[kykylalac]

Hello, please, I'm trying to count not difficult task: at the inlet to a vessel supplied with gas at a pressure 40torr, in the vessel a vacuum, but as I understand it, in ANSYS CFX can not set the vacuum cleaner and you have at least some small pressure set , right?

I wanted to ask 40torr for entry and for example, 1 torr in the vessel, but errors all the time. I increase a little pressure in the vessel - still error at the same place, but when put in the vessel 20 Torr, all consider, but it is not a vacuum.

just tell me what is the difference between the reference pressure in the basic setting and relative pressure in the initialization?

ask reference pressure, relative pressure when tried everything took off, regardless of the pressure enter values

[ghorrocks]

Before you do anything you need to make sure you are in a regime suitable for continuum models (such as CFX with the Navier Stokes equations). You need to work out if your Knudsen number (https://en.wikipedia.org/wiki/Knudsen_number) is OK for a continuum model.

If your Knudsen number is OK then we will discuss how to implement your model.

[kykylalac]

thanks for answer, my Knudsen number is normal for continuum model.
may be, does the error associated with time step?

[ghorrocks]

OK, good.

Then to answer your original question:
Reference pressure is used to normalise all other pressures to values near zero. This reduces numerical round off errors and makes the solver more accurate and more numerically stable.

All pressures are relative to the reference pressure.

[kykylalac]

okay, if i use reference pressure with value 2torr and other pressures is 0,it's mistake?

[ghorrocks]

No, that approach would be good if the system you are modelling had the system running at about 2 torr.

[kykylalac]

may be error is in something else . could you tell me about vacuum in ansys cfx? What are settings for it?

[ghorrocks]

As I said previously, as long as the Knudsen number is in the regime indicating the continuum assumption is appropriate then CFX is OK to model it.

What error are you getting?

[kykylalac]

ERROR #004100018 has occurred in subroutine FINMES. | | Message: | | Fatal overflow in linear solver.
very popular error

[ghorrocks]

Yes, very popular error. Popular enough to get its own FAQ: http://www.cfd-online.com/Wiki/Ansys...do_about_it.3F

Have you tried the suggestions in the FAQ?

[kykylalac]

thanks for link. very useful.
could you explain me what does this notice mean?
****** Notice ****** |
| A wall has been placed at portion(s) of an OUTLET |
| boundary condition (at 44.2% of the faces, 44.3% of the area) |
| to prevent fluid from flowing into the domain. |
| The boundary condition name is: free. |
| The fluid name is: Fluid 1. |
| If this situation persists, consider switching |
| to an Opening type boundary condition instead.

[ghorrocks]

That one is the very next FAQ on the list: http://www.cfd-online.com/Wiki/Ansys...f_an_OUTLET.22

[kykylalac]

what minimum pressure can i use? how can i count the min pressure?
i must have vacuum

[ghorrocks]

CFX is a fluid dynamics solver. A total vacuum means there is no fluid so you can't use a fluid dynamics solver (if that is what you mean).

The minimum pressure CFX can use depends on the application. You simply have to try some values until you find what fails to converge in your case.

And what are you trying to model a vacuum for? What pressure are you intending to use?

[kykylalac]

I'm trying to count not difficult task: at the inlet to a container supplied with gas at a pressure 40torr, in the container a vacuum.
i tried 1 torr in the container.
and i must to use heat transfer on wall. i want to set wall tempreture condotion, but i have error that solver is overflow. i know about link for solve this problem. i don't know where is my mistake.
whene i use heat transfer coef my task solved, but i don't have heat transfer

[ghorrocks]

If there is a flow from a pressure source at 40 torr into a chamber at a vacuum - well the vacuum chamber has to be at some finite pressure. It cannot be absolute zero pressure. So you should look at whether CFX can model the vacuum chamber at its pressure, and check that the knudsen number is OK for it to be assumed as a continuum.

[kykylalac]

Okay, I understand it, thank you. Could you help me with wall heat transfer and more specifically with wall temperature. When the wall has adiabetic and heat transfer coef. (random number) conclusion the task solves, when I set wall temperature the task has error.

[ghorrocks]

Please attach an image of what you are modelling, an image of your mesh and you output file (including both the CCL at the start and the error which finished it)

[kykylalac]

my task: you know the wording

outputSTART (1).txt

??????????1.jpg

??????????12.jpg

??????????3.jpg

[Antanas]

Have you specified proper initial conditions? May you share your project?