[supermanks]

Hello all,

I'm trying to simulate the flow through a centrifugal compressor (geometry made in Creo 1.0, exported to ICEM for meshing, and Fluent for solution). I have given a mass-flow-inlet BC at the inlet, and I have given the static pressure at inlet where it says "supersonic/initial gauge pressure", and the mass flow rate in the corresponding field. Also, I created monitors of "area-averaged" and "mass-averaged" "static pressure" at inlet, just to make sure. However, as the iterations began, it was surprising to find that the averaged static pressure was not the straight line of the constant BC value that I had expected, but was a value different value. Afterwards, it appeared to approach the input value, but it straightened out, and never reached the BC value of static pressure that I had input.

I am totally confused regarding this as I do not understand if I entered the wrong pressure in the "supersonic/initial gauge pressure" blank, or if my mesh is not good enough, or if it is some other factor at play, but I thought fluent worked its way to the flow field inwards from the BCs that we input, and hence the BC monitors should show a graph of constant value that was input. I'd be really grateful if anyone could give their input as to what is going wrong here. [Operating pressure = 0, and all other pressures entered at absolute values, as learned from different tutorials] Thanks a lot !

[pakk]

Out of memory, so I could be wrong: if you use the pressure-based solver, the boundary pressure that you supply is the total pressure. The difference is the dynamic pressure, rho*v*v. Could the difference be that?

[supermanks]

Hello Pakk,

I had the same confusion. After lots of tutorial-referring and manual-checking, I found that supersonic/initial gauge pressure refers to static pressure (please see the fluent manual for this). But this alone is not sufficient to make the problem a properly constrained one, as velocity or stagnation pressure has to be supplied. Thus, we can choose mass-flow-inlet, or velocity-inlet, if we choose to enter velocity (and static pressure), or we can choose a pressure-inlet condition, thus supplying the appropriate stagnation and static pressures directly.

Of course, I might be gravely mistaken about this so please do provide your inputs on this. Am I taking the wrong operating pressure of zero, or is my mesh quality not up to the mark? Also, the convergence pattern of residuals and monitors changed when I changed the turbulent length scale, from a very small value (0.05 mm), to that of the inlet pipe diameter (40 mm). Could this be making a difference in the static pressure BC? I am really confused as tho what the length scale value should be and I'd be grateful if someone could clarify this as well.

Thanks !

[amin.z]

Hi!

As it's referred in Fluent tutorial "supersonic/initial gauge pressure" is a initial guess of static pressure for subsonic condition and it's updating along the iterations! So difference between your input pressure and final pressure of Fluent isn't irrational!
If all of the condition of your problem be constant and just you change the initial pressure, I guess you must finally reach same pressure!

Finally I think the pressure inlet or mass-flow rate BCs is more appropriate in your case against velocity inlet!

[supermanks]

Hi amin.z,

Are you quite sure that it's only a "guess" value for static pressure???

Because to have a properly constrained problem (so that fluent can't operate with junk values where we don't input them), we need inlet static pressure PLUS something to calculate stagnation pressure at inlet as well. The latter component can be supplied as a direct stagnation pressure (if we choose pressure-inlet BC), or a velocity (which will be used indirectly to calculate stagnation pressure, by specifying mass-flow or velocity itself)

If it indeed is a guess value, how is Fluent solving all the equations without the actual BC, because not specifying static pressure results in Fluent being able to manipulate the variable freely, and it is possible we may never reach the desired the inlet BC !!! I thought a BC meant that the particular value HAD to be imposed on that boundary no matter what !

Finally is my turbulent length scale (being equal to inlet diameter) valid at all?

Kindly help with my above queries ! Thanks a lot !

[amin.z]

Originally Posted by supermanks

Hi amin.z,

Are you quite sure that it's only a "guess" value for static pressure???

Because to have a properly constrained problem (so that fluent can't operate with junk values where we don't input them), we need inlet static pressure PLUS something to calculate stagnation pressure at inlet as well. The latter component can be supplied as a direct stagnation pressure (if we choose pressure-inlet BC), or a velocity (which will be used indirectly to calculate stagnation pressure, by specifying mass-flow or velocity itself)

If it indeed is a guess value, how is Fluent solving all the equations without the actual BC, because not specifying static pressure results in Fluent being able to manipulate the variable freely, and it is possible we may never reach the desired the inlet BC !!! I thought a BC meant that the particular value HAD to be imposed on that boundary no matter what !

Finally is my turbulent length scale (being equal to inlet diameter) valid at all?

Kindly help with my above queries ! Thanks a lot !

My dear friend!
This is true that "supersonic/initial..." is a guess for the static pressure! I'm attached a snapshot of fluent tutorial that valid my idea! this tutorial is available in help of your fluent or in below link:

http://www.google.com/url?sa=t&rct=j...,d.d2k&cad=rja

I think the static pressure can be measured with using "total pressure" and "total temperature"! if you see equations of total temp. and pre. you will see this is possible!

finally about your turbulence parameter! like you I'm working on a centrifugal compressor!but my skills about turbulence is limit! Based on my search about turbulence models I think your model for first try is appropriate but you must try another turbulence parameters and select best model based on your results! this is my idea and certainly there is better ideas!
If you find answer for turbulence model Notify me!
Good luck

[supermanks]

Alright, so "supersonic/initial..." refers to the static pressure guess value.

But, there has GOT to be some way to restrict the inlet static pressure to our desired value, otherwise it's all a bit vague and unconstrained. I'm sure I'm missing out something here.

Any more ideas people?

Thanks !

[supermanks]

Any ideas people?? I'm really stuck with this.