Calculating work done to a fluid by a rotor and stator in a sealed unit

[r.lawrence88]

Hi,

I am looking to calculate the work done to a fluid that is sealed inside a unit between a rotor and a stator type design (no inward or outward flow). Essentially the rotor is rotated by an external motor to a specific RPM to agitate the fluid between the two solid bodies, there is space for the fluid to move between the solids.

I want to be able to change fluid parameters and cavity shape (fluid contained between the Rotor and Stator design) by changing the Rotor and Stator design and compare the results.

(Ultimately I would also like to calculate the heat transfer from the internal fluid through to an external body but i realise this is another question in itself.)

I am very new to CFD and have worked through a few of the tutorials and read a couple of threads on here but nothing specifically like this. 

I have been advised to use boundary conditions on rotational speed, fluid, temperature (assumed), kinematic viscosity and density with a turbulence model of k-w SST. The solution scheme for pressure-velocity coupling is the SIMPLE scheme. The simulations of the rotor/casing covered are single phase, steady state, isothermal in a moving reference frame.

What i have done so far is:

• modelled the fluid cavity as a solid in Inventor then pulled it by itself into CFD 2019. I've read that cutting this into a section would give faster results. So I cut it to a 90° section of the cavity.
• change the material to a fluid (I'm starting with water to get it to work but I have specific heat transfer fluid I would like to model that I have created in 'My Materials')
• added the boundary conditions but I am unsure as to where these go in regards to the many faces i could select since we don't have an inlet or outlet
• Autosize the mesh. I've also read that a really fine mesh is needed for this sort of simulation.
• hope for the best 🙂 

Any help with how and where to enter these boundary conditions would be very much appreciated. Thank you for reading if you have made it this far.

Thanks 

Ray 

[David.Short.]

Hi Ray,

Sound fun! I am still not clear on the geometry can you send some pics of what the model looks like in reality?

We have had a few people modelling leakage between rotors and stators maybe take a look at these threads as they should give you some pointers.

gas-turbine-wheel-space-modelling

modelling-lekaage-flows-in-gas-turbines

We managed to get some nice 2D sims running. Could you run your model as 2D?

All the best,

David

[r.lawrence88]

Hi David,

Thanks for the reply.

I have attached a pic of a section of the fluid void.

I will have a look at these threads, thanks.

I think this is essentially how I'm trying to start, I'm putting a 90° 1mm section of the attached and trying to put boundary conditions to the faces.

Thanks 

Ray 

[David.Short.]

I have just done a quick test with a 360 degree 2D model and think all you need to do is add a rotational velocity to the outer edge (where the stator begins) to replicate the rotor spinning. Things seem to be working well.

To do a 90 degree section you would need to utilise periodic boundaries however i don't think it is necessary as you should have enough PC power to adequately mesh a complete 360 degree model as it is just 2D.

What do you think?

All the best,

David

[r.lawrence88]

Hi David,

Thanks again for your reply.

I have got this working in the way you have expressed.

It outputs fluid velocity in mm/s. I am attempting to find work done in Watts so I would need either internal energy or torque applied to the fluid. Do you know how i would be able to display this or find this?

Thanks

Ray

[Jon.Wilde]

Could you use the Wall Calculator?

[r.lawrence88]

Hi John,

I have attempted to use the wall calculator tool and I'm getting a value exactly equal to my initial condition temperature set.

After having a good look around at all settings available I came across these:

• A global result for Wall Film Coefficients (W/mm2/K)
• A surface Boundary Condition for Total Heat Generated (W)
• A surface Boundary Condition for Heat Generation (W/mm3)

I am struggling to get the heat transfer side of the results to output anything I have attached images of the settings i have entered and my results.

Are you able to give some advice on the best way to set a heat transfer function so I can get results and maybe if any of the tools mentioned above could be used in aiding my solution.

(Ultimately I want to 3D model the full assembly and analyse the heat transfer to an outer steel body. If this could be kept in  mind for adapting the 2D process to suit)

Thanks again for the help so far guys.

[Jon.Wilde]

For some reason I cannot see the model above but can share an idea based on the screenshot from the model settings.

Your image shows that there is a zero heat load (so not a lot should change 😉 ) and also there is no way for heat to escape, is that intended? This would simply continue to heat indefinitely as it is, perhaps intended though.

Hope that helps,

Jon

[r.lawrence88]

Hi Jon,

Our intention is to see how much the fluid is heated when enclosed in the system through friction of the rotating Rotor. 

We would then want to know how much of this heat is transferred outward through to an external body.

Our idea was that you have the rotor hot and the external surface cold to give a temperature gradient through the fluid.

We are setting all these values so the system isn't really reacting to a real world situation. It doesn't tell us how much work the rotor is doing as we are telling it what it is doing. We are interested in how much torque the rotor is giving to the fluid to give the fluid internal energy which we can use as a proxy for heat transfer. 

Thanks

Ray

[matt.bemis]

Hi @r.lawrence88,

I gave this thread a quick read and wanted to provide some input. Sorry if I am repeating information already stated. Autodesk CFD will not show you a increase in temperature/enthalpy due to compression of a fluid. For example, it will not accurately model a classic fluid in a piston thermo example. It will also not compute heat up to friction. Maybe some hand calculations will allow you to move forward. Some people do the heat gen via hand calcs and then translate that to a BC in the CFD model.

Thanks, 

[r.lawrence88]

Hi Matt,

I understand that it cant be used for heat in that way. We have had a consultancy CFD project done and i am just trying to recreate the results that they were given. they focused on an isothermal system with Boundaries stated above and calculated work done to the fluid and used it for a proxy for heat transfer. Basically they took work done and worked out how much internal energy was going into the fluid.

How would I set up my CFD model to output work done / internal energy?