Challenges of Bridge Cable Modelling

1
Challenges of Bridge Cable Modelling

• Neil Botterill
• PhD Student, School of Civil Engineering
• January 2006

2
Background
3
Motivation

• Tacoma Narrows Bridge Collapse 1941
• University of Nottingham pursuing a body of work
  based around the Kessock Bridge using mostly CFD

4
Work done so far

• Bridge Deck CFD models Dongke Sun, Phrain Gu
• Cable CFD modelsKai Fan Liaw, Neil Botterill
• Field measurementsJohn Owen
• Wind tunnel experiments

5
Cable Modelling Applications
6
Cable Aeroelastic Behaviour

• Vortex Induced Vibration
• Galloping
• Buffeting
• Rain-Wind Vibration
• Can the above be reproduced in CFD models?

7
Major Considerations

• Structural problem as well as a fluid problem
• Fluid-Structure Interaction (FSI) coupling
  procedure
• Meshing strategy
• Turbulence model

8
Meshing Strategy

• Usually Eulerian mesh is used
• Arbitrary Lagrangian Eulerian (ALE) formulation
  used to account for movement of bluff object

9
Turbulence

• What effects are important to capture?
• Zdravkovich proposes the categories
• Laminar
• Transition in Wake
• Transition in Shear Layers
• Transition in Boundary Layers
• Full Turbulence
• Drag Crisis is important

10
Drag Crisis

• Significant drop in drag in a range of Reynolds
  numbers easily achievable in the real world
• Nobody has successfully modelled this yet
  although some have come close (Wang et al. 2001)
• What turbulence model is suitable for this?

11
Large Eddy Simulation

• LES calculates large eddies (turbulence)
• LES only uses a turbulence model for the smallest
  eddies
• Separation is achieved by filter function
• Usually grid based so this introduces extra
  considerations for meshing
• Large resource requirements

12
LES Work done so far
Reynolds Number Grid Size (no. of nodes) Number of parallel clusters Wall clock time (per 60s of solution time)
16x103 1,533,525 6 7.95 days
32x103 1,533,525 12 7.09 days
64x103 2,224,323 12 17.22 days
128x103 2,238,816 10 21.37 days
13
Drag crisis the story so far
14
Summary of Goals / Challenges

• Continue pursuit of Drag Crisis
• Develop a suitable FSI coupling procedure
• Adopt a suitable meshing strategy for large
  motion
• Extend 2d work on circular cylinders to a 3d
  cable
• Parametric study

15
Thank You For Listening

• Neil.Botterill_at_nottingham.ac.uk
• January 2006