definition of various pressures

[mattkoch]

I am trying to use code_saturne to determine the pressure loss across custom fittings in air ducts (HVAC). To this end, I am modeling a straight duct, comprising an entrance and an exit, and determine the total pressure drop between inlet and outlet. Then I insert the fitting in question between entrance and exit and again determine the total pressure drop. The difference between the two total pressure drops should be the pressure loss in the fitting.

However, I am not having much luck matching even textbook results for known fittings, which I am looking at first in order develop some confidence in my method. I think it has to do with the fact that I really do not understand how the various pressures are defined that get reported in ParaVis.

I am using the unsteady method on an incompressible fluid in code_saturne. Per Bernoulli, the following should apply between inlet (1) and outlet (2), whereby dp12 is the pressure loss (due to friction) I am trying to find:

p1 + 0.5*rho1*v1^2 + rho1*g*h1 = p2 + 0.5*rho2*v2^2 + rho2*g*h2 + dp

In ParaVis, both "pressure" and "total pressure" are reported, but in code_saturne, one must also specify a "reference pressure" for the total pressure, i.e. p0. Hence my questions:

1) is "pressure" as reported by ParaVis identical to p1 and p1 above?
2) is "total pressure" as reported by ParaVis identical to p1 + 0.5*rho1*v1^2 and p2 + 0.5*rho2*v2^2 above?
3) where does the "total pressure reference pressure" p0 fit into this?
4) What would Bernouli look like when written in code-saturne pressures?

By the way, I am determining the inlet and outlet conditions by creating ParaVis "slices" in these two locations, and then doing a ParaVis "integrate variables" on these slices, but as I indicated, I cannot make much sense of the results. For example, combining "pressure" and velocity and elevation in any form known to me never comes out to be "total pressure".

Just for reference, my geometry is a 48" in x 18" duct cross section, with the entrance being 120" long and the exit being 240" long with a 1" body-fitting grid from Salome, using k-omega-SST turbulence in code_saturne. The fitting in question can have any shape (e.g. elbow) and length (but typically on the order of 36"), and has the same cross section. Per inlet boundary condition, I am flowing 1 m3/s through this duct. I understand that this is not perfect in terms of allowing fully developed flow, but it is turbulent and the runs do seem to indicate that the velocity profiles at the inlet and outlet are "plausible".

Any insight would be very much appreciated! Thank you kindly in advance.

[Yvan Fournier]

Hello,

The reference pressure should not impact the head loss.

For a good estimate of the head loss, you should use the built-in head losses measurement functionality, which is also available through the GUI.

Also, mesh quality and refinement may have a large importance.

Regards,

Yvan

[mattkoch]

Thanks Yvan. I cannot find the head loss functionality in 4.0.6. I suppose it was introduced later? The website says that only the 4.0 series is stable and the next stable series is 5.0, but that is not out yet? I like to work with the stable versions only.

In the meantime, I think it would really help to understand how the pressures are defined. For example, yesterday, I also found a reference to the following pressure under RANS k-omega-SST:
p* = p - rho0*g*x + 2/3*rho*k. So, is the "pressure" reported in ParaVis really this p* instead of the p? Is there no succinct summary somewhere that relates the ParaVis output to actual physical quantities?

I ought to be able to use the ParaVis output for the static pressure, dynamic pressure and elevation pressure in order to come up with a head loss estimate. I am just not sure how to interpret the "pressure", "total pressure", "velocity" and, if need be, "k" reported by ParaVis. Ideally, you would point me to equations in the code_saturne manual(s) which correspond to these pressures? The way I would interpret these is as follows, but I get the sense that is incorrect here?

"pressure" = p = static pressure
from "velocity": 1/2*rho*v^2 = dynamic pressure
from geometry: rho*g*x = elevation pressure
"total pressure" = p + 1/2*rho*v^2 + rho*g*x = total pressure
e.g., I do not even see how exactly the reference pressure p0 enters here?

Finally, I suppose one could also go by "effort", but I feel even less confident about interpreting these in the ParaVis output. I suppose I would be looking for "tangential efforts" rather than "normal efforts"?

Thanks,

Matt Koch

Hello,

The reference pressure should not impact the head loss.

For a good estimate of the head loss, you should use the built-in head losses measurement functionality, which is also available through the GUI.

Also, mesh quality and refinement may have a large importance.

Regards,

Yvan

[Yvan Fournier]

Hello,

I am not a specialist in the numerics of the code, so I can't tell you much more than what's in the theory manual. I'll let others answer more in detail.

But version 5.0 is now released, and is the new stable version, so you may try that.

