# FEM ConstraintFlowVelocity

Other languages:
Menu location FEM ConstraintFlowVelocity Model → Fluid boundary conditions → Flow velocity boundary condition FEM None - Constraint initial flow velocity
Solvers
Elmer

## Description

Applies a flow velocity as boundary condition to an edge in 2D or to a face in 3D.

## Usage

1. Press the button or select the menu Model → Fluid boundary conditions → Flow velocity boundary condition.
3. Select the target Edges or Faces.
4. Uncheck unspecified to activate the necessary fields for editing.
5. Set the velocity values or (introduced in version 0.21) specify a formula.

## Formulas

It is possible to define a velocity by specifying the velocity profile as formula. In this case the solver sets the velocities at the different positions according to the profile.

To specify for example the velocity profile

${\displaystyle \quad v_{x}(y)=6\left(y-1\right)\left(2-y\right)}$

for ${\displaystyle y\in [1;2]}$ (assuming that e.g. a pipe has the wall at y = 1 m and y = 2 m)

enter this to the Formula field:
 Variable Coordinate 2; Real MATC "6*(tx-1)*(2-tx)"

This code has the following syntax:

• the prefix Variable specifies that the velocity is not a constant but a variable
• the variable to calculate the velocity is Coordinate 2, meaning y
• the velocity values are returned as Real (floating point value)
• MATC is the prefix for the Elmer solver that the following code is a formula
• tx is always the name of the variable in MATC formulas, no matter that tx in our case is actually y

That y will only be in the range ${\displaystyle y\in [1;2]}$ is set because MATC only evaluates the tx range where the result is positive. This behavior is a bit special but has the advantage that one does not need to specify the range manually.

It is also possible to use more than one variable. See as example the definition of rotations in the displacement constraint.

## Notes

• Any vector component that should be the result of the solver must be set as Unspecified.
• If the target face or edge is not aligned with the main Cartesian coordinate system, it is possible to set the option Normal to boundary.
If Normal to boundary is checked, the normal vector to the selected edge or face is X and it will be oriented away from the mesh domain.
For example, if a flow of 20 mm/s of air should enter the domain, then with Normal to boundary one must input -20 mm/s in the Velocity x field.
• For a wall with non-slip condition, set all velocity components to 0.
• For a symmetry condition, set the the flow to (0, Unspecified, Unspecified) if Normal to boundary is checked.