1. INDIAN INSTITUTE OF TECHNOLOGY GANDHINAGAR
33rd AIAA Applied Aerodynamics Conference
22-26 June 2015
Dallas, TX, USA
Computational Assessment of Rainfall Effects on Aircraft Aerodynamic Characteristics
H. Sharma†, S. Vekaria*, S. Harshe*, M. Singh*, M. Damodaran‡ and B.C. Khoo**
INDIAN INSTITUTE OF TECHNOLOGY GANDHINAGAR
Discipline of Mechanical Engineering
INDIA
Temasek Laboratories
National University of Singapore
Singapore
† Graduate Student
* Undergraduate Student
‡ Professor and AIAA Fellow
** Director and Professor
25th June 2015

2. Acknowledgement 1. Temasek Laboratories:
Supported the summer internships of the first 4 authors ( IITGN students) at Temasek
2. Indian Institute of Technology Gandhinagar:
Simulations have been carried out at Also IIT Gandhinagar have supported for travel of the interns.

3. Overview Introduction and Motivation
Mathematical and Computational Modeling
Rainfall Modelling
Computational Domain and Meshing
Flow Conditions
Results and Discussion
Conclusion and Future Work

4. Introduction
Experiments have been conducted at NASA for investigating effect of rainfall
Discrete Phase Modelling (DPM) approach is used for computational modelling by Valentine and Decker
Effect of rainfall is of special interest for low altitude flights
Surveillance UAVs fly at low altitude
Photograph of wing (with cross-section of NACA airfoil) model installed in the NASA Langley 14 by 22 Foot Subsonic Tunnel.

5. Mathematical And Computational Modeling
Navier Stokes Equations
Lagrangian Particle Equations
Schiller-Naumanns Correlation
Schiller, L. and Naumann, A..“A Drag Coefficient Correlation”, VDI Zeitschrift, Vol. 77, pp , 1933.

6. Mathematical And Computational Modeling
Particle Modelling Equations
𝐹 𝐷𝑟𝑎𝑔 𝑈 − 𝑣 𝑝 =𝐹𝑜𝑟𝑐𝑒 𝑝𝑒𝑟 𝑈𝑛𝑖𝑡 𝑚𝑎𝑠𝑠= 1 2 𝜌 𝑣 𝑝 − 𝑈 𝑈 − 𝑣 𝑝 𝐶 𝑑 𝜋 𝑑 𝑝 𝜋 𝑑 𝑝 𝜌 𝑝 = 18𝜇 𝐶 𝑑 𝑅𝑒 𝑝 24 𝜌 𝑝 𝑑 𝑝 2
𝑑 𝑣 𝑝 𝑑𝑡 = 𝑣 𝑝 𝑘+1 − 𝑣 𝑝 𝑘 ∆𝑡
𝑣 𝑝 = 𝑑 𝑥 𝑝 𝑑𝑡 = 𝑥 𝑝 𝑘+1 − 𝑥 𝑝 𝑘 ∆𝑡

7. Raindrop Modelling Rainfall Rate : 1490 mm/hr Injection Grid
Injection Grid Parameters
Injectors per unit length
Mass flow rate at injector
Velocity of Injection
Location of Injection Grid

8. Wall Impingement Model
Six possible outcome after Droplet-Wall interaction
Bai C., and Gosman A.D., "Development of Methodology for Spray Impingement Simulation," SAE Technical Paper , doi: / in Proceedings of Society Automation Engineering (SAE),69–87 (1995)

9. Wall Impingement Model
(a) Wetted Wall
(b) Dry Wall
T12 = Boiling Temperature
T23 = Leidenfrost Temperature
Wc = Critical Weber Number
A = Splash onset coefficient

10. Thin film modelling
Source Terms:
Phase interaction
Source: OpenFOAM

11. Thin film modelling Force due to momentum transfer by particles
Accounted in momentum equation.
Viscous force due to liquid water film.
Air
Uinf
Air
Air –Water Interface: Vt = Uinf
Water-Wall
Interface : Vt = 0
Film
Body

12. OpenFOAM Implementation
reactingParcelFilmFoam rainFallFoam
Removed:
- Combustion Modelling
Break up Models
Evaporation
Utilities :
Injection grid generation code.
Based on length, width, height, AoA, grid points per unit length and rainfall rate generates the injection grid.
Force Calculation:
At wet surface, force on the surface is calculated based in pressure from the Eularian field and shear force from the surface film equation

13. Aerodynamic Configurations Considered
Wing with NACA airfoil as cross-section z x y
Model UAV with wing of NACA 0012 airfoil as cross-section

