CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   "Full buoyancy" option (https://www.cfd-online.com/Forums/main/111795-full-buoyancy-option.html)

 arashsk January 15, 2013 13:26

"Full buoyancy" option

Hello,

I am modelling a room ventilation scenario in FLUENT. I use boussinesq approximation with the k-epsilon standard turbulence model, turning on the "Full buoyancy" option as well. I have tried many strategies to get the solution to converge but I am having difficulties specially when turning on the "Full buoyancy" option.

In any case, I am wondering what the "turning on" of the "Full buoyancy" means. As I read the guidelines and the theory, this includes the buoyancy influence in epsilon-equation in addition to the k-equation in k-epsilon model. I have read the theory guideline but failed to understand the effect or importance of including the buoyancy effect in epsilon-equation in practice.

In theory guide, it talks about the degree in which the epsilon is affected by buoyancy and introduces "C_3_epsilon" which is the tangent hyperbolic of the two components of the velocity (parallel to the gravitational force, and perpendicular to the gravitational vector). It is 1 when the main flow aligns with the direction of gravitation and it is 0 when the main flow is perpendicular.

Obviously I can not obtain a single value (or even an average) to associate a "C_3_epsilon" for my whole domain. It is definitely different in each region. But overall, the room airflow pattern is mainly dominant by the airflow from supply air opening momentum rather than buoyancy. Also in case that the air is supplied at the ceiling level toward the floor, the velocity vectors are mainly aligned with the gravitational vector. That means the local "C_3_epsilon" might be 1 in many regions within the room. But that does not mean that the airflow is dominantly buoyant.

Can anyone help me to understand this contradictory situation and whether it is important to include the buoyancy in epsilon-equation? (Do I need to turn on "Full buoyancy effect" option?)