[looee]

Hi everyone!
I've been studying CFX and have faced with misunderstanding.
How does ansys calculate total pressure?
Can anybody explain me it using my example?
Geometry and BC see on fig. 1.
I've simulated this problem three times:
1) Mesh~93,300 cells (fig.2). Referense pressure = 0, pressure outlet = 1 bar.
2) The same mesh. Ref pressure = 1 bar, pressure outlet = 0.
3) Mesh ~3,500,500 cells (fig.3). Ref pressure = 0, pressure outlet = 1 bar.

The results of these simulations see on fig. 4,5,6 correspondingly (F column for difference between upper and bottom planes)

As I've understood, In Ansys:
pressure+reference pressure = static pressure
pressure+reference pressure + *g*h=absolute pressure.
In Ansys help is written, that total pressure = static pressure + dinamic pressure. So why total pressure is less then static pressure in my simulations?
Moreover, why are results for total pressure difference (6F cell) in (1,2) and (3) simulations so different (In about 2,5 times)? All problems have converged to my precision very fast.

[looee]

The 6th figure (results for the 3rd simulation) is it.

[ghorrocks]

I cannot figure out which image refers to which simulation. As they are just tables of numbers can you put the relevant numbers directly in the post, not as images?

[looee]

Originally Posted by ghorrocks

I cannot figure out which image refers to which simulation. As they are just tables of numbers can you put the relevant numbers directly in the post, not as images?

I've united tables. Maybe it'll be more clearly.
I think, that putting the relevant numbers directly in the post will be less clearly for understanding (without table).

[evcelica]

Pay attention to reference density and location. These may make a difference. By default, location is automatic, so who knows where that would be.

It also may depend on whether you are using full buoyancy model or boussinesq approximation.

Easiest thing to do would be to make a test model where you know what the results should be already, then you can figure out how CFX calculates these things.

Start with free slip walls, and a very low flow, so you are only figuring out the static head portion of the pressure variables. Remember to set a reference density and location.
Or use 1 simple 1D model using symmetry for the walls, so there is no pressure drop whether you have flow or not.

Start adding to the test model to figure out how other components contribute.

PS. That is a very strange mesh for fluids. It does not look like good boundary layer resolution. (Not suggesting that is making a difference here)

[looee]

Originally Posted by evcelica

Pay attention to reference density and location. These may make a difference. By default, location is automatic, so who knows where that would be.

It also may depend on whether you are using full buoyancy model or boussinesq approximation.

Easiest thing to do would be to make a test model where you know what the results should be already, then you can figure out how CFX calculates these things.

Thank you for your respond.
I didn't pay attention to ref. density and location because I haven't understood yet what they mean. But I'll figure it out now.
For this simulation I already know what results should be (I've made hand calculations). And all values (except total pressure) correspond to these results.

[Opaque]

For clarification, in ANSYS CFX

Absolute Pressure = Static Pressure + Reference Pressure

Static Pressure = "solved" Pressure (variable in the momentum equation) + hydrostatic pressure

For incompressible flow:
Total Relative Pressure = Static Pressure + "dynamic" pressure

For compressible flow:
Total Relative Pressure = Absolute Total Pressure - Reference Pressure
Absolute Pressure = Absolute Pressure if the flow is isentropically slowed down to rest.

[looee]

Originally Posted by Opaque

For clarification, in ANSYS CFX

Absolute Pressure = Static Pressure + Reference Pressure

Static Pressure = "solved" Pressure (variable in the momentum equation) + hydrostatic pressure

For incompressible flow:
Total Relative Pressure = Static Pressure + "dynamic" pressure

For compressible flow:
Total Relative Pressure = Absolute Total Pressure - Reference Pressure
Absolute Pressure = Absolute Pressure if the flow is isentropically slowed down to rest.

Thanks!
I've understood what absolute pressure, static pressure, ref. pressure and "solved" pressure means.

But I supposed, that:
Absolute Pressure = Static Pressure + hydrostatic Pressure
Static Pressure = "solved" Pressure (variable in the momentum equation) + reference pressure.

