AutoCFD5 presentations and recordings

Summary of Workshop Series and Introduction to AutoCFD5 Case 1 and 2

Gary Page & Burkhard Hupertz

Meshing TFG Context and Summary

Chair: Vangelis Skaperdas

Effect of Turbulence Modeling and Automatic Mesh Refinement on the External Aerodynamics of Automobiles

CNRS / Centrale Nantes

Additional Mesh Coarsening Approaches and Their Effects

Multimatic

Automated Parallel Hex-Dominant Volume Mesh Generation for the DrivAer Notchback Using ChopMesh

Voxshell

Scale Resolving Simulation TFG Context and Summary

Chair: Charlie Mockett

High Order Spectral Element Method DrivAer

Barcelona Supercomputing Center

Effect of time-step and grid refinement on DDES approach on flow topology and separation behaviour for the DrivAer model

Bramble CFD

Evaluation of Hybrid RANS–LES Scale-Resolving Simulations for the DrivAer Notchback Configuration

ENGYS

Towards Fast and Accurate DrivAer Simulations

Flexcompute

Effect of LES model on the onset of separation

KM Turbulenz

Assessment of the open-source CFD code code_saturne on the AutoCFD workshop cases: from steady RANS to hybrid RANS-LES

Simvia / EDF Group

Investigation of the Effect of Targeted Octree Mesh Refinement on WFLES and SBES solutions for the DrivAer Notchback Configuration

Synopsys

The Role of Numerical Scheme Accuracy in WMLES of the Aerodynamics of the Windsor Body

Universitas Mercatorum

Scale-resolving OpenFOAM simulations of the DrivAer case 2 using a novel family of dynamic RANS closures

University of Modena and Reggio Emilia

Volcano ScaLES Results for AutoCFD5 Cases

Volcano Platforms

HPC TFG Context and Summary

Chair: Herbert Owen

GPU-Accelerated Scale Resolving Simulations for AutoCFD5 on Fixed and Adaptive Octrees

Institute for Advanced Study, Beijing Normal-Hong Kong Baptist University

Detailed comparison of performance and quality between GPU and CPU simulations with OpenFOAM

Upstream CFD

Noise Factors TFG Context and Summary

Chair: Burkhard Hupertz

Pacefish's LBM-based GPU-native CFD

Numeric Systems

Quantification of uncertainties from mesh approaches and hardware architecture using PowerFLOW and DrivAer model

Dassault Systèmes

Using Ensemble Averaging to Accelerate Uncertainty Quantification in Automotive Aerodynamics Simulations

Ford Motor Company

Towards Multi-Factor Noise Assessment in Automotive CFD Using Design of Experiments

ICON

Ensemble Averaging Method for Efficient Transient Simulations of External Automotive Aerodynamics

Škoda Auto

Quantifying Noise Factors: Sampling Length and CrossSolver Uncertainty Analysis

Škoda Auto

Revisiting DrivAer Aerodynamics: Updated results integrating Noise Factors and AI

Siemens Digital Industries Software

AI/ML TFG context, datasets and summary

Chairs: Neil Ashton & Astrid Walle

Mistral DrivAerML results using AB-UPT

Mistral

Impact of Input Quantities on Aerodynamic Predictions Using Geometric Deep Learning

Cadence

IngeniNetSim: Fast and Scalable Geometric Deep Learning for DrivAerML Aerodynamic Prediction

IngeniAI

Spatially Smooth Multi-Granularity Conformal Prediction for Reliable Automotive Aerodynamic Surrogate Modeling

KIT

Trustworthy AI Surrogates for DrivAerML: Cross-Architecture Evaluation, Data-Source Ablation, and Calibrated Uncertainty Quantification

Luminary AI

Single-Pass versus Sequential Multi-Field Prediction with the Courant Neural Surrogate for the DrivAer Case

Pasteur Labs

Data-Centric Large Physics Models for Automotive Aerodynamics: From Architecture to Scalable Training Data

PhysicsX

OEM Panel Discussion

TFG Lead Panel Discussion

Summary, Closing Remarks and Future Workshop Discussions

New case proposal: Gary Page