1 of 18

AN ACTUATOR LINE METHODOLOGY FOR HIGH-LIFT AIRCRAFT WAKE SIMULATION

S. Bennie, M. Fossati

SU2 Conference 2023. Varenna, 23-25 October

AEROSPACE CENTER - FASTlab

page

2 of 18

Outline

  1. Motivation
  2. Traditional wake vortex analysis strategies
  3. Source based on ”lifting-line” (actuator line) methodology

​

  • Use case: Approach of EETAR12 with continuously varying high-lift devices
      • EETAR12 geometry and high-lift operation
      • Span-wise loading extraction
      • Data-driven Reduced-Order-Model (ROM) via Proper Orthogonal Decomposition (POD)
      • Source term initialization method

​

  • Concluding remarks and further outlook

​

2

AEROSPACE CENTER - FASTlab

page

3 of 18

Wake vortex risk assessment

3

  • Wake vortices are the major driving influence in the categorization of safe aircraft separation distances

​

  • The evolution of vortical structures and the associated pressure variation might lead to mechanical damage to buildings

Courtesy: Halcrow Group Limited Wake Turbulence Study

GOAL: how to simulate LTO cycles in a cost-effective manner while retaining key flow physics of vortices

AEROSPACE CENTER - FASTlab

page

4 of 18

Wake analysis strategies

  • Environmental noise and inclement weather create a challenging environment for vortex observation
  • LIDAR remains limited in its applicability to vortex problems
  • Available data that can be garnered from observation remains limited
  • Advancements in CFD has made airframe simulations accurate and computationally affordable

4

​

  • Full-scale simulation of trailing vortices continue to be cost prohibitive due to the need to simulate the aircraft geometry explicitly

​

  • To date wake simulations are often performed using idealized mathematical representation of the vortex system

​

  • Simulation of vortices allows on demand evaluation of additional fluid variables such as pressure, vorticity, density

AEROSPACE CENTER - FASTlab

page

5 of 18

Data-driven methodology

5

Generate aircraft geometry of interest (high-lift, landing gear…)

Perform near-field high-fidelity CFD simulations (URANS)

Generate solution database of spanwise loading

Create data-driven model capturing the span-wise loading for different high-lift deployment

Perform LES far-field simulation using span-wise loads obtained at each time-step form the data driven ROM

Simulate vortex time evolution and propagation

AEROSPACE CENTER - FASTlab

page

6 of 18

Source term methodology

  • Lifting line (Actuator Line Model, ALM) can be included within the governing fluid transport equations via a residual term
  • Source terms are included in both compressible and incompressible flow solvers
  • Definition of the source force values can be performed prior to simulation via a pre-processing script

6

AEROSPACE CENTER - FASTlab

page

7 of 18

Source term methodology

  • Source terms are coded via creation of a new vortex_force class within the flow_sources
  • Upon initialization of the class, a connectivity table is generated that maps the local SU2 point ID to the nodal ID assigned within the .su2 mesh file
  • Force data is then included on a per timestep basis, using the connectivity table as a means of rapidly assigning the force to each node

​

​

​

​

​

​

7

Volume discretization

SU2

Force pre-processing

Unsteady force tables

Connectivity Table

Local nodal residual source values

Wake

propagation

Advance in time

AEROSPACE CENTER - FASTlab

page

8 of 18

EETAR12 - Geometry

  • Energy-Efficient-Transport Aspect-Ratio 12 aircraft represents typical medium-haul aircraft
  • Possesses leading edge slats and slotted flap geometries
  • Simplification of EETAR12 was performed to improve quality of final volume mesh
  • Unstructured grid generation was performed via GMSH, BL was inserted post generation via a front advancing method

8

AEROSPACE CENTER - FASTlab

page

9 of 18

EETAR12 – Geometry

9

AEROSPACE CENTER - FASTlab

page

10 of 18

Spanwise loading extraction

10

AEROSPACE CENTER - FASTlab

page

11 of 18

POD-based ROM & validation

  • Parameters could be combined into a general deployment setting, due to the linear relationship between each
  • Validation was performed via a Leave-One-Out (LOO) approach
  • Inner points are selectively excluded from training set and using the depleted ROM reconstructions can be compared to the original snapshot

11

AEROSPACE CENTER - FASTlab

page

12 of 18

Use case – approach trajectory

12

  • EETAR12 following a typical approach glidepath of -3 degrees
  • Flaps assumed to deploy from setting 2 to 3 over a period of several seconds
  • Domain length spans ~210m with ~14x106 nodes
  • Structured grid could be used as the wing geometry has been successfully decoupled
  • Minimum cell spacing based upon assumed vortex core radius

AEROSPACE CENTER - FASTlab

page

13 of 18

Farfield time evolution

13

AEROSPACE CENTER - FASTlab

page

14 of 18

Farfield time evolution

14

AEROSPACE CENTER - FASTlab

page

15 of 18

Comparison of source simulation to EETAR12 simulation

15

AEROSPACE CENTER - FASTlab

page

16 of 18

Remarks and future outlook

  • Lifting-line (actuator line) method for vortex simulation successfully captured the physical trends associated with wake vortex generation absent of the parent geometry
  • Addition of the ROM allows for dynamic alterations of the spanwise loading to account for continuously variable high-lift wakes
  • Modularity of the approach allows for a wide range of trajectories to be considered
  • Approach could now be used for further geometries that feature dynamic deployments such as propellers, landing gear or spoilers
  • Addition of a ROM to the ALM method promising with respect to higher fidelity propeller modelling via the addition of variable pitch airfoils

16

AEROSPACE CENTER - FASTlab

page

17 of 18

Physics isn’t the most important thing

- Richard Feynmann

AEROSPACE CENTER - FASTlab

page

18 of 18

Idealised vortex models

18

 

AEROSPACE CENTER - FASTlab

page