Numerical simulation of seismic wave propagation: solution verification

岩土地震工程建模的未来将几乎完全基于性能的抗震设计理念，即基于位移或变形。地震荷载作用下岩土结构变形的真实数值模拟是非常复杂的。虽然有许多先进的工具可用，一个主要的问题时，提出的数值分析的结果进行评估是：“我们有多少可以（应该）相信数值模拟？“对数值分析结果的信任是通过验证和确认获得的。验证确定数值模型是否正确地求解了目标数学方程，而验证确定了模型的代表性水平。本文将对FLAC3D程序在地震波传播问题中的应用进行详细而严格的验证。验证将包括评价的数值精度的程序，通过比较其性能的一套数学解决方案，关于弹性动力学和poroelastodynamic问题。数值模拟的关键要素，如空间和时间的离散化，以及初始和边界条件，将与嵌入式固流相互作用的耦合公式一起研究。此外，一个先进的非线性模型新实现的程序将进行评估，比较在OPENSEES程序的各种验证问题。这些计算平台和预期的模型越来越多地用于学术界和先进的工程实践，因此，这项研究是支持进一步研究地震波传播的先进数值模拟的重要一步。该提案的结果还将使人们更好地了解地震波建模过程中的力学作用，使加拿大专业实践能够利用本研究的结果来开发更安全、更可靠、更经济的设计。

The future of modeling in geotechnical earthquake engineering will be based almost exclusively onPerformance-Based Seismic Design philosophy that is based on displacements or deformations. A realisticnumerical simulation of deformation of geo-structures subjected to earthquake loading is very complex. Whilethere are many advanced tool available, one of the main questions raised when results of a numerical analysisare evaluated is: "How much can (should) we trust numerical simulations?" Trust in results obtained fromnumerical analysis is gained through verification and validation. Verification determines if the numericalmodel is correctly being solved for the intended mathematical equation of interest, while validation establishesthe level of representativeness of the model. This proposal will conduct a detailed and critical verification ofthe program FLAC3D for the problem of seismic wave propagation. The verification will consist on evaluatingthe numerical accuracy of the program by comparing its performance to a suite of mathematical solutionsregarding elastodynamic and poroelastodynamic problems. Crucial elements of numerical modeling such asspatial and temporal discretization, and initial and boundary conditions will be investigated together with theembedded coupled formulation of solid-fluid interaction. Furthermore, an advanced nonlinear model newlyimplemented in the program will be evaluated in comparison to the one in OPENSEES program for a variety ofverification problems. These computational platforms and the intended model are being increasingly used inacademia and advanced engineering practice, and for that reason this study is a paramount step towardssupporting further research on advanced numerical simulation of seismic wave propagation. Results from thisproposal will also allow a better understanding of the mechanics at play during modelling of earthquake wavesenabling the Canadian professional practice to use outcomes of this study to develop safer, more reliable, andmore economical designs.