Workshop on Regression of Validation Data to an Application Point
The ASME Sub-Committee of Verification, Validation and Uncertainty Quantification on Computational Fluid Dynamics and Heat Transfer is preparing a supplement to the ASME VVUQ 20 Standard. The supplement presents a regression technique to estimate modeling errors at application points where experimental data is not available. The regression is applied to the data obtained from the point wise validation metric described in the ASME VVUQ 20 Standard.
A Workshop to evaluate the performance of the proposed regression technique will be held at ASME Symposium on Verification, Validation and Uncertainty Quantification of 2025. The selected test case is a statistically steady, two-dimensional flow of an incompressible fluid around the Eppler 387 airfoil at a Reynolds number of 3×105. The selected quantities of interest are the lift, drag and pitching moment coefficients and the validation space includes six angles of attack, for which there are experimental data available including uncertainty estimates.
Simulations have been performed with a Reynolds-Averaged Navier-Stokes solver using the SST k-∞ two-equation eddy-viscosity model with and without the local correlation transition model γ-Reθ. Results of the quantities of interest are available for each of the validation space points in five geometrically similar grids.
Participants are requested to perform two exercises:
- Determine the comparison errors Ei and the validation uncertainties Uvaι of the selected quantities of interest applying the ASME V&V20 validation metric to the data available from the experiments and simulations for the six angles of attack of the validation space
- Apply the proposed regression technique to estimate the modeling error at three angles of attack that are not included in the validation space, i.e. the application points.
The results will enable the comparison of the intervals obtained with the proposed regression technique with the point wise procedure, because experimental data are also available for the application points. Furthermore, the existence of results for two different models with significantly different modeling errors will allow to check the performance of the regression technique for different levels of comparison errors.
Deadline for data submission is the 1st of April 2025. Participants are encouraged to present their findings in the Workshop sessions of the ASME Verification and Validation Symposium of 2025. To that end, a "Technical Presentation Only" must be submitted by December 2, 2024.
Complete information about the Workshop is available here. For more information contact vvuq20workshops@gmail.com.
We look forward to seeing you at the next Symposium on Verification, Validation and Uncertainty Quantification in April 9-11, 2025 in College Station TX.
Kevin Dowding, Luís Eça, Nima Fathi, Urmila Ghia,
Joel Peltier and Laura Savoldi