FENST-2D

• Velocity
• Pressure
• k & ε
• SPM concentration
• (free surface elevation)

• Hybrid mixture/2-phase flow
• Suspension mass balance (continuity)
• Suspension momentum balance (RANS)
• Viscosity closure (suspension rheology)
• Turbulence closure (customized k-ε model, incl. low-Reynolds)
• SPM mass balance (unsteady adv.-diff. eq.)
• Particle/fluid momentum exchange closure (simplified particle force balance)
• Roughness closure
• Optional: Free surface eq. (kinematic condition)

traditional boundary conditions for RANS solvers

Measured data:

• Particle properties

For validation:

• profiles of velocity
• SPM concentration
• Reynolds stress
• TKE / free surface elevations
• pressures

Model generated data:

• DNS data for validation
• Mixing-length theory profiles for initialization

• Horizontal: carthesian
• Vertical: carthesian

• Fixed or moving (ALE method)
• Mixed structured/unstructured grid (block-structured)
• Arbitrary angles (but the more deviating from orthogonality, the worse the performance)

• Second order acurate in space
• Minimal numerical diffusion

• Slow solver --> no 3D version, small problems
• No higher order time schemes available (2nd order has bugs)
• Not very robust (to minimize numerical diffusion effects)
• Continuously under development, unsuitable for external users (poorly documented)

Parameters:

1. Fluid density
2. Particle density
3. Settling velocity
4. Nikuradse roughness
5. Stabilization parameters (AD, pseudo Δt)

Procedure

• 1-2: usually known
• 3-4: start from known, modify if necessary
• 5: trial & error

Procedure

• Comparison on profiles with measurements

Validation test cases available

• Flume experiments
• DNS data