![Forces, coefficients & representative scales
CD =
2FD
ρU2
A
Coefficent of drag
FD = drag force [N]
ρ = fluid density [kg/m3]
U = freestream velocity [m/s]
A = projected area [m2]](https://image.slidesharecdn.com/f9dcf613-599b-498d-8736-e14baf05cb91-150608162410-lva1-app6892/85/johan_malmberg_thesis_pres_final-15-320.jpg)
![Forces, coefficients & representative scales
CD =
2FD
ρU2
A
Coefficent of drag
FD = drag force [N]
ρ = fluid density [kg/m3]
U = freestream velocity [m/s]
A = projected area [m2]
Re =
ρUL
µ
Reynolds number
L = characteristic length [m]
µ = dynamic viscosity [PaŸs]](https://image.slidesharecdn.com/f9dcf613-599b-498d-8736-e14baf05cb91-150608162410-lva1-app6892/85/johan_malmberg_thesis_pres_final-16-320.jpg)
johan_malmberg_thesis_pres_final
- 1. KTH ROYAL INSTITUTE OF TECHNOLOGY Improving Aerodynamic Performance of a Truck a Numerical Based Analysis Master’s thesis presentation by Johan Malmberg, 15th June 2015
- 2. Presentation Outline • Background & Motivation • Bluff body aerodynamics • Forces, coefficients & representative scales • Governing equations & approach • Case description and setup • Results • Conclusions & discussion • Future work
- 3. Background & Motivation • New EU-legislation to improve the aerodynamics of vehicles and their energy efficiency Image from SAE Image from MAN/Krone
- 4. Background & Motivation • Rear flaps and boat-tails • Base drag accounts for 29% of total drag Image from TrailerTail Image from Seattle Pi/Kenworth
- 6. 80 km/hr Background & Motivation -12.5% CD -5% fuel
- 7. 80 km/hr 160 000 km/year Background & Motivation -12.5% CD -5% fuel 5000 EUR/year saved
- 9. Bluff body aerodynamics Streamlined Viscous forces Bluff Pressure forces
- 10. Bluff body aerodynamics Bluff Pressure forces Turbulent flow
- 11. Bluff body aerodynamics Bluff Pressure forces Turbulent flow Stagnation point – zero velocity – maximum pressure
- 12. Bluff body aerodynamics Bluff Pressure forces Turbulent flow Stagnation point – zero velocity – maximum pressure Boundary layer – separation point
- 13. Bluff body aerodynamics Bluff Pressure forces Turbulent flow Stagnation point – zero velocity – maximum pressure Boundary layer – separation point Wake – recirculation bubble – self-similarity von Kármán vortex street Image from NASA
- 14. Bluff body aerodynamics Bluff Pressure forces Turbulent flow Stagnation point – zero velocity – maximum pressure Boundary layer – separation point Wake – recirculation bubble – self-similarity
- 15. Forces, coefficients & representative scales CD = 2FD ρU2 A Coefficent of drag FD = drag force [N] ρ = fluid density [kg/m3] U = freestream velocity [m/s] A = projected area [m2]
- 16. Forces, coefficients & representative scales CD = 2FD ρU2 A Coefficent of drag FD = drag force [N] ρ = fluid density [kg/m3] U = freestream velocity [m/s] A = projected area [m2] Re = ρUL µ Reynolds number L = characteristic length [m] µ = dynamic viscosity [PaŸs]
- 17. Governing equations & approach Navier-Stokes equations For incompressible, Newtonian fluid
- 18. Governing equations & approach Navier-Stokes equations For incompressible, Newtonian fluid
- 19. Governing equations & approach Navier-Stokes equations For incompressible, Newtonian fluid Direct Numerical Simulation (DNS) takes time Number of operations grows as Re3
- 20. Governing equations & approach Navier-Stokes equations For incompressible, Newtonian fluid Direct Numerical Simulation (DNS) takes time Number of operations grows as Re3 Reynolds decomposition
- 21. Governing equations & approach Reynolds Stress Tensor Reynolds decomposition
