# Discussion Forum

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

## natural convection

Hi everyone,
I'm trying to make a model of natural convection in a close system (air CANNOT go in or go out), but when I start to solve there is not solution and gave the error:

Time 0:
Failed to find consistent initial values.
Last time step is not converged.

while in a open system (air CAN go in or go out) solution is quickly given.

Can anyone help?

Best regards
Simone

4 Replies Last Post Aug 21, 2009, 7:21 PM EDT

Hi everyone,
I'm trying to make a model of natural convection in a close system (air CANNOT go in or go out), but when I start to solve there is not solution and gave the error:

Time 0:
Failed to find consistent initial values.
Last time step is not converged.

while in a open system (air CAN go in or go out) solution is quickly given.

Can anyone help?

Best regards
Simone

Dear Simone,

This is probably due to the pressure variable. When you have a closed system, the reference pressure for the system is undefined when using the "Incompressible Navier-Stokes". You therefore need to lock the pressure in one point.

The best approach is probably to lock the average pressure and adding a constraint on the average pressure (using a coupling variable and a global equation). However, this is an expensive solution from the computational point of view.

A possible cheap workaround is to specify the pressure variable in one point. You can do this through the following:

1. Select "Point Settings" from the "Physics" menu.
2. Select a point in the "Point selection" list
3. Check the "Point constraint" box
4. Specify a value for the pressure
5. Click "OK".

Best regards,

Ed

Dear Ed,

I'm using 'Weakly Compressible Navier-Stokes', and the floating law as a volume-force in y-direction (2 dimensional model).

I try to use 'Incompressible Navier-Stokes' and the 'Point constraint'.

Thanks a lot

Regards, Simone

Hi,
I tried to use 'Incompressible Navier-Stokes' and the 'Point constraint' for my model, and on my personal notebook (PC with a client-server connection) works well, but on Mac of my study (the server) it does not work.
Is possible that the same file works on a Pc and does not work on a Mac?

thanks all

Simone