Actually these definitions aren't making a significant difference here.

The values of these quantities (absolute pressure and static pressure) in my simulation are looking sanity (in my definitions they coincide with the hand calculated values).

I also understand that total pressure should be = static pressure + .

But in my simulation total pressure is less then ("solved" Pressure + ref. pressure). It's wrong, isn't it?

[Opaque]

Actually these definitions aren't making a significant difference here.

It may not make a "quantitative difference" for you, but it makes it understandable for others to speak using a common vocabulary for CFX users.

But in my simulation total pressure is less then ("solved" Pressure + ref. pressure). It's wrong, isn't it?

Do not try to compare Total Pressure to Solved Pressure + Ref Pressure. I was clear earlier, they are not defined that way for ANSYS CFX. If you read it that way, it would be great to point us to the documentation section where it says so.

Does it looks off? Yes it does, but as Evcelica mentioned, more information (in CFX speak) is needed.

Hope the above helps

[looee]

Originally Posted by Opaque

For compressible flow:
Total Relative Pressure = Absolute Total Pressure - Reference Pressure
Absolute Pressure = Absolute Pressure if the flow is isentropically slowed down to rest.

Can you explain what is absolute total pressure?
There is only the "total pressure" in CFX (not total relative pressure or ablosute total pressure).

[looee]

Originally Posted by evcelica

Pay attention to reference density and location. These may make a difference. By default, location is automatic, so who knows where that would be.

It also may depend on whether you are using full buoyancy model or boussinesq approximation.

Analyzing my results I can suppose, that reference location is located in pipe outlet.
I think, that there aren't any problems with reference density too.
So what is wrong?

[Opaque]

Absolute Total Pressure is effectively Total Relative Pressure + Reference Pressure.

That variable is not presented by ANSYS CFX anywhere, but if you were to evaluate Total Density, you will use Absolute Total Pressure and the Total Temperature in the equation of state

In ANSYS CFX, "Total" prefix is what others call "Stagnation".

Total Pressure = Stagnation Pressure
Total Enthalpy = Stagnation Enthalpy
Total Density = Stagnation Density,

...

[looee]

Originally Posted by Opaque

Absolute Total Pressure is effectively Total Relative Pressure + Reference Pressure.

That variable is not presented by ANSYS CFX anywhere, but if you were to evaluate Total Density, you will use Absolute Total Pressure and the Total Temperature in the equation of state

In ANSYS CFX, "Total" prefix is what others call "Stagnation".

...

Thanks a lot for your respond again!

But what is total relative pressure then?
You said, that Total Relative Pressure = Absolute Total Pressure - Reference Pressure, then I asked you what is Absolute Total Pressure.

You answered that Absolute Total Pressure is effectively Total Relative Pressure + Reference Pressure

I've also just found some proves of your answers (see fig.) - maybe It'll helpful for other people, like me

But I still don't understand how total pressure in my simulations has been calculated

[Opaque]

Absolute Total Pressure is effectively Total Relative Pressure + Reference Pressure.

That variable is not presented by ANSYS CFX anywhere, but if you were to evaluate Total Density, you will use Absolute Total Pressure and the Total Temperature in the equation of state

[AtoHM]

Honestly, I have no solution and can't see what might be wrong. However, can you extract the static and total pressure values directly at the inlet/outlet boundaries instead of these planes?

As the total pressure relates to density and velocity, also check carefully that these values are in the expected range, calculate the value yourself and see if there's a difference.

[looee]

Originally Posted by AtoHM

Honestly, I have no solution and can't see what might be wrong.
However, can you extract the static and total pressure values directly at the inlet/outlet boundaries instead of these planes?

As the total pressure relates to density and velocity, also check carefully that these values are in the expected range, calculate the value yourself and see if there's a difference.

Can I send you a link to a cloud?
Except the extracted values for these planes, I've done it also for inlet and outlet directly (see fig. on the 4th message in this theme).

Values of density and velocity are in the expected ranged. I've calculated them manually.