[Transient rotating domain initialization from steady state run

 Register Blogs Members List Search Today's Posts Mark Forums Read

June 13, 2019, 17:28
[Transient rotating domain initialization from steady state run
#1
New Member

Andrea Sanchez
Join Date: Nov 2018
Location: Montreal, CA
Posts: 1
Rep Power: 0
Hello,

I'm want to run a transient simulation of an axial pump (stator-rotor - stator) using a converged steady-state result file as the initialization for my transient simulation.

***I started a simple model in which I would like to rotate the second section of the cylinder for a better understanding of the complexity of the problem***

While the initial conditions for the stationary domains are successfully interpolated, I'm receiving the following error:

_______________________________________

| FATAL ERROR :
|
| Initial values are required for all variables in TRANSIENT runs.
| In this simulation, no initial value was set for
|
| Variable: Velocity
| Domain: Impeller
|
| The value can be set using the Initialisation Branch in CFX Pre.
|
| To bypass this message and use default solver initial values,
| set the expert parameter "transient initialisation override = t"

___________________________________

I know that the Domain Initialization tab of the Impeller Domain should be defined in order to proceed with the simulation.

Therefore, I decided to use a Rotating Frame Type to apply the initialization values relative to the Impeller domain and the Automatic Cartesian Velocity components to evaluate the last iteration of my steady run, but this option didn't work.

I would like to use the cartesian velocity components (u,v,w) of the last iteration of the steady state run to start the transient run.

Is there any way I could storage these values on the steady run and use them as a CEL function on my transient run? or which would be the best way to do it? (I'm connecting the CFX blocks instead if loading the solution file for the Initial Values)

I've played around with a Domain Initialization using 0 [m s^-1]for each velocity component (u v w) and static pressure of 0 Pa and the solution correctly integrates the steady and transient run (Image below). But I know this affects the transient run and I would like to take the steady continuous to start the transient one.

I've attached several attachments for a better understanding of my problem

Many thanks!

Andrea
Attached Images
 steady and transient.png (14.2 KB, 17 views) final velocity steady run.jpg (85.5 KB, 23 views) solution steady run and transient.jpg (117.6 KB, 16 views) solution steady run.jpg (109.8 KB, 13 views) initial timestep transient simulation.jpg (96.2 KB, 18 views)