Drag Prediction on DLR-F6 Wing-Body Configuration

Abstract:

This paper mainly investigate the accuracy of the computed drag on the DLR-F6 Wing-Body configuration, and analyze effect of grid and the turbulence models including the Spalart-Allmaras model, Wilcox’s k-ω model and Menter shear-stress transport model on aerodynamic forces for wing-body configuration. The computed results show that grid refinement has little effect on the pressure distributions, significant effect on drag. The turbulence models have certain effects on the pressure distributions, especially positions of the shock wave. They have obvious effects on drag, particularly friction drag. This study shows that performing the CFD calculation at the same angle-of-attack as experiment resulted in good comparisons with wing surface pressures.

• Periodical: Applied Mechanics and Materials (Volumes 110-116)
• Pages: 1506-1511
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.110-116.1506
• Online since: October 2011
• Authors: Qiu Ya Zheng, San Yang Liu
• Keywords: Drag, Navier-Stokes Equation, Turbulent Model, Wing-Body Configurations

References

[4] and the computing results in [5] are also included in Figs. 1-5. It can been obviously found that grid refinement leads to convergent results for SA model, and our computing drag is better than those using other software when compared with experimental data. Figs. 1-5 show that drag decreases with grid refinement, and the angle of attack and pitching moment increase with grid refinement. Figure 1. Comparison of angle-of-attack versus number of cells to the -2/3 power, CL=0. 5, SA . Figure 2. Comparison of pitching moment versus number of cells to the -2/3 power, CL=0. 5, SA. Figure 3. Comparison of total drag versus number of cells to the -2/3 power, CL=0. 5, SA. Figure 4. Comparison of pressure drag versus number of cells to the -2/3 power, CL=0. 5, SA. Figure 5. Comparison of friction drag versus number of cells to the -2/3 power, CL=0. 5, SA. Table I shows force and moment results for Case 1 along with the experimental data for reference. Total drag is 293. 4, 288. 1 and 286. 6 counts for coarse, medium and fine grid, respectively. The total drag range decreased is 6. 8 counts from the coarse grid to the fine grid. Where the pressure drag range decreased 6. 5 counts, the friction drag decrease 0. 5 counts. The results demonstrated that grid refinement has significant effect on pressure drag at a constant lift. TABLE I. Force and Moment Results for Case 1 Mesh Aerodynamic parameter a/(deg. ) CD CDp CDv CM Coarse.

125.

02934.

01597.

01337 -0. 138 Medium.

165.

02881.

01546.

01335 -0. 143 Fine.

165.

02866.

01532.

01334 -0. 146 Exp.

52.

0295 - - -0. 121 Fig. 6 shows the comparison of the wing surface pressure distributions between coarse grid and fine grid. The computational results are shown at four (y/b=23. 9%, 37. 7%, 51. 4% and 84. 7%) of the eight experimental span locations along with the experimental pressure coefficients for reference. The comparison of the pressure distributions in Fig. 6 shows that the pressures changed very little with grid refinement. The pressure distributions predicted compare well with the experimental values on the inboard span of the wing, but the mid-span of the wing the predicted shock is forward of the experimental data. And at the tip the predicted shock is much weaker and forward of the experimental data. Figure 6. Comparison of the wing surface pressure distributions both coarse grid and fine grid, CL=0. 5, SA. Case 2: Fixed Angle-of-attack Condition The0. 490 case was computed fully turbulent on medium grid using SA, SST and Wilcox model respectively, whose purpose was to assess the effect of turbulence models on drag and pressure distribution. Force and moment results of faxing angle of attack and the experimental data were tabulated in Table II. It can be found that total drag varied from a low of CD = 0. 02965 for SST to a high of CD = 0. 03104 for Wilcox (13. 9 drag count difference), SA was between. Most of the total drag difference was from the friction drag (8 drag count difference). Total drag and friction drag component for SA are close to that of Wilcox (6. 9 drag count difference and 2. 3 drag count difference , respectively), but pressure drag component for SA close to that of SST (2. 1 drag count difference). In general, the computed results using SST model has the highest accuracy compared with the experiment. TABLE II. Force and Moment Results for Case 1 Model Aerodynamic force coefficient CL CD CDp CDv SA.

54691.

03043.

01755.

01288 SST.

53460.

02965.

01734.

01231 SST.

55871.

03104.

01793.

01311 Exp.

4967.

