[amodpanthee]

Does any body have link for references to execute cfd analysis of Pelton turbine using ANSYS CFX? Any link for tangential flow turbines like turgo turbines might be beneficial.

[oj.bulmer]

Have a look at the CFX tutorials, they provide ample information.

OJ

[amodpanthee]

Is there any specific tutorial for tangential flow turbines?

[Far]

I think there is no specific tut on pelton turbine. For pelton turbine you have to model full wheel and place the jets at appropriate location. Placing GGI between jets and wheel may not be recommended but you can try.

Meshing definitely would be the most time consuming part.

Make the CFD model (meshing, boundary conditions) and come back to further details. If you need help in meshing, you are also welcome.

[amodpanthee]

[IMG]file:///C:/Users/RENP-PID-437-1/Desktop/domain.JPG[/IMG]

[amodpanthee]

I have set the domain as shown in the pic....Is it OK to simulate with it?

[ghorrocks]

This will probably work, as long as you only want to model 5 blades.

It can probably be optimised somewhat by doing a few things:
1) Put a wall instead of your outlet, excpet for at the bottom. This will converge much easier and should not affect things too much.
2) You probably do not need to model the nozzle in too much detail. I would just make it an inlet with the necessary velocity - unless the small velocity distribution across the jet is important.

[amodpanthee]

I am running a trial on the model. The boundary conditions are set as shown in the attachment. It gave an unexpected result. Why is the direction of jet changed when it passes from stationary domain to the rotating domain. I think I have some problem with rotating domain.

I have 22 buckets rotating at 600 rpm and 2 nozzles of which I took 5 buckets and lower nozzle. The nozzles are not in symmetric locations. How to choose pitch value in such cases?

[Far]

At-least you need eleven blades and one nozzle. If nozzles are not symmetric then you need to use full wheel with two nozzles.

[ghorrocks]

The jet is not straight because you are plotting velocity in the local frame of reference. To get a straight jet plot in CFD-Post using "Velocity in Stationary Frame of Reference".

[amodpanthee]

The direction does not change when we see it in stationary frame too

[Far]

Originally Posted by ghorrocks

The jet is not straight because you are plotting velocity in the local frame of reference. To get a straight jet plot in CFD-Post using "Velocity in Stationary Frame of Reference".

did you give the rpm to runner?

[amodpanthee]

Yes of course..... Runner is rotating in anticlockwise direction. I have assigned value 600 rpm.

[amodpanthee]

I have problem with the direction of jet at interference between rotating and stationary domain, attached picutre. I think I have problem with defining the flow direction at Inlet.

I was trying to simulate a complete runner and got an error on setting the cartesian co-ordinates 0,0,0, attached picture. So, I tried simulating taking normal to boundary which runs, attached picture.....I ran it although it does not give desired result. So, I think the flow direction is the reason for deflecting the jet.

Is there any rule of thumb like right hand thumb rule or left hand thumb rule to define the flow direction? The two jets in my model are placed inclined at an angle 45 degree in second quadrant and third quadrant in XY plane. Changing the flow direction will probably help run the simulation correctly.

Any suggestions would be highly appreciated.

[Magma]

Originally Posted by amodpanthee

I have problem with the direction of jet at interference between rotating and stationary domain, attached picutre. I think I have problem with defining the flow direction at Inlet.

I was trying to simulate a complete runner and got an error on setting the cartesian co-ordinates 0,0,0, attached picture. So, I tried simulating taking normal to boundary which runs, attached picture.....I ran it although it does not give desired result. So, I think the flow direction is the reason for deflecting the jet.

Is there any rule of thumb like right hand thumb rule or left hand thumb rule to define the flow direction? The two jets in my model are placed inclined at an angle 45 degree in second quadrant and third quadrant in XY plane. Changing the flow direction will probably help run the simulation correctly.

Any suggestions would be highly appreciated.

Sir, What I believe is : if you are working with Cartesian coordinates, you must specify each components of velocity!! cheers!!!

[ghorrocks]

Just use normal to the boundary for the inlet. No need to specify a direction.

Your jet is being deflected before it even gets to the pelton blades. This means something is fundamentally wrong with your setup to do this. What are streamlines of the air phase look like? If you have over-constrained the air phase it will deflect the water jet like this. I mean the air needs somewhere to go to get out of the way of the water jet.

[Far]

Why are you using symmetry condition? if it strange to use symmetry condition for this type of flow where you don't know exactly where flow is going...


My suggestion :
we did similar study on tip injection effect on stall margin enhancement of transonic axial compressor rotor ( to be specific, NASA rotor 37)

http://arc.aiaa.org/doi/abs/10.2514/6.2012-329

http://www.researchgate.net/publicat...ow_Compressors

The point which I want to discuss is that " we have divided tip clearance into two parts : one was with rotating domain of rotor and second was with stationary domain of tip injection. So instead of putting interface at the intersection of jet and casing of compressor we gave it some space into the rotating domain.

I can locate that paper and send you if needed.

[amodpanthee]

Far

Can I get the full length paper? amodpanthi@ku.edu.np

[Far]

Sorry for delay. did you get paper?

[amodpanthee]

Yes, I received the attachment. I am going through it.

Thank you