- 22. Governing equations & approach RANS DES
- 23. Case description and setup Solver • 3D space • Steady-state RANS • Segregated flow • Incompressible (Ma<0.3) • Turbulence model: Eddy-Viscosity SST k-ω • Wall functions to resolve boundary layer
- 24. Case description and setup Ground Transportation System (GTS) 1/8 scale l = 2.4761 m w = 0.3238 m h = 0.4507 m
- 25. Case description and setup Boat-tails Straight 80 cm / 10° Smooth 1 m Truncated wing Round 1.1 m Designers 1st draft Also simulated with suction slots
- 26. Case description and setup Mesh 8.8 million cells
- 27. Case description and setup Mesh Inlet Outlet Floor Symmetry 8.8 million cells U = 91.64 m/s Re = 2 million
- 28. Results: Grid convergence study
- 29. Results: Grid convergence study
- 30. Results: i-Velocity Field XY Baseline
- 31. Results: i-Velocity Field XY Comparison Straight Baseline Smooth Round
- 32. Results: i-Velocity Field XZ Comparison Straight Baseline Smooth Round
- 33. Results: Zero i-Velocity Isosurfaces StraightBaseline RoundSmooth
- 34. Results: Zero i-Velocity Field XY lc/w=1.95, CD=0.325 lc/w=1.79, CD -18.1% lc/w=1.54, CD -21.2% lc/w=1.02, CD -9.2% StraightBaseline RoundSmooth
- 35. Results: Velocity profiles XY StraightBaseline RoundSmooth
- 36. Results: Pressure field XY Straight Baseline Smooth Round
- 37. Results: TKE XY Comparison Straight Baseline Smooth Round
- 38. Results: TKE XZ Comparison Straight Baseline Smooth Round
- 39. Results: TKE profile XY, at X=1w, Z=w/2 StraightBaseline RoundSmooth
- 40. Results: Round tails with suction slots No suction Suction hi Suction mid Suction low XY XZ
- 41. Results: Suction slots velocity profiles XY Hi slot, CD -4%No suction Low slot, CD -1.4%Mid slot, CD +2.8%
- 42. Results: TKE XZ Comparison No suction Suction hi Suction mid Suction low
- 43. Conclusions & discussion • Best case: 21.2% reduction in drag
- 44. Conclusions & discussion • Best case: 21.2% reduction in drag • Designing boat-tails is not intuitive
- 45. Conclusions & discussion • Best case: 21.2% reduction in drag • Designing boat-tails is not intuitive • The round tail does not have a clear separation point and no clear steady-state solution
- 46. Conclusions & discussion • Best case: 21.2% reduction in drag • Designing boat-tails is not intuitive • The round tail does not have a clear separation point and no clear steady-state solution • Suction can decrease drag, but slot location is critical
- 47. Future work • Run simulations with unsteady models, for example DES, to better capture the flow field
- 48. Future work • Run simulations with unsteady models, for example DES, to better capture the flow field • Suction slots could be elaborated on: suction & blowing, active control, plasma actuators etc.
- 49. Future work • Run simulations with unsteady models, for example DES, to better capture the flow field • Suction slots could be elaborated on: suction & blowing, active control, plasma actuators etc. • Use more realistic truck model to make an optimized boat- tail shape
- 50. Future work • Run simulations with unsteady models, for example DES, to better capture the flow field • Suction slots could be elaborated on: suction & blowing, active control, plasma actuators etc. • Use more realistic truck model to make an optimized boat- tail shape • Experiment full size with straight boat-tail on real truck
- 51. KTH ROYAL INSTITUTE OF TECHNOLOGY Thank you!
- 53. Numerical approach Wall treatment Flow near wall depends on viscosity Free flow is inviscid viscous sublayer buffer layer log- layer outer region inner region Wall function blends near wall with outer region