
[Sponsors] 
October 8, 2003, 10:57 
Gravity (g) Influence on Simulation Results

#1 
Guest
Posts: n/a

Hi, thank you in advance for your taking time and energy to give me your valuable proposals. The problem is as follows.
When we solve NS equation to obtain the velocity and pressure fields, if gravity g has influence on the final simulation results? If the answer is 'yes', how do we reflect gravity(g)'s influence on the results using Fluent? Or what should we do using Fluent to consider g's influence? Any suggestions are highly appreciated. 

October 8, 2003, 11:35 
Re: Gravity (g) Influence on Simulation Results

#2 
Guest
Posts: n/a

You have to specify gravity acceleration in the Define>Operating conditions panel.
There, you can define along which axis gravity acceleration is aligned, or insert its components along the axes, if it's not aligned with one of them. Hi ap 

October 8, 2003, 13:58 
Re: Gravity (g) Influence on Simulation Results

#3 
Guest
Posts: n/a

Hi, AP,
Thank you for your kind response. The problem I met is the simulation results seem not to change with the gravity levels (g=0, g=2, and g=9.81 m/s2) although I made the same specifications as you suggested. Therefore I wonder if I should specially add g in source term with Fluent. Hope your further suggestions. 

October 8, 2003, 15:57 
Re: Gravity (g) Influence on Simulation Results

#4 
Guest
Posts: n/a

If you use FLUENT internal models you don't need to add a source term for gravity, but you just have to specify gravity acceleration as you have already done.
Which model are you using? Exactly, which parameters don't change in your simulations? Hi ap 

October 8, 2003, 16:43 
Re: Gravity (g) Influence on Simulation Results

#5 
Guest
Posts: n/a

I used a ke turbulent model to calculate a simple channel flow. For this channel flow the inlet and outlet are periodic boundary conditions with constant pressure gradient.
The fluid properties are constant. I changed the gravity levels from g=0 to g=9.81. But it seems the velocity field and pressure field didn't make a difference at all. So I wonder where the problem may be. 

October 9, 2003, 07:37 
Re: Gravity (g) Influence on Simulation Results

#6 
Guest
Posts: n/a

Hi Colin,
what kind of problem are you trying to solve. Do you think that your problem is strongly buoyancy dependent (buoyancy driven flow). If not, there is no reason to make your task more complex than necessary (do not impose gravitational acceleration). Hope it helps. 

October 9, 2003, 08:04 
Re: Gravity (g) Influence on Simulation Results

#7 
Guest
Posts: n/a

I don't know the configuration of the system you are trying to model, but maybe it's not deeply influenced by gravity acceleration.
Anyway, activa Is your channel horizontal or nearly horizontal? Or are there significant variation in quote? Hi ap 

October 9, 2003, 10:09 
Re: Gravity (g) Influence on Simulation Results

#8 
Guest
Posts: n/a

Hi, Vladimir and AP,
Your advices are quite reasonable and lightening. I should announce more clearly when I presented an "asking help" here. Actually the phenomenon studied is a vertical upward flow in a circular tube. The gravity should play a role (although not very important here). But from simulation results, I can not expect this. Today I plan to check again if there is something wrong with Fluent Setting, I found somebody also met the similar problem. Thanks for your taking the time and effort to give me your kind help. 

October 9, 2003, 17:20 
Re: Gravity (g) Influence on Simulation Results

#9 
Guest
Posts: n/a

Hi, ap and Vladimir,
I think I find out where is the problem. This can be said a bug of Fluent. If the gravity is wanted in simulation, the fomer procedure is just as ap said: define>operating conditions>gravity this is not enough. Further work should be done as follows: Specified Operating Density should also be checked, and add Zero to Operating Density. Then the gravity is added. For vertical upward or inclined flow in a tube, gravity should play a kind of role in results. For horizontal flow, it doesn't make a difference. 

October 12, 2003, 22:06 
Re: Gravity (g) Influence on Simulation Results

#10 
Guest
Posts: n/a

hi, colin, I think that you make operating density to zero need think it over.if the problem you simulate occurs in the air(my estimation),and it should be 1.225 or some value else(if not in the air).and i want to know what you do about with the operating density in the former simulations occuring the errors? I did a simulation like yours recently and didn't do the cases you have done and thanks for your making me aware the problem


October 16, 2003, 14:33 
Re: Gravity (g) Influence on Simulation Results

#11 
Guest
Posts: n/a

Hi, Richard,
I would like to suggest you try the following two cases: (1)Horizontal pipe laminar flow (2)Vertical pipe laminar flow For all these two cases, a theoretical pressure drop or velocity profile can be drived. You can make a comparsion between your simulation results with analytical results. Use different operating density in your simulation work, then you would make your own conclusions. 

November 17, 2011, 14:24 
gravity or not in vertical pipe

#12 
New Member
Juan
Join Date: Dec 2010
Posts: 17
Rep Power: 7 
Hi
I am trying to simulate a vertical. I have done a simulation without considering gravity or operating density value. Should I considered gravity value in order to Fluent considered the different heigth between 2 points. I am using water. Should I use any special density? Tnahks 

September 23, 2015, 09:45 
very good!

#13  
New Member
zhouwen
Join Date: Sep 2015
Posts: 1
Rep Power: 0 
Quote:
your answer helps me so much. 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Initializing simulation from previous results help  siw  FLUENT  10  July 12, 2016 15:45 
Unsteady Simulation  Wierd results from Fluent  nikhil  FLUENT  3  October 3, 2011 07:13 
Transient Simulation. Is antialiasing of the results possible?  Darren Leong  ANSYS  0  July 18, 2011 23:39 
Simulation of sloshing by time varying gravity  Manoj Kumar  FLUENT  3  June 13, 2011 03:34 
Batch issues to get results from a transient simulation  aero  ANSYS  0  November 10, 2009 07:32 