SimScale CAE Forum

MRF vs AMI rotational zone


#1

@PowerUsers_CFD
What is the difference between AMI and MRF rotation? For what I know, in AMI rotation , the rotating zone will physically rotate. Is it the cylinder that is rotating? Or the things inside the cylinder is rotating eg: blades of wind turbine


#2

Hi @100067610!

Check out this link: MRF vs. AMI - cheers!

Jousef


#3

https://www.simscale.com/docs/simulation-setup/advanced-concepts/rotating-zones/
check this document
Rohit.


#4

Just to confirm, so basically AMI rotational zone will give us the overall result. For example, if I am simulating a wind turbine, AMI rotation will let me know the wind turbine’s force averagely for 360 degree. Meanwhile for MRF, it will only let me know the force at one particular angle.

Am i right?


#5

Hi @100067610: your worries are justified.

With axial flow you can use MRF or AMI. MRF will give good approximation of forces on blade(s).

With lateral flow, your Savonius turbine will be impossible to evaluate with MRF, as forces fluctuate widely depending on angle of rotation. This is not related to initial angle of blades in your geometry. MRF will give you an average (you need to measure one blade), but that data may not allow you to find better setup.

So bottom line is that you are wrong but I suggest you find examples of simulation of water turbines (internal CDF, using MRF) or cooling fans within an enclosure, where you have lateral inflow and axial outflow, as it could inspire your research of better Savonius turbine (in a box).

Good luck,

Retsam


#6

@Retsam sir, that means taking octogonal average value is needed in the Savonius wind turbine simulation right? Because the forces output changes a lot when the blade is at different degrees.


#7

Lets say I am using a duct with bigger inlet area than outlet area to boost the velocity of fluid, Venturi effect applied, can i run it with mrf rotation?

And also, should I use internal mesh for that case?


#8

Yes, you can use ‘internal mesh’. Your slots would need to be defined as Velocity inlet / Pressure outlet and it will be easier to mesh. :yum:


#9

Ok sir , one more question, the pressure outlet suppose to be at another end of the duct OR at the outlet of the whole system?
Hold on , i will show you the project link
@Retsam


#10

Pressure outlet is the place where air is leaving the box…


#11

https://www.simscale.com/workbench/?pid=2514634742243032734&mi=spec%3A95b0ee7e-c579-4baf-89bd-546563e42218%2Cservice%3ASIMULATION%2Cstrategy%3A2

@Retsam


#12

@Retsam and also can you check if all my refinements are necessary? I feel like some of the refinements is extra…


#13

Internal CFD would need CLOSED volume containg your Savonus blades (without rod) and MRF zone.

  • You define Velocity inlet pointing to surface of your air duct (it does not exist yet).
  • You define Pressure outlet pointing to other side of the box (now it is open complex shape, you would need to close it, but it can be of complex shape).

Now resuming internal CFD (you can find my post on that topic on the forum):

You need four solids:

  • Box where simulation is run
  • Two solids which are your Savonius blades (those two blades should be “extracted” by boolean operation from the Box")
  • One solid which is MRF zone

Regarding your refinement, most is useless, as you define minimum length to 0.1 m (~1/2 of blade chord). As you use automatic mesh, very coarse, due to curves in the geometry it will create enormous mesh you should avoid to refine in first place. Once mesh is generated you can use ‘Mesh Clip’ button to split the mesh and see how ‘fine’ it is inside. However with your big mesh split operation may fail, alas.


#14

I already had a closed volume for the duct right? Now i need to edit the outlet to make it become a