[Viddesh]

Dear All,

I have been facing some dilemma in CFD post recently. I have simulated a steady state full 3D Impulse steam turbine analysis (with nozzles, 4 rotor rows, stators and extension domain) with the following parameters;
Inlet BC: 1.8 kg/s mass flow rate
Outlet BC: 0.25 bar Pressure
Inlet steam temp is 250 deg C
Rotor speed: 700 rpm
Heat Transfer model: Total energy [Incl. viscous work term] with high speed (compressible) heat transfer model turned on.
Turbulence model is SST
I have used frozen rotor interface for rotor stator interface.

Now coming to the point I'm having issues with;
1. I have extracted massflowave variables like temp, enthalpy, Mach no. etc. using the function calculator option in CFD post.
I have also plotted temperature contour on the same patch/Interface where I have used function calculator.

My observations (only w.r.t Temperature (as that is my concerned variable));
1. The value which function calculator returns for temperature at patch say above rotor1 (patch above rotor1 blades) is significantly lower than temperature contour plot of the same patch.

Thank you for any help provided.
Regards

[ghorrocks]

Have a look at the flow direction. Some areas of your plot may have backflow and that will affect the calculation. Also look at how the massFlowAve function handles backflow - there is another function which only considers the forward flow which might be more suitable for you.

[Viddesh]

Originally Posted by ghorrocks

Have a look at the flow direction. Some areas of your plot may have backflow and that will affect the calculation. Also look at how the massFlowAve function handles backflow - there is another function which only considers the forward flow which might be more suitable for you.

Thank you for your reply ghorrocks. I appreciate the prompt response.

Yes as you have said there is some backflow/recirculation zones in some areas. As you suggested I have extracted temperature using massflowaveAbs function but there is only a marginal increase in the value from 311 K to 317 K. Any suggestions?

Also, for my steam turbine casing design, what temperatures would be more realistic from this CFD simulation; Total or static temperature? FYI, my CFD consists of only fluid domain.

Best Regards,
VB

[ghorrocks]

How are you estimating what you expect the average temperature to be? Are you just looking at the image you posted? You need to take into account the variation in mass flow and the backflow regions if you want your average to be realistic. Are you doing that?

If not, you could create a new variable and define it to be mass flow * temperature, and then just draw the positive flow regions. This should show an image which is representative of your calculation.

Total or static temperature? I don't like answering questions like that as if you are doing analysis on devices like this you really should know what variables like this mean and when they are appropriate. So have a look into this yourself and let us know what you think the answer is. (And the answer to which temperature to use is: it depends on what you are trying to do, so it needs more information to be answered anyway.)

[Viddesh]

Originally Posted by ghorrocks

How are you estimating what you expect the average temperature to be? Are you just looking at the image you posted? You need to take into account the variation in mass flow and the backflow regions if you want your average to be realistic. Are you doing that?

If not, you could create a new variable and define it to be mass flow * temperature, and then just draw the positive flow regions. This should show an image which is representative of your calculation.

Total or static temperature? I don't like answering questions like that as if you are doing analysis on devices like this you really should know what variables like this mean and when they are appropriate. So have a look into this yourself and let us know what you think the answer is. (And the answer to which temperature to use is: it depends on what you are trying to do, so it needs more information to be answered anyway.)

Thank you for your response.
I'm relying on CFD values for temperatures and not carrying out any calculations.
As of now, I'm considering the average temperatures by looking at the temp contours and using the same for my thermal analysis. But what bothers me is that the average value by function calculator is significantly different from temp contours.

As you have suggested, I've created a new variable (massflow*Temperature) and used it to plot the contour for the same (Image attached). From the picture, does it mean that +ve values indicate backflow / recirculation areas and -ve values are the region/area where flow is flowing normally into the domain. (at the outlet of the nozzles)
(My flow direction is Z axis which is the normal practice for turbomachinery CFD)
From the image, does it mean the at the outlet of my nozzle region shows -ve values as the flow is moving into the next domain and +ve values (indicated by orange color) mean that those areas are where backflow/recirculation is happening.

Is there any way to get average value calculated by function calculator closer to temp contour average value?

Coming to Total vs static temp. I have gone through multiple websites and forums and kind of know that temperature to be considered is static temperature as static temp is when you measure the temp. with no relative velocity effect and total / stagnation is when fluid is brought to rest isentropically and its kinetic energy is changed into internal energy. Also, I have read an article where it says usually the actual fluid temp is somewhere between static and total.
I just want an insight from your end on my interpretation of the same.

Regards,
VB

[Opaque]

Just a warning when dealing with fluid with considerable variation of properties. Steam is one of them, in particular, if near the saturation lines (dome).

The conserved quantity is energy, and temperature is a derivation used because it can be measured with respect to a reference.

What are you trying to figure out by comparing temperature averages of sorts?

The only thing that can be guaranteed is the amount of energy, and depending on the average you expect a similar temperature. Doing such calculations directly is bound to differences unless the specific heat capacity is constant, and the gas behaves ideally

[ghorrocks]

Good point - is temperature the right variable to be averaging?

Is there any way to get average value calculated by function calculator closer to temp contour average value?

If you are only going to accept the average when it is your pre-conceived value, then why not just say the average is what you want it to be and ignore the reported value?

Or maybe a better question is to ask "Why does the reported average differ from what looks like the average temperature from the cross section?"

[Viddesh]

Originally Posted by Opaque

Just a warning when dealing with fluid with considerable variation of properties. Steam is one of them, in particular, if near the saturation lines (dome).

The conserved quantity is energy, and temperature is a derivation used because it can be measured with respect to a reference.

What are you trying to figure out by comparing temperature averages of sorts?

The only thing that can be guaranteed is the amount of energy, and depending on the average you expect a similar temperature. Doing such calculations directly is bound to differences unless the specific heat capacity is constant, and the gas behaves ideally

Dear Opaque,
Thank you for the response.

Coming to my problem, I have simulated superheated steam that will remain in superheated state at the exit of the turbine as this is a requirement of the system.


My goal from this CFD analysis is to;
1. Extract the temperatures at each row which will help me to determine temperatures of bearings during separate thermal analysis.
2. Exit temperature of steam.

When i extracted above temperatures, i noticed patch/interface average values are significantly different than contour values.
So as Mr. ghorrocks has pointed out it could be due to averaging method used by function calculator in CFD post is not suitable in my current case.

I want to know how best can i predict the temperatures from my cfd analysis for the above points.

Regards,
VB