CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Help: Set Up model in CFD with time (

Ranie Go October 3, 2003 02:19

Help: Set Up model in CFD with time
Hello guys! i'm just new here. Just like to ask some help regarding the CFD. My boss want to run a fluid with a specfic temp at a certain period of time (1 hr). THis could be my first analysis with a given time = hr. I've been searching in CFD where i could possibly input a time of an hr but it seemed i cant find it. Does anyone know about transient flow analysis here? pls help me how to apply time in CFD so i could start my analysis. Thanks in advance! Ranie pls reply at <>

Helper. October 3, 2003 05:26

Re: Help: Set Up model in CFD with time
Deatails answers will be fortcoming without details!

P. Birken October 6, 2003 05:53

Re: Help: Set Up model in CFD with time
It seems to me that you are using a specific code. Maybe posting the question in the forum for that code might be better? If not, the answer to your problem is probably nondimensionalization. The time used in the code is not the physical time, but a dimensionless time. To obtain the physical time from the time in the code, you have to multiply it with a reference value, which is usually x_ref/v_ref (the reference values for the nondimensionalization of the length and the velocity).

Ranie Go October 6, 2003 21:37

Re: Help: Set Up model in CFD with time
Let me elaborate more on what my boss want me to do.

I have a specific component/part that need to be heated at around 200 deg C in 1 hr. We will be installing a bracket on that part so we will try to know the max temp that that part would have with the specific condition. This i smy first time running a transient analysis and dont know the exact set-up for this. Theres no problem with me during those days wherein i set-up a part in CFD with non-transient flow. Could someone here guide me how to do this? How to set a transient flow with the specific condition? i dont really enough idea what is a time-step and the stop time in the after i click analyse in the solver. How could i set up time there? How could i know the unit time if that is a minute or a second?

Thanks in advance. hope this is clear.

thomas October 7, 2003 05:11

Re: Help: Set Up model in CFD with time
Hey man, Your problem pretty seems complex :(. Let me try to bring ya further info.

1 - About time step ? Looking the equations of transport, energyn and conitinuity, your variable depends on the time and the space. To solve the space dependance you are using a mesh. The mesh is teh discretization of your domain. By analogy for the time dependance you need to do the same thing : discretize the time. By analogy A time step roughly represent a cell. Your size of your time step and your mesh are couple. For most instable cases I recommand to make sure a Fluid/solid/energy particle does go through more than one cell. This has to be a compromise between accuracy and computational time calculation. Lets take the example of a simulation of solid particle. A too high time step will perform very fast calculation but during the time step you might loose information on the trajectory of the particle. On the other hand, a small time step will give long CPU time calculation but will improve the calculation of your particle trajectory. Hope these notes will complete what you needed.

2 - About your case ? Your seems specific and maybe confidential so you cannot bring too much information. Just follow what berkin told ya to get the physical time.

hope I help ya to get a further idea. thomas ( I do not have time for proof reading so sorry if my english is so so)

P. Birken October 7, 2003 05:12

Re: Help: Set Up model in CFD with time
I'm afraid we can't be of much help, because we don't know your code. Anyhow, if you're using en explicit method, your timestep should be OK for transient ( I think you mean unsteady) calculations. If you're using an implicit method, it depends on your flow how big you can choose the timestep. You need experience to determine that. As for the physical time that corresponds to a numerical timestep, you probably need to know what your velocity and lenght reference values are. The first one is probably computed from the inflow condition of velocity. The second one is probably set somewhere in the code. Note that I say probably. You should consult the guy who wrote the code or the documentary.

All times are GMT -4. The time now is 06:47.