Learning Forum

[hzgour]

Hello all,

I am running a simulation in natural convection where I am adding momentum source terms. In order to check how these terms are added in the simulation, I ran a first case where I defined a gravity term "-rho*9.81" and assigned it to a X-momentum source term, where rho is a polynomial function of temperature. I noticed that I don't obtain the same solution as the case where gravity is assigned by enabling the "Gravity" in "General" and imposing -9.81 m s^-2 in X componant (I am running in 2D-Axisymmetric)

I ran different cases where I defined the density differently, and each time, the only difference is between the cases with gravity as source terme and gravity in the "Gravity" box.

Is the gravity term treated differently when assigned in "General"-->"Gravity" and as a source term? any correction factors added or interpolations?

Thank you

[Karthik Remella]

Hello:
My recommendation would be to enable gravity (and not as a source term). Between this and the material property you pick (variable density), Fluent will automatically add the necessary body force terms for your simulation. There is a lot going on under the hood and I will not be able to go into the details in an open forum.
Karthik

[hzgour]

Thank you for your answer Karthik.
The primary objective of my simulation is to study the influence of a x,y-dependent gravity field on the natural convection. For this; I calculate body forces on matlab (complex integrals) and import them on Fluent as profiles. But since I can't assign a profile to the gravity in the "Gravity" box, I have to add them as source terms. I found out that the difference between the simulations is mainly because of buoyancy effects on turbulence production : when gravity is enabled, these effects can be taken into account. The problem is that the body forces added as momentum source terms wouldn't be taken account of in the turbulence production. Am I right? If so, should I manually calculate the buoyancy source terms of TKE and Dissipation Rate and add them as source terms too ?

Thanks

[Mahdi]

I think for the issue with gravity, when you activate gravity in the "general" section, you should also take a modification in the "operating conditions". You must define an operating density. When you add the operating density, its effect will appear as an effective gravity in the simulation.
Regrading the turbulence, as far as I remember, when you activate the energy equation, in newer versions of Ansys Fluent, the consideration of the buoyancy effects on turbulence generation and dissipation appear in the "turbulence model" part. If you activate "full buoyancy", the effect appears in both generation and dissipation in k-e and k-w model.