National Technical University of Athens

School of Mechanical Engineering

Computational Studies on the Cut-Cell, Immersed

Master Thesis
Thanos A. Natsikas
Supervisor: K. C. Giannakoglou

February 2017
2
Immersed Boundaries Method / Cut-Cell Approach

Solid Boundary
Representation

Solid Boundary Locating &

3
 Euler Equations
• Finite-Volumes approach
• Compressible, inviscid flow
• Conservative form of governing equations (momentum, mass, energy):

𝜕𝑈
𝑅𝑃 = 𝑑𝛺 + Φ 𝑑𝑆
𝜕𝑡
𝑘+1
𝛺𝛲 𝑆𝛲𝑘+1
Time terms Inviscid Flux

4
• Time Terms 𝑢𝑤𝑎𝑙𝑙,𝑛
𝜕𝑈 𝜕 𝑘+1
𝑑𝛺 = 𝑈 𝑑𝛺 − 𝑢𝑔𝑟𝑖𝑑 ∙ 𝑛𝑘+1 𝑑𝑠
𝜕𝑡 𝜕𝑡
𝑘+1
𝛺𝛲 𝑘+1
𝛺𝛲 𝑘+1
𝑆𝛲
𝑘+1
𝑘+1 𝑘+1 𝜕𝛺𝛲
2𝑛𝑑 𝑡𝑒𝑟𝑚 ≅ 𝑈𝑤𝑎𝑙𝑙 𝑢wall,𝑛 𝑑𝑠 = 𝑈𝑤𝑎𝑙𝑙
𝜕𝑡
𝑘+1
𝑆𝛲
𝑘+1
𝑈𝑤𝑎𝑙𝑙
= (3𝛺𝛲𝑘+1 − 4𝛺𝛲𝑘 + 𝛺𝛲𝑘−1 ) 𝒖𝒘
2𝛥𝑡 𝒖𝒘,𝒏

Fluidification? Division/Merge?
Solidification?

𝑡𝑘 Intermediate step 𝑡 ∗ 𝑡 𝑘+1

5
Computational Investigations

6
1. Undulating Airfoil
2𝜋 2𝜋
• Undulating Camber Line 𝑦 𝑥, 𝑡 = 𝐴𝑒 𝑎(𝑥−1) 𝑠𝑖𝑛 𝑥− 𝑡
𝜆 Τ
Maximum Rate of
Amplitude Wave Undulation
Lengthwise
Length of Period
growth of
Undulation
Amplitude

7
• Undulating Propulsion Physics
𝟐𝑨𝒇
• Defining factor: 𝐒𝐫 = Wave Propagating inversely of the
𝑼∞
• For hydro-organisms: Sr≈[0.25:0.35] swimming direction γ (=λ/Τ) > 𝑼∞

Eel-Like Locomotion (Anguiliform)

8
Dynamic Meshing
• Most of the mesh remains uninfluenced from the
local-solid boundary transformation
lower re-meshing time

9
 2𝐴𝑓
Investigation of Propulsive Capabilities Sr =
𝑈∞
𝟎. 𝟑 ≤ 𝑺𝒓 ≤ 𝟎. 𝟕𝟓
𝑈∞ ∈ 1.5: 10 m/s
• Wave Length λ=0.63 (eel-like locomotion)
• Sr ⋲[0.3:1.1]…..but must also: λ/Τ (=γ)>𝑈∞
• Friction driven propulsion (here: inviscid flow)
• Mach<0.1, in order to keep undulating frequency
relatively low 𝐒𝐫 = 𝟎. 𝟗𝟓

𝟎. 𝟕𝟏 ≤ 𝑺𝒓 ≤ 𝟏. 𝟎𝟔
U∞ ∈ 15: 20 m/s

10
• Wave Length λ=1.6283 : Aerodynamics>friction (also easier potential 2𝐴𝑓
realisitation) Sr =
𝑈∞
• U∞ = 20 m/s, 𝐴 = 0.4
• Propulsive Efficiency:
𝑇𝑈∞
• 𝜂𝑝 =
𝑇𝑈∞ +𝑃𝑙𝑎𝑡𝑒𝑟𝑎𝑙
𝑡+𝑇
• 𝑃𝑙𝑎𝑡𝑒𝑟𝑎𝑙 = 𝑡
−𝑝𝑛𝑦 𝑉𝑦 𝑑𝑠 𝑑𝑡

11
• 𝜼𝒑𝒓𝒐𝒑 =74%

𝑪𝑻 =0.63

12
• Velocity & TE Vorticity Vector field

13
2. Tandem Cylinders
Steady Problem: 𝚳∞ ≈ 𝟎. 𝟑
• 2 solid boundaries
• Mixing of spheres of influence C𝑝 Distribution

14
2α ) Tandem Oscillating Cylinders
• Α=r=0.2, T≈0.6 sec Aerodynamic Coefficients of Top Cylinder
• δ=0.3:0.7 (min Distance Range)

15
2β ) Tandem Vibrating Cylinders( A<< )
• Α=r/4, T≈0.6 sec
• δ=0.2/0.3

16
3) Radial Stator Flow Simulation
𝑐
• 40 blades: y=αlog(βx) , solidity: 𝜎 = = 1.33
𝑡
• Circular input-output

17
18
 Conclusions
• Confirmed capability of the Cut-Cell/Euler solver to simulate flows round
undulating foils
• Mapping of various undulating foils behavior in aerodynamic environment
• Locating of propulsive undulating modes
• Solver-external geometry files cooperation automated
• Correct flow prediction in cases of close approaching of different solid boundaries
• Local Refinement when flow grows supersonic
• Meet of expectations in radial flows
• Spotting of meshing defects and Correction

19