Regards,

Yvan

[mattkoch]

Most excellent, thanks Yvan. I am downloading it right now and will start playing with it. Perhaps (hopefully) the head loss measurement feature will take care of everything. Really appreciate your support on this forum. Always great to be able to put questions to the community.

Hello,

I am not a specialist in the numerics of the code, so I can't tell you much more than what's in the theory manual. I'll let others answer more in detail.

But version 5.0 is now released, and is the new stable version, so you may try that.

Regards,

Yvan

[xingjian]

Hello Matt,

Reference Pressure is atmospheric pressure = 101325 Pa, as you said pressure is static pressure.
Normally, static pressure = rho*g*h + atmospheric pressure, if you make atm = 0 Pa, static pressure becomes gauge pressure, so it should have no effect on the pressure loss.

Best Regards

XING Jian

[mattkoch]

Thanks Xing Jian,

are you saying then that "pressure" as reported by Salome/ParaVis includes BOTH the traditional static pressure AND the elevation pressure. In other words, in my world, total pressure = static pressure + dynamic (or velocity) pressure + elevation (or hydrostatic)pressure, but in Salome/ParaVis, "static pressure" = static pressure + elevation pressure?

Also, what is the Salome/ParaVis "total pressure"? Is it as I defined in the previous paragraph? Plus, is the total reference pressure the same as the (static) reference pressure? Why is the manual distinguishing between a reference pressure and a "reduced" reference pressure? As you can tell, I am fairly lost.

In fact, in the manuals, I have come across the following definitions:

User Manual 4.0, Page 172, ro0 (reduced pressure p*)

p* = p - rho_0*g*(x-x_0) + p_0* - p_0

There it is also stated that both p and p* appear in the post-processing output. Does that mean that "pressure" in Salome/ParaVis is actually p* and "total pressure" in Salome/ParaVis is actually p? In that case, the "reference value for total pressure" in the "Physical properties - > Reference values" part of the GUI seems to refer to p_0? Where do I specify p_0* then? Plus, the way code_saturne uses total pressure is totally different from (my) industry practice, where this pressure would be referred to as static pressure only?

In addition, I found the following:

Theory Manual 4.0, Equation I.2.21, Page 16

p* = p - rho_0*g*x + 2/3*rho*k

These can't possibly be the same definitions for pressure as from the User Manual?

This is all very confusing to me, so sorry for the dumb questions. And my apologies for being so dogged with my questions, but I feel understanding this is a key part in understanding the results from code_saturne.

Thanks,

Matt Koch

Hello Matt,

Reference Pressure is atmospheric pressure = 101325 Pa, as you said pressure is static pressure.
Normally, static pressure = rho*g*h + atmospheric pressure, if you make atm = 0 Pa, static pressure becomes gauge pressure, so it should have no effect on the pressure loss.

Best Regards

XING Jian

[xingjian]

Hello Matt,

I'am so sorry, in fact, I learned slome and code saturne less than two months, I do not even understand that User Manual, so I do not know how it is defined these pressures.

But according to my previous experience, whether the fluid is moving or still, the pressure that we mentioned is static pressure, that is also the pressure which we can measure through the pressure sensor.

And my study the incompressible flow, is to ignore the changes of the elevation, so in my opinion:
total pressure = p + 1/2*rho*v^2 = Static pressure + dynamic pressure

So I think that the pressure in Salome is static pressure, the total pressure in Salome is the total pressure.

I also asked a bunch of questions on another post, I understand you are confused.

Best Regards

XING Jian

[mattkoch]

Hello Xing Jian,

not that we will be able to find the answer here, but based on the manuals, I am beginning to think code_saturne's "total pressure" is actually the traditional static pressure, and code_saturne's "pressure" is actually some sort of reduced pressure they use in their calculations. It sure would be nice if someone fully familiar with code_saturne could confirm or correct this assumption.

Regards,

Matt Koch

Hello Matt,

I'am so sorry, in fact, I learned slome and code saturne less than two months, I do not even understand that User Manual, so I do not know how it is defined these pressures.

But according to my previous experience, whether the fluid is moving or still, the pressure that we mentioned is static pressure, that is also the pressure which we can measure through the pressure sensor.

And my study the incompressible flow, is to ignore the changes of the elevation, so in my opinion:
total pressure = p + 1/2*rho*v^2 = Static pressure + dynamic pressure

So I think that the pressure in Salome is static pressure, the total pressure in Salome is the total pressure.

I also asked a bunch of questions on another post, I understand you are confused.

Best Regards

XING Jian

[Yvan Fournier]

Hello Matt,

For the pressure, I think that's the general idea. There might be a few additional subtleties, but I'd need to check with colleagues for that.

Regards,

Yvan