When doing a fluid flow simulation, the first step after selecting the analysis type is to select a Turbulence Model.
Before selecting one of the options from the dropdown menu, the first step is to determine whether your simulation occurs in the **laminar** or **turbulent** flow scheme.
So how can this be done?
The choice of a turbulence model depends on the flow conditions, mainly on the Reynolds number (Re):
Re = ρvL/μ
ρ is the density of the fluid (SI units: kg/m3)
v is the maximum flow velocity (SI units: m/s)
L is a characteristic length (depending on the application, it can be hydraulic diameter, geometry length etc) (SI units: m)
μ is the dynamic viscosity ((SI units: Pa·s or N·s/m2 or kg/(m·s))
The Reynolds number is a dimensionless quantity
Typically if the Reynolds number is < 2300 then laminar model is used otherwise a turbulent model is used.
When the material and geometric properties are fixed then the only variable which effects the Reynolds number is inlet velocity. In this case respective model will be used based on the inlet flow velocity.
- For internal flows the characteristic length is equal to the hydraulic diameter of the channel.
- For the external flows usually the length of obstacle in the direction of flow is taken as the characteristic length. For example for flow around a car, the length of car will be used as characteristic length.
To simulate air flow around a truck with given dimensions and velocity
First we will calculate the Reynolds number.
Since this is an external flow simulation so
Characteristic length = 16.6m (length of the truck)
Dynamic viscosity of air = 0.0000186N.s/m^2
Density of air = 1.19198kg/m^3
Inlet velocity = 28m/s
Re = ρvL/μ = 1.19198x28x16.6/0.0000186
Re = 29786683.01 > 2300
So turbulent model will be used.
Going back to the Turbulence Model options on SimScale, you will find both laminar and turbulent models in the dropdown menu. For example, for an incompressible flow simulation:
LES Spalart - Allmaras
For more details about selecting LES or RANS k-epsilon, k-omega, or k-omega SST see: https://www.simscale.com/forum/t/a-discussion-on-turbulence-models/29836