Characterizing parameter and model uncertainty in rock engineering

近年来，岩石工程设计的数值方法和获取岩体数据的技术都取得了相当大的进展。随着人口的增长延伸到高山地区和深度超过1公里的大型露天矿井，对地质力学模型和数据收集方法的需求不断增加。越来越多的地下矿山和隧道必须在更大的深度和更复杂的地质和岩体环境中挖掘。在充满挑战的经济时代，工程师需要在相关风险方面加强尽职调查，以优化经济发展。尽管我们的模拟和数据收集能力有了巨大的发展，但由于意外的岩石边坡和矿山故障，我们仍然经常面临安全和经济问题；最近在加拿大发生的一个例子是2008年Porteau Cove岩石边坡的破坏，导致通往2010年惠斯勒奥运会的海天高速公路关闭。岩石工程设计与分析面临的一项基本任务与参数和模型不确定性的影响有关。本研究旨在开发新的方法来表征岩石工程中参数和模型的不确定性。综合岩体表征-数值模拟方法将使用最先进的技术来表征不同地下和地表环境中的岩体。输入数据的不确定性对岩石破坏机制数值模拟的影响将使用数值方法工具箱进行探讨。模型不确定性的影响将使用从简单的2D到3D分析的复杂性范围内的连续、非连续和混合代码的频谱进行研究。脆性断裂建模的重点将探讨不确定性对离散断裂网络和随后的合成岩体方法的影响。该研究将努力将一套不确定性分析工具应用于岩石工程问题，并证明这种方法对提高岩石工程设计信心的重要性。

Recent years have seen considerable advances in both numerical methods for rock engineering design and techniques for acquiring data on rock masses. The demands placed on both geomechanical models and data collection methods continue to increase as population growth extends into high mountain areas and large open pits reach depths of over 1km. Underground mines and tunnels increasingly have to be excavated at greater depths and in more complex geological and rock mass environments. In challenging economic times the engineer is required to optimize economic development with increased due diligence in terms of associated risk. Notwithstanding the enormous developments in our simulation and data collection capabilities we are still all too frequently faced with safety and economic problems due to unexpected rock slope and mine failures; a recent example in Canada was the 2008 Porteau Cove rock slope failure which closed the Sea to Sky highway, the route to the 2010 Olympics at Whistler. A fundamental task faced in design and analysis in rock engineering is associated with the implications of both parameter and model uncertainty. This research endeavours to develop new methodologies for characterizing parameter and model uncertainty in rock engineering. An integrated rock mass characterization-numerical modelling approach will use state-of-the-art techniques to characterize rock masses in varied underground and surface environments. The implications of uncertainty in input data for numerical modelling of failure mechanisms in rock will be explored using a toolbox of numerical methods. The influence of the model uncertainty will be investigated using a spectrum of continuum, discontinuum and hybrid codes ranging in complexity from simple 2D to 3D analyses. An emphasis on brittle fracture modelling will explore the implications of uncertainty on discrete fracture networks and subsequent synthetic rock mass approaches. The proposed research will endeavour to apply a suite of uncertainty analysis tools to rock engineering problems and demonstrate the importance of this approach toward achieving increased confidence in rock engineering design.