[ghorrocks]

If your simulation is close enough to validation results then you have some confidence the simulation is working accurately. There is a small chance the result was just a fluke and if you make a change it will be miles off - running a different mesh size would confirm whether this has happened.

Your second question is an FAQ: https://www.cfd-online.com/Wiki/Ansy...gence_criteria

[Opaque]

Just be careful.

Solutions are ONLY converged when the mathematical statement is satisfied, i.e. the residual of the system is ZERO.

Solutions can ONLY be unique for a given method when they no longer depend on any of the discretizations parameters, i.e. mesh size, timestep, under relaxation, etc

Achieving the above is most of the time too costly, and sometimes not worth the additional effort since the models being solved already include a certain amount of error, i.e. they only approximate the real problem. Then, the experience working with specific models comes in, and in the end, it is YOUR criteria and responsibility to decide.

However, be careful of the practice of "solution is converged because my variable of choice is no longer changing". Here is a counterexample,

Say we want to find the root of f(x) = x^2 + 1, i.e. solving the problem

f(x) = x^2 + 1 = 0

If you use an iterative method, it will lead towards a constant value of 0, i.e. exactly where the residual is at its minimum, but a real-valued solution does NOT exist. What will you do then?

Food for thought

[stfahim]

Thank you for your reply. I got a better understanding.

[nicolas.sargi]

Hello.

Did you figure out how to specify a pressure drop across a device?

I have a 4.5m*2m cylinder, and I need to specify a pressure drop across the midplane of this cylinder.

An easy way would be to specify this plane as an inlet, and make the relative pressure be the pressure drop, but then I would need to specify the other parameters (temperature etc..), while I only have a pressure drop across the device.

Any help would be appreciated.

Thanks.

[ghorrocks]

Make it an interface and set the interface parameters to have a pressure drop (or mass flow rate) over the interface. Easy.

[nicolas.sargi]

Thanks for the answer!

I am struggling to define the interface so that CFX understands that the pressure drop is occurring throughout the device.

Do you have a tutorial on simple steps to set up an interface?

Thank you

[ghorrocks]

Have a look at the CFX tutorials. Several of them use interfaces.

[nicolas.sargi]

Could you please point out a fluid-fluid interface for me?

Thank you

[ghorrocks]

Have a look at the CFX tutorials. Several of them use interfaces.

...... really, just look at the tutorials. It is very basic stuff.