[Sasquatch]

Dear Users,

I use kin.e. loss definition provided by the famous paper of Denton (1992) on loss mechanisms defined as:
(static enthalpy exit - isentropic static enthalpy exit) / (total enthalpy exit - static enthalpy exit)

To compute my kinetic energy loss in the stator I use:

(massFlowAveAbs(Static Enthalpy)@S1R1 Side 1 -
massFlowAveAbs(Isentropic Static Enthalpy)@S1R1 Side 1)
/
(massFlowAveAbs(Total Enthalpy)@S1R1 Side 1 -
massFlowAveAbs(Static Enthalpy)@S1R1 Side 1)

This yields a 20% loss due to massive separation, and now for the rotor, Total Enthalpy will be automatically referred to rotating frame of reference so the definition will be:

(massFlowAveAbs(Static Enthalpy)@R1S2 Side 1 -
massFlowAveAbs(Isentropic Static Enthalpy)@R1S2 Side 1)
/
(massFlowAveAbs(Total Enthalpy)@R1S2 Side 1 -
massFlowAveAbs(Static Enthalpy )@R1S2 Side 1)

but this gives me 44% of loss and my Ma contour map suggest that the loss is smaller there. Is this approach correct or there is an error here?

[turbo]

As I had posted long before, CFX has been using VERY confusing terminology which is against turbomachinery community convention. They admitted the problem, but won't fix it being stubborn. Here is my post summary again.

------------------------------------------------------------------------------------------------
For pressure, temperature and density, CFX has 3 kinds :

1. No extensions
2. in Stn Frame
3. in Rel Frame

which mean,

1. Based on the rothalpy, I = h + (W^2-U^2)/2
2. Based on the absolute total enthalpy, ho = h + C^2/2
3. Based on the relative total enthalpy, ho,rel = h + W^2/2

where h = static enthalpy, W = relative velocity and C = absolute velocity.

Accordingly, the 3 kinds of total temperatures are (for a perfect gas),

1. “Total Temperature” = To,rothalpy = T + (W^2 – U^2)/2Cp
2. “Total Temperature in Stn Frame” = To,abs = T + C^2/2Cp
3. “Total Temperature in Rel Frame” = To,rel = T + W^2/2Cp

and also the 3 kinds of total pressures (and total densities) are in the same way.
-------------------------------------------------------------------------------------------------------


Therefore, you will need to change,
massFlowAveAbs(Total Enthalpy)@R1S2 Side 1 --> massFlowAveAbs(Total Enthalpy in Rel Frame)@R1S2 Side 1

[Sasquatch]

Dear turbo,

Thank you for the reply however, surprisingly, I do not have Total Enthalpy in Rel Frame - there is no such option in CFX.

After referring to manual, I understood that, using "Total Enthalpy" for rotating frame of reference I will refer automatically to the relative one.

Could you please refer to this? It is a strong assumption, here is the text from the manual:

"1.1.3.14.2. Ideal Gases
In a rotating frame of reference, the CFX-Solver solves for the total enthalpy, which includes the relative kinetic energy."

[turbo]

If you are using an ideal gas, try just a simpler one like,
Total Temperature in Rel Frame (ho,rel = Cp*To,rel)

[Sasquatch]

Unfortunately I am forced to only real gas due to conditions in the domain.

But does what I wrote earlier mean that my approach is correct?

[turbo]

Yes, those loss definitions are correct, suggested by Denton.

Now you will agree with me in terms of the very confusing and annoying terminology CFX has been sticking to. They have no turbomachinery experts in there, but are not willing to fix this mess. This morning I found another (endless) annoying error in CFX-Post, and reported to Ansys support team.

For your case, I suggest you get pressure and temperature at each station to go to Refprop program to read enthalpy correctly.

[Sasquatch]

Thank you for your reply, I will go for a Refprop I think,

but one thing:

by saying that the definitions are correct you meant that the paper is correct, but my approach with the Total Enthalpy by default understood as relative in the rotor is incorrect?

For example, Velocity at the outlet of rotating frame will be automatically my relative velocity (in turbomachinery known as w), correct?

Just wanted to make it clear, and I think I will send them an email too ...

Best regards,

[turbo]

Yes, "Total Enthalpy" of CFX-Post looks based on the relative velocity, and thus it should be called "relative total enthalpy", ho,rel. Another confusion comes in here.

If you plot the Inlet-to-Outlet plot in turbo mode, you can figure out what the variables are really to be called.

Another uncertainty would be how CFX-Post calculates the "Isentropic" enthalpy. You will need to double-check them using Refprop.

[Sasquatch]

Thanks for the clarification!

I have another doubt, related more to the physics, rather than solver, namely:

If I want to evaluate the TOTAL-TO-TOTAL efficiency of the turbine stage, shall I refer to absolute values or relative values (in Stn Frame or in (Rel) default frame)

This:
( (massFlowAveAbs(Total Enthalpy)@R1S2 Side 2) -
(massFlowAveAbs(Total Enthalpy)@R2S3 Side 1) )
/
(massFlowAveAbs(Total Enthalpy)@R1S2 Side 2 -
(massFlowAveAbs(Isentropic Static Enthalpy)@R2S3 Side 1 +
0.5 * massFlowAveAbs(Velocity^2)@R2S3 Side 1) )
gives 35%.

and this:

( (massFlowAveAbs(Total Enthalpy in Stn Frame)@R1S2 Side 2) -
(massFlowAveAbs(Total Enthalpy in Stn Frame)@R2S3 Side 1) )
/
(massFlowAveAbs(Total Enthalpy in Stn Frame)@R1S2 Side 2 -
(massFlowAveAbs(Isentropic Static Enthalpy)@R2S3 Side 1 +
0.5 * massFlowAveAbs(Velocity in Stn Frame^2)@R2S3 Side 1) )
gives 64%.

Which one is physically appropriate? According to Dixon's book on Turbomachinery, ABSOLUE should be always taken into account - hence only in STN FRAME, so 64% is the result.

[turbo]

Absolute enthalpy changes.

[Sasquatch]

Thanks, therefore the latter (with Stn Frame), of 64% is correct.

Similarly with the load coefficient:

(massFlowAveAbs(Total Enthalpy in Stn Frame)@R1S2 Side 2 -
massFlowAveAbs(Total Enthalpy in Stn Frame)@R2S3 Side 1)
/
(0.5 * massFlowAveAbs(Velocity in Stn Frame Circumferential^2)@S2R2 Side 2)

which yields 2.1 [-]

That makes sense, thank you for clarification !!!