0295 - Fig. 7 show the effect of viscous turbulence model on wing surface pressure distributions using SA, SST and Wilcox. The computational results are shown at four (y/b=23. 9%, 37. 7%, 51. 4% and 84. 7%) of the eight experimental span locations along with the experimental pressure coefficients for reference. Figure 7. Effect of viscous model on surface pressure distribution using SA, SST and Wilcox at a=0. 490. The comparison of the wing surface pressure distributions in Fig. 7 shows that the pressures distributions and shock position predicted by SA and SST have little difference. The pressure distributions predicted by SA, SST and Wilcox turbulence models compare well with the experimental values at a=0. 490, but the shock position predicted by Wilcox is afterward of that of SST. The results in Fig. 7 show that the turbulence models have certain effects on the pressure distributions, particularly on the positions of the shock wave. Surface Pressure Comparison of Matching Lift vs. Matching Angle-of-attack. Comparisons of wing surface pressure distribution between matching lift and matching angle of attack was shown at two span locations (23. 9% and 84. 7%) in Fig. 8, using medium grids and SA turbulence model. The experimental pressure coefficients are also plotted here for reference. Constant lift resulted in poor comparison with experimental surface pressures. On the other hand, running CFD at the same angle of attack as experiment (a= 0. 49) yielded much better comparisons, including correct shock position prediction. The fact is that at a fixed angle of attack (a= 0. 49) CFD predicted wing surface pressures distribution in very good agreement with experiment. Figure 8. Surface pressure Comparison of matching lift vs. matching angle-of-attack, SA, medium grid. Conclusions In the paper, the calculations for two cases were preformed to investigate the effect of grid refinement and turbulence models on forces and moments for WB. The results indicate that grid refinement has little effect on the pressure distributions, has significant effect on drag, especially on the pressure drag. The turbulence models have certain effects on the pressure distributions, especially on the positions of the shock wave. They have obvious effects on drag, especially friction drag. This study demonstrated that performing the CFD calculation at the same angle-of-attack as experiment resulted in good comparisons with wing surface pressures. Acknowledgement The project is supported by China Aviation Science Foundation (NO. 20081431) and The Fundamental Research Foundation for the Central Universities (NO. CHD2010JC101).

[1] Levy, T. Zickuhr, J. Vassbery, S. Agrawal, R. Wahls, S. Pirzadeh, and M. Hemsch. Summary of data from the First AIAA CFD Drag Prediction Workshop. Journal of aircaft, vol. 40, October, 2003, pp.875-882.

DOI: 10.2514/2.6877

[2] M. Hemsch, Statistical Analysis of CFD Solutions from the Drag Prediction Workshop. Journal of aircaft, vol. 41, Deccember, 2004, pp.95-103.

[3] K. Laflin, S. Klausmeyer, T. Zickuhr, J. Vassberg, R. Wahls, J. Morriso, et al. Summary of data from the second AIAA CFD drag prediction workshop. Journal of aircaft, vol. 42, October, 2005, pp.1165-1178.

DOI: 10.2514/1.10771

[4] Rumsey, S. Rivers, and J. Morrison. Study of CFD variation on transport configurations from the second drag prediction workshop[R]. Comuter & Fluids, vol. 34, August, 2005, pp.785-816, doi: 10. 1016/j. compfluid. 2004. 07. 003.

DOI: 10.1016/j.compfluid.2004.07.003

[5] Y. T. Wang, G. X. Wang, and Y. L. Zhang. Drag of DLR-F6 wing-body configuration. Chinese Journal of Computational Physics, vol. 25, April, 2008, pp.145-160.

[6] Q. Y. Zheng, S. Y. Liu, and Y. H. Liang. Numerical simulation of the flowfields around the DLR-F4 wing-body configurations. Chinese Journal of Harbin Engineering University, vol. 31, August, 2010, pp.1029-1033.

[7] Q. Y. Zheng, S. Y. Liu, T. X. Zhou. Transonic drag computation around DLR-F6 wing-body configurations. Chinese Journal of XI'AN JIAOTONG University, vol. 44, September, 2010, pp.115-121.

[8] P. R. Spalart, S. R. Allmaras. A One-equation turbulence model for aerodynamic flows. La Recherche Aerospatiale, vol. 1 1994, pp.5-21.

[9] R. Menter. Two-equation eddy-viscosity turbulence models for engineering applications. AIAA Journal, vol. 32, August, 1994, pp.1598-1605, doi: 10. 2514/3. 12149.

DOI: 10.2514/3.12149

[10] D. C. Wilcox. Re-assessment of the scale-determining equation for advanced turbulence models[J]. AIAA Journal, vol. November, 1988, pp.1299-1310, doi: 10. 2514/3. 10041.

DOI: 10.2514/3.10041