Hello,

Up to recently I have been using COMSOL but have changed to fluent for various reasons. In COMSOL I was using a boundary condition called, "Neutral boundary condition". Their definition of this boundary condition is - "The neutral boundary condition means that no forces act on the fluid." Also in respect to my model, "The neutral condition defines that you assume the computational domain extends to infinity."

My model is a small heater sitting on what can be described as an infinite base with and infinite fluid above it. The small heater is creating a convective flow in the fluid. Basically, I want to model a finite part of the infinite domain - i.e. mainly the convective stream. It was suggested that I model this with walls far away from the heater. In other words the walls are far enough away from the heater and the convection, that it is essentially the same as an infinite domain, because the no-slip boundary is not close enough to interact with the convection stream. I semi-successfully modelled this but I found the walls have still have a large influence. Also, there is the obvious increased computation time due to the increased size of the mesh when modelling my problem in this way.

Does anyone know how to model a "neutral" boundary condition? It seems like it would be commonly used, as part of the boundary condition set within fluent, but I can't find anything. I can post some more info on my problem if needed.

Regards,

MW

It seems that the Far Field BC is the closest to your description

I thought that initially also, but you can't use this boundary condition unless you use ideal-gas law. I use the boussinesq approximation. The boussinesq approximation was used in the COMSOL model also. This part is taken from the fluent help,

"! This boundary condition is only applicable when the density is calculated using the ideal-gas law (see Section 7.2). Using it for other flows is not permitted. To effectively approximate true infinite-extent conditions, you must place the far-field boundary far enough from the object of interest. For example, in lifting airfoil calculations, it is not uncommon for the far-field boundary to be a circle with a radius of 20 chord lengths."

An update,

The neutral boundary condition states that transport by shear stresses is zero across a boundary. It is called "neutral" because it does not put any constraints on the velocity, and it states that there are no interactions across the modelled boundary. In other words, if I change within the "wall" boundary conditions from "no slip" to "specified shear" and define the specified shear in each component direction as 0 Pascals, this will change the total stress tensor to 0 and therefore the total boundary condition equals 0. I believe this is the correct way to define a neutral boundary within fluent. Does anyone have any opinion on this?

The method I suggested in my previous post does not work. Still stuck.

[rishabh_g]

Have you solved the problem because I am solving a similar problem for natural convection in an open cavity