interdisciplinary co-supervisions with Dr. D. Inkpen, expert in ML algorithms and Dr. D. Lau, expert in SHM bridge data interpretation. Thus, the HQP will gain interdisciplinary technical skills and soft skills through participation in planned activities.