Application of Chimera with Hexahedral Blocks in Solar Meshes Simone Crippa C²A²S²E, Institute of Aerodynamics and Flow Technology, DLR Braunschweig

Transcription

1 Slide 1 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Application of Chimera with Hexahedral Blocks in Solar Meshes Simone Crippa C²A²S²E, Institute of Aerodynamics and Flow Technology, DLR Braunschweig

2 Slide 2 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Overview Introduction AIAA Fourth Drag Prediction Workshop (DPW 4) DPW 4 Results for CRM WBT Grid Convergence at Cl=0.5 SOB Separation Discretization Improvements First AIAA High Lift Prediction Workshop (HiLiftPW 1) Observation from Grid Convergence Study on Trap Wing Discretization Improvements Conclusions

3 Slide 3 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Introduction Motivation for the presented studies originates from lessons learned during past validation exercises with TAU using Solar grids Fourth AIAA Drag Prediction Workshop First AIAA High Lift Prediction Workshop Solar generates unstructured, quad/hexa dominant grids with an advancing front near field, i.e. O type. Chimera was used to locally improve the discretization deficiencies found in Solar grids for steady state computations. Selected case studies were required to assess: Feasibility, performance penalties using TAU with Chimera, and effect of alternative grid generation approaches.

4 Slide 4 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > AIAA Fourth Drag Prediction Workshop (DPW 4) Workshop on June 2009 Test case: NASA Common Research Model (CRM); WBT Cruise configuration of conventional transport aircraft Natural application of Solar due to available philosophy files, which automate the distribution of sources Vassberg et al., Summary of the Fourth AIAA CFD Drag Prediction Workshop, AIAA

5 Slide 5 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > DPW 4 Results Grid Convergence at Cl=0.5 image courtesy: Tinoco et al.

6 Slide 6 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > DPW 4 Results SOB Separation Wing body separation bubble at trailing edge Some participants predicted one, others did not Solar: no separation (SA, SST, RSM) No reported separation ID Turb. Model C D L M N O S T SA EARSM SA SST k ω SA SST k ω SSG/LRR SST k ω Reported Separation ID Turb. Model A B E F H I J P R U V W X 2 4 SST k ω SA SA SA SA SA SA SA SST k ω SA SST k ω k ε SA SA SA image courtesy: Tinoco et al.

7 Slide 7 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT SOB Separation Wing body separation bubble at trailing edge... orientation

8 Slide 8 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements Near field layer contraction; feature of Solar adaptive advancing layer method Cell extraction at constant x plane

9 Slide 9 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements TAU computations on Chimera grid Solar medium grid (11.7 mio. points) Hexahedral CH type blocks in wing/body and HTP/body junctions (5.25 & mio. points) Hexahedral HC type blocks around wing and HTP trailing edges (2.39 & mio. Points) Total size mio. points ( blanking interpolation = mio. points)

10 Slide 10 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements

11 Slide 11 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements

12 Slide 12 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements

13 Slide 13 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements H topology in SOB more appropriate to resolve the body and wing boundary layers Cell extraction at constant x plane

14 Slide 14 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements Improved discretization resolves SOB separation Iso surface of x velocity= 10m/s in red, on cf coded wing/body junction. Critical cp in blue Solar Chimera

15 Slide 15 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements Separation bubble triggered by shock BL interaction Shock is correctly resolved with baseline Solar grid, but not its interaction with the BL. Shock at x/croot=73.8% Bubble start at x/croot=68.2% Span wise extension Δy/b/2=0.6% (Δy 6.8'' ; b/2= '')

16 Slide 16 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements At CL=0.5 the discretization improvements are not dramatic in terms of integrated coefficients At off design the discretization improvements are more clear, e.g. at the angle of attack of 4

17 Slide 17 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements At aoa=4, massive separations dominate the SOB junctions Skin friction lines in wing/body Solar and HTP/body junctions Chimera

18 Slide 18 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > CRM WBT Discretization Improvements At aoa=4, SOB and wake discretization improvements with Chimera result in a significantly different flow field picture Solar Chimera

19 Slide 19 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > First AIAA High Lift Prediction Workshop (HiLiftPW 1) Workshop on June 2010 Focus NASA Trapezoidal Wing (Trap Wing) three element high lift section medium/high aspect ratio Among other tasks, perform grid convergence study to separate discretization from modeling errors New application for Solar grid generation, no available philosophy files Slotnick et al.,overview of the 1st AIAA CFD High Lift Prediction Workshop, AIAA

20 Slide 20 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > HiLiftPW 1 Observation from Grid Convergence Study Grid independence not achieved in terms of flap/body separation, over entire angle of attack range coarse Significant discretization errors medium fine case 1; aoa=13 ; flap/body junction separation topology changes between the grid levels Crippa et al.,dlr Contribution to the First High Lift Prediction Workshop, AIAA

21 Slide 21 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > HiLiftPW 1 Solar Grid Generation Peculiarity Solar automatically decreases local near field expansion ratio from target value in concave surface intersections config. 1 with brackets

22 Slide 22 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar grid shows major near field layer contraction, which also leads to poor shear layer resolution of upstream elements Possible discretization improvements are evaluated only for the angle of attack of 13 in respect to the SOB resolution Solar Sample structured grid

23 Slide 23 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements TAU computations on Chimera grid Solar medium grid with brackets (39.7 mio. points) Hexahedral box extracted from HiLiftPW 1 participant provided grid (JAXA, Gridgen, Str OnetoOne E v1, medium level) Mixed CH/O type resolution in slat, main, flap to body junctions (5.01mio. points) Total size mio. points ( blanking interpolation = mio. points)

24 Slide 24 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements

25 Slide 25 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Skin friction lines on the flap reveal a similar SOB separation on both grids Differences appear also on the body pod and in the main element SOB Solar Chimera

26 Slide 26 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Skin friction lines show a different flow topology starting at the main element leading edge Solar Chimera

27 Slide 27 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Horse shoe vortex from slat/body Horse shoe vortex from main/body Solar ¼ cmac field cut Chimera

28 Slide 28 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Horse shoe vortex from slat/body Wake of first slat support bracket Horse shoe vortex from main/body Solar Flap trailing edge normal field cut 16 Chimera

29 Slide 29 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 12 Chimera

30 Slide 30 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 8 Chimera

31 Slide 31 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 4 Chimera

32 Slide 32 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Flap/body SOB separation Solar Flap trailing edge normal field cut Chimera

33 Slide 33 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 4 Chimera

34 Slide 34 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 4 Chimera

35 Slide 35 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 4 Chimera

36 Slide 36 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Trap Wing Discretization Improvements Solar Flap trailing edge normal field cut 4 Chimera

37 Slide 37 > Application of Chimera with Hexahedral Blocks in Solar Meshes > S. Crippa > Conclusions Solar grids lack sufficient discretization in critical areas Concave junctions Wakes of thin lifting surfaces Chimera grids offer the flexibility to enhance poor discretization regions at the cost of Higher computational effort: wasted points being blanked out or used for block to block interpolation (CRM: 28%, trap wing: 7%) poor solver performance (RK, CFL < 1) Chimera Search Time Level t Increased problem setup time: FILE 'chimera_search.c', LINE 1501: WARNING: in total xxx points not found for all domains No automatic hole definition capabilities No generic component level grid generation capabilities Accuracy? TAU Chimera is not flux conservative! Unstructured solvers are able to deliver flexibility on matching grids, now it is up to the unstructured grid generation algorithms to deliver appropriate grid topologies