14. Computational Domain Wing NACA 64-210
Star-CCM+
OpenFOAM A B C D E F H G I
Face ABCD: Inflow Boundary, V : 25 m/s
Face ABEF, CGHD, BFGC: Farfield
Face EFGH: Outflow Boundary, P = Pa
Face AEHD: Symmetry Plane (Star-CCM+)
Farfield (OpenFOAM)
Face I: Velocity = 0 m/s
Farfield Velocity : 25 m/s
Farfield Pressure : Pa
Farfield Temperature : 300K
Farfield Density : 1.18 kg/m3

15. Computational Domain UAV
Star-CCM+
OpenFOAM E F A B H J G D C
Face ABCD: Inflow Boundary, V = 25 m/s
Face ABEF, CGHD, BFGC, AEHD: Farfield
Face EFGH: Outflow Boundary, P = Pa
Face J: Velocity = 0 m/s
Farfield Velocity = 25 m/s
Farfield Pressure = Pa
Farfield Temperature = 300K
Farfield Density = 1.18 kg/m3

16. Mesh Cell Count: Wing NACA 64-210: Star-CCM+: 1.25 million
OpenFOAM: 1.2 million
UAV:
Star-CCM+: 1.9 million :

17. Results and Discussion
Preliminary Computational Assessment on an Aircraft Wing in Star-CCM+.
Flow past an aircraft rectangular wing of NACA with OpenFOAM solver.
Computational modeling of the Heavy rainfall over a UAV using both commercial code Star-CCM+ and open-source software OpenFOAM.

18. Ratio of finest to coarsest mesh
Results and Discussion
Mesh Convergence Studies – NACA
Star-CCM+
OpenFOAM
Order of Convergence
Grid Convergence
Index ,
Ratio of finest to coarsest mesh

19. Results and Discussion
Comparison of Aerodynamic Characteristics for NACA wing in Star-CCM+

20. Results and Discussion
Contours for with rain and without rain case for NACA wing in Star-CCM+
Computed Coefficient of Pressure contours
Without Rain
With Rain

21. Particle Trajectories
Results and Discussion
Splashing for NACA wing in Star-CCM+
Particle Trajectories

22. Results and Discussion
Comparison of Aerodynamic Characteristics for NACA wing using OpenFOAM

23. Results and Discussion
Contours for with rain and without rain case for NACA wing in OpenFOAM
Computed Wall Film Thickness contours
Splashing
With Rain

24. Results and Discussion

25. Results and Discussion
Comparison of Aerodynamic Characteristics for UAV using Star-CCM+

26. (at 0.1m in spanwise direction)
Results and Discussion
Comparison of Contours for with rain and without rain case for UAV using Star-CCM+
Without Rain
With Rain
Computed Contours(4 AoA)
Velocity
(at mid plane)
Velocity
(at 0.1m in spanwise direction)
Vorticity

27. (at 0.1m in spanwise direction)
Results and Discussion
Comparison of Contours for with rain and without rain case for UAV using OpenFOAM
Without Rain
With Rain
Computed Contours(4 AoA)
Velocity
(at mid plane)
Velocity
(at 0.1m in spanwise direction)
Vorticity

28. Results and Discussion
Comparison of Aerodynamic Characteristics for Star-CCM+ and OpenFOAM

29. Results and Discussion
Contours of wall film thickness on surface of a UAV
Star-CCM+
OpenFOAM

30. Results and Discussion

31. Results and Discussion

32. Conclusion and Future Work
Present computational study has show a potential for using open source software OpenFOAM for high fidelity study for complex computational problems
The results computed using OpenFOAM have shown expected aerodynamic behavior which is in general agreement with available experimental data as well as match closely to the data computed from the commercial software Star-CCM+
Results predicts higher percentage of performance degradation at lower AoA which matches with earlier data
The current work have been resulted in a dedicated solver to solve rainfall problem in OpenFOAM

33. Conclusion and Future Work
Novel impingement, wall film and separation module which can work with one equation SA turbulent model will help in predicting early stall as shown by experimental data obtained at NASA.
Experiments will be carried out in future to verify the computational data and making more robust models to predict the performance more accurately.
A flight dynamics coupling to define the unsteady aerodynamic characteristics of a propeller powered UAV will be carried out as part of Future Work

34. Thank You
Questions !!!

35. Backup Slide -1

36. Backup Slide - 2 Pre-processing (Mesh Generation):
snappyHexMesh ( OpenFOAM inbuilt mesher )
Gmsh ( 3D Unstructured mesher)
Solver :
OpenFOAM
Post-Processing:
Paraview

37. Backup Slide - 3