A methodology for the development of discrete adjoint solvers using automatic differentiation tools

A methodology for the rapid development of adjoint solvers for computational fluid dynamics (CFD) models is presented. The approach relies on the use of automatic differentiation (AD) tools to almost completely automate the process of development of discrete adjoint solvers. This methodology is used to produce the adjoint code for two distinct 3D CFD solvers: a cell-centred Euler solver running in single-block, single-processor mode and a multi-block, multi-processor, vertex-centred, magneto-hydrodynamics (MHD) solver. Instead of differentiating the entire source code of the CFD solvers using AD, we have applied it selectively to produce code that computes the transpose of the flux Jacobian matrix and the other partial derivatives that are necessary to compute sensitivities using an adjoint method. The discrete adjoint equations are then solved using the Portable, Extensible Toolkit for Scientific Computation (PETSc) library. The selective application of AD is the principal idea of this new methodology, which we call the AD adjoint (ADjoint). The ADjoint approach has the advantages that it is applicable to any set of governing equations and objective functions and that it is completely consistent with the gradients that would be computed by exact numerical differentiation of the original discrete solver. Furthermore, the approach does not require hand differentiation, thus avoiding the long development times typically required to develop discrete adjoint solvers for partial differential equations, as well as the errors that result from the necessary approximations used during the differentiation of complex systems of conservation laws. These advantages come at the cost of increased memory requirements for the discrete adjoint solver. However, given the amount of memory that is typically available in parallel computers and the trends toward larger numbers of multi-core processors, this disadvantage is rather small when compared with the very significant advantages that are demonstrated. The sensitivities of drag and lift coefficients with respect to different parameters obtained using the discrete adjoint solvers show excellent agreement with the benchmark results produced by the complex-step and finite-difference methods. Furthermore, the overall performance of the method is shown to be better than most conventional adjoint approaches for both CFD solvers used.     

A NEW FLUX-CONSERVING NEWTON SCHEME FOR THE NAVIER-STOKES EQUATIONS

SUMMARY

A new numerical method is developed for the two-dimensional, steady Navier-Stokes equations. Using local polynomial expansions to represent the discrete primitive variables on each cell, we construct a scheme which has the following properties: First, the local discrete primitive variables are functional solutions of both the integral and differential forms of the Navier-Stokes equations. Second, fluxes are balanced across cell interfaces using exact functional expressions (to the order of accuracy of the local expansions). No interpolation, flux models, or flux limiters are required. Third, local and global conservation of mass, momentum, and energy are explicitly provided for. Finally, the discrete primitive variables and their derivatives are treated in a unified and consistent manner. All are treated as unknowns to be solved together for simulating the local and global flux conservation.

A general third-order formulation for the steady, compressible Navier-Stokes equations is presented. As a special case, the formulation is applied to incompressible flow, and a Newton's method scheme is developed for the solution of laminar channel flow. H is shown that, at Reynolds numbers of 100, 1000, and 2000, the developing channel flow boundary layer can be accurately resolved using as few as six to ten cells per channel width.     

Kernel polynomial representation for imaginary-time Green's functions in continuous-time quantum Monte Carlo impurity solver

Inspired by the recently proposed Legendre orthogonal polynomial representation for imaginary-time Green s functions G(τ),we develop an alternate and superior representation for G(τ) and implement it in the hybridization expansion continuous-time quantum Monte Carlo impurity solver.This representation is based on the kernel polynomial method,which introduces some integral kernel functions to filter the numerical fluctuations caused by the explicit truncations of polynomial expansion series and can improve the computational precision significantly.As an illustration of the new representation,we re-examine the imaginary-time Green's functions of the single-band Hubbard model in the framework of dynamical mean-field theory.The calculated results suggest that with carefully chosen integral kernel functions,whether the system is metallic or insulating,the Gibbs oscillations found in the previous Legendre orthogonal polynomial representation have been vastly suppressed and remarkable corrections to the measured Green's functions have been obtained.     

Integrating 1D and 2D hydrodynamic,sediment transport model for dam-break flow using finite volume method

The purpose of this study is to set up a dynamically linked 1D and 2D hydrodynamic and sediment transport models for dam break flow.The 1D-2D coupling model solves the generalized shallow water equations,the non-equilibrium sediment transport and bed change equations in a coupled fashion using an explicit finite volume method.It considers interactions among transient flow,strong sediment transport and rapid bed change by including bed change and variable flow density in the flow continuity and momentum equations.An unstructured Quadtree rectangular grid with local refinement is used in the 2D model.The intercell flux is computed by the HLL approximate Riemann solver with shock captured capability for computing the dry-to-wet interface for all models.The effects of pressure and gravity are included in source term in this coupling model which can simplify the computation and eliminate numerical imbalance between source and flux terms.The developed model has been tested against experimental and real-life case of dam-break flow over fix bed and movable bed.The results are compared with analytical solution and measured data with good agreement.The simulation results demonstrate that the coupling model is capable of calculating the flow,erosion and deposition for dam break flows in complicated natural domains.     

用鲁棒Riemann求解器和运动重叠网格计算旋翼粘性绕流

发展了一种基于鲁棒Riemann求解器和运动重叠网格技术计算直升机悬停旋翼流场的方法。基于惯性坐标系,悬停旋翼流场是非定常流场,控制方程为可压缩Reynolds平均Navier-Stoke方程,其对流项采用Roe近似Reimann求解器离散,使用改进的五阶加权基本无振荡格式进行高阶重构,非定常时间推进采用含牛顿型LUSGS子迭代的全隐式双时间步方法。为实施旋转运动和便于捕捉尾迹,计算采用运动重叠网格技术。计算得到的桨叶表面压力分布及桨尖涡涡核位置都与实验结果吻合较好。数值结果表明:所发展方法对桨尖涡具有较高的分辨率,对激波具有较好的捕捉能力,该方法可进一步推广到前飞旋翼粘性绕流的计算。     

一种适用于磁流体力学切向间断的HLLC黎曼算子

磁场的存在使得磁流体力学特征波模不同于流体力学, 因此直接由流体力学HLLC黎曼算子导出的HLLC双中间态在交界的间断面会出现不守恒的问题。通常降级采用HLL单磁场中间态代替HLLC双磁场中间态以实现守恒和计算稳定, 代价是切向间断的模拟精度不足。本文对此进行改进, 在模拟切向间断时仍然保留原有的HLLC双磁场中间态, 同时各守恒量仍然能够满足Toro相容条件; 改进型HLLC算子在间断两侧的磁场分量存在差异, 因此能够更精确还原切向间断面。基于数值测试, 包括一维激波管和切向间断的时变模拟, 以及地球磁层三维数值模拟, 将模拟结果进行对比, 结果表明: 相比于已发展的HLLC算子, 改进型HLLC算子对切向间断具有更好的捕捉精度, 能够达到或接近耗时更多的HLLD算子的模拟精度。     

A continuous adjoint formulation for the computation of topological and surface sensitivities of ducted flows

Topology optimization of fluid dynamic systems is a comparatively young optimal design technique. Its central ingredient is the computation of topological sensitivity maps. Whereas, for finite element solvers, implementations of such sensitivity maps have been accomplished in the past, this study focuses on providing this functionality within a professional finite volume computational fluid dynamics solver. On the basis of a continuous adjoint formulation, we derive the adjoint equations and the boundary conditions for typical cost functions of ducted flows and present first results for two‐ and three‐dimensional geometries. Emphasis is placed on the versatility of our approach with respect to changes in the objective function. We further demonstrate that surface sensitivity maps can also be computed with the implemented functionality and establish their connection with topological sensitivities. Copyright © 2008 John Wiley & Sons, Ltd.     

基于level set的Eulerian-Lagrangian耦合方法及其应用

提出了一种基于水平集的Eulerian-Lagrangian耦合方法,其中Lagrangian方法采用相容显式有限元拉氏方法,Eulerian方法采用基于近似Riemann解的有限体积Eulerian方法,多介质界面处理采用新的水平集和Ghost方法计算. 给出了若干数值算例,包括激波管问题以及金属和气体的运动界面及其大变形问题,并分别与精确解和相容显式有限元拉氏方法的计算结果进行了对比. 数值结果表明,该方法计算结果正确,精度较高,能够准确捕捉物质界面,适用于处理大变形问题.      

Hartmann共振管及超音速雾化喷嘴流场的数值模拟

采用基于Roe解法的有限体积法,对Hartmann共振管中的气体流场进行了数值模拟,研究了当喷嘴轴线处存在针型激励器的情况下流场的振动情况,数值计算的结果与理论和相关的实验结果符合得较好．计算结果表明移除或引入激励器,将会使Hartmann共振管的共振模式发生转换．通过对超音速雾化喷嘴流场的数值模拟,研究了其中Hartmann共振腔和二级共振腔共同作用下的振动现象以及各物理参数对振动的影响,并对喷嘴中气流从亚音速向超音速的转变机理进行了研究．     

适用于等熵流动的交错拉氏Godunov方法

为消除传统单元中心型Godunov方法在求解稀疏波问题时的非物理过热现象,发展一种适用于等熵流动的交错拉氏Godunov方法.主要的特征是采用速度与热力学变量交错分布的形式,避免在单元内进行速度平均,从而消除由于动量平均过程导致的动能耗散.与传统的von Neumann型交错网格方法相比,网格的边界通量由节点处的多维黎曼求解器提供,克服了多维人工粘性选取带来的困难.为减少多维黎曼求解器在求解稀疏波问题时的非物理熵增,给出稀疏波出现的合理判据,从而保证了热力学关系式的满足.数值实验表明：该方法能很好地消除稀疏波的过热现象,同时在求解激波问题时又能保持与传统单元中心型拉氏方法相同的激波捕捉能力.