GPU-accelerated simulations for eVTOL aerodynamic analysis

A computational Fluid Dynamics (CFD) code for steady simulations solves a set of non-linear partial differential equations using an iterative time stepping process, which could follow an explicit or an implicit scheme. On the CPU, the difference between both time stepping methods with respect to stability and performance has been well covered in the literature. However, it has not been extended to consider modern high-performance computing systems such as Graphics Processing Units (GPU). In this work, we first present an implementation of the two time-stepping methods on the GPU, highlighting the different challenges on the programming approach. Then we introduce a classification of basic CFD operations, found on the degree of parallelism they expose, and study the potential of GPU acceleration for every class. The classification provides local speedups of basic operations, which are finally used to compare the performance of both methods on the GPU. The target of this work is to enable an informed-decision on the most efficient combination of hardware and method when facing a new application. Our findings prove, that the choice between explicit and implicit time integration relies mainly on the convergence of explicit solvers and the efficiency of preconditioners on the GPU.

2020

Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2009

The present paper provides an up-to-date survey of the use of large eddy simulation (LES) and sequels for engineering applications related to aerodynamics. Most recent landmark achievements are presented. Two categories of problem may be distinguished whether the location of separation is triggered by the geometry or not. In the first case, LES can be considered as a mature technique and recent hybrid Reynolds-averaged Navier–Stokes (RANS)–LES methods do not allow for a significant increase in terms of geometrical complexity and/or Reynolds number with respect to classical LES. When attached boundary layers have a significant impact on the global flow dynamics, the use of hybrid RANS–LES remains the principal strategy to reduce computational cost compared to LES. Another striking observation is that the level of validation is most of the time restricted to time-averaged global quantities, a detailed analysis of the flow unsteadiness being missing. Therefore, a clear need for detaile...

2015

Current trends in high performance computing include the use of graphics processing units (GPUs) as massivelyparallel co-processors to CPUs in order to accelerate numerical operations common to computational fluid dynamics (CFD) solvers. This paper examines GPU characteristics for various CFD methods and provides examples of current implementations for commercial CFD software. In order to increase adoption of GPUs for commercial CFD, a linear solver library called AmgX was developed by NVIDIA that offers an algebraic multigrid (AMG) solver and other features, with parallelization of all phases of AMG including hierarchy construction and ILU factorization and solve. Examples relevant to CFD practice are investigated in order to demonstrate the use of AmgX in ANSYS Fluent for industry-scale applications. Hardware system configuration is also discussed that examines directions on CFD solver development.

Computer Physics Communications, 2020

In this work we present ZEFR, a GPU accelerated flow solver based around the high-order accurate flux reconstruction (FR) approach. Written in a combination of C++ and CUDA, ZEFR is designed to perform scale resolving simulations within the vicinity of complex geometrical configurations. A unique feature of ZEFR is its support for overset grids; a feature which greatly expands the addressable problem space compared with existing high-order codes. The C++ implementation of FR in a manner which is suitable for modern hardware platforms is described in detail. Additionally, an overview of the input deck used by ZEFR is included. Validation results are presented for a range of steady and unsteady flow problems including Couette flow, the Taylor-Green vortex, and flow around an SD7003 airfoil. Single node performance on a NVIDIA V100 GPU is analyzed where it is shown that all of the kernels in ZEFR attain a high proportion of peak memory bandwidth. Moreover, multi-node performance is also assessed with strong scalability being demonstrated from 60 to 3840 NVIDIA V100 GPUs.

2002

Aerodynamic design and development of civil and military aircraft relies to an increasing extent on Computational Fluid Dynamics (CFD) analysis tools as time goes on. These tools are seen as complementary to wind tunnel tests and flight tests. Computational methods can provide a better insight into specific aerodynamic features. For example they can give a better insight in installation effects of geometry components, such as engine(s), flaps, slats, flap tip devices, bay doors, weapons, fuel tanks and pods on the flow field. CFD tools can be used to assess the aerodynamic characteristics of an aircraft in terms of aerodynamic coefficients such as lift, drag and moments, and to examine the behaviour of aircraft under aerodynamic high loaded conditions. Hybrid grid technology, combining prismatic grid generation near aerodynamic surfaces with automatic tetrahedral volume grid generation, has become an important tool for aerodynamic analysis and design because it allows a higher level...

Orthogonal, structured grids allow flow simulations in simple geometries with high efficiency and accuracy. In contrast, complex and realistic flow problems have traditionally required the use of curvilinear or unstructured meshes, which require large computational costs and reduced accuracy due to limited grid smoothness and orthogonality. In recent years an alternative approach which combines the advantages of simple Cartesian grids with the ability to deal with complex geometries has been developed. In this technique, named the immersed boundary method ), the complex object is immersed in a regular grid and the body effect on the flow is accounted for by prescribing an appropriate body force in the momentum equations in the first computational cell outside the immersed body. This is a de facto grid-free numerical method in the sense that the time-consuming construction of the smooth mesh fitted to the body is avoided.

ASME Turbo Expo 2025 Turbomachinery Technical Conference & Exposition, 2025

The rapid development of the Graphics Processing Units (GPU) enables affordable Large Eddy Simulations (LES) for gas turbine combustion. This work evaluates the accuracy and the computational efficiency of a newly-developed native GPU solver on single and multiple GPUs, develops the optimal GPU numerical formulations for LES, and analyzes the solver's capability to resolve large meshes and complex physics for a gas turbine combustor, which substantially reduces the computational costs whilst ensuring sufficient solution convergence and accuracy. The combustor investigated is the swirl-stabilized, partially-premixed methane/air PRECCINSTA burner, using LES with the WALE sub-grid scale model for turbulence, and the species transport-based Finite Rate model for combustion, combining with the Stiff Chemistry solver. Rich experimental data are available for this combustor, ideal for GPU validation purposes. To optimize the GPU calculations, the Rapid Octree meshing technique has also been utilized in this work, generating high-quality octree cells for the combustor mesh. A detailed mesh sensitivity analysis has been performed to identify the optimal mesh for the present combustor with the GPU solver. This study then compares two different LES numerical settings on their solution accuracy for the same GPU solver, finding that the GPU-dedicated, Optimized LES Numerics offer slightly more accurate results. The GPU solutions have also been validated against the available experimental data. Finally, the speed-up and the scalability performance of the GPU solver have been discussed. In summary, the best practice settings and workflows developed in this work may be helpful to the deployment of this new GPU solver to more complicated industrial gas turbine combustor simulations.

ASME Turbo Expo 2025 Turbomachinery Technical Conference & Exposition, 2025

The rapid development of the Graphics Processing Units (GPU) enables affordable Large Eddy Simulations (LES) for gas turbine combustion. This work evaluates the accuracy and the computational efficiency of a newly-developed native GPU solver on single and multiple GPUs, develops the optimal GPU numerical formulations for LES, and analyzes the solver's capability to resolve large meshes and complex physics for a gas turbine combustor, which substantially reduces the computational costs whilst ensuring sufficient solution accuracy. The combustor investigated is the swirl-stabilized, partially-premixed methane/air PRECCINSTA burner, using LES with the WALE sub-grid scale model for turbulence, and the species transport-based Finite Rate model for combustion, combining with the Stiff Chemistry solver. Rich experimental data are available for this combustor, ideal for GPU validation purposes. To optimize the GPU calculations, the Rapid Octree meshing technique has also been utilized in this work, generating high-quality octree cells for the combustor mesh. A detailed mesh sensitivity analysis has been performed to identify the optimal mesh for the present combustor with the GPU solver. This study then compares two different LES numerical settings on their solution accuracy for the same GPU solver, finding that the GPU-dedicated, Optimized LES Numerics offer slightly more accurate results. The GPU solutions have also been validated against the available experimental data. Finally, the speed-up and the scalability performance of the GPU solver have been discussed. In summary, the best practice settings and workflows developed in this work may be helpful to the deployment of this new GPU solver to more complicated industrial gas turbine combustor simulations.

Theoretical and Computational Fluid Dynamics, 2014

We propose and analyze a wall model based on the turbulent boundary layer equations (TBLE) for implicit large-eddy simulation (LES) of high Reynolds number wall-bounded flows in conjunction with a conservative immersed-interface method for mapping complex boundaries onto Cartesian meshes. Both implicit subgrid-scale model and immersed-interface treatment of boundaries offer high computational efficiency for complex flow configurations. The wall model operates directly on the Cartesian computational mesh without the need for a dual boundary-conforming mesh. The combination of wall model and implicit LES is investigated in detail for turbulent channel flow at friction Reynolds numbers from Re τ = 395 up to Re τ = 100,000 on very coarse meshes. The TBLE wall model with implicit LES gives results of better quality than current explicit LES based on eddy viscosity subgrid-scale models with similar wall models. A straightforward formulation of the wall model performs well at moderately large Reynolds numbers. A logarithmic-layer mismatch, observed only at very large Reynolds numbers, is removed by introducing a new structure-based damping function. The performance of the overall approach is assessed for two generic configurations with flow separation: the backward-facing step at Re h = 5,000 and the periodic hill at Re H = 10,595 and Re H = 37,000 on very coarse meshes. The results confirm the observations made for the channel flow with respect to the good prediction quality and indicate that the combination of implicit LES, immersed-interface method, and TBLE-based wall modeling is a viable approach for simulating complex aerodynamic flows at high Reynolds numbers. They also reflect the limitations of TBLE-based wall models.