For fluid flow scenarios where solid-fluid interfaces are present, boundary layer effects can be critical to accuracy of results meaning accurate modelling of the interfaces is imperative. The two methods typically applied in CFD include applying a wall model or fully resolving wall.
However, in pursuit of observing the boundary layer effect, an important parameter must first be explained. This parameter is the y+, which is a non-dimensional wall distance for a wall-bounded flow that characterizes the flow regime near the wall. This is important in observing turbulence due to existing solids in the domain or if there is flow separation in the domain. The y+ value will determine whether or not your wall modelling strategy is valid. For full resolution the y+ value should be kept below 1, ensuring that and when using a wall model the first cell center is placed in the viscous sub-layer. When using a wall model the y+ value should be kept between 30 and 300.
In choosing which method to use, there are several considerations to take:
- Full resolution requires extremely fine mesh near the wall in order to achieve a y+ value less than 1, meaning large mesh sizes will be required. If a large mesh size can be used for the simulation then a full wall resolution can be used, on the other hand if only a small mesh size can be implemented for the simulation, then a wall model must be chosen to simulate the boundary layer effects.
- Complexity of the expected boundary layer effects should also be considered. Wall models can only capture basic behavior. If for example, there is significant surface curvature the flow separation and reattachment may be to complex for the wall model to capture.
The calculation of y+ can be turned on in the SimScale platform by enabling “Turbulence” in the “Field calculations” settings under “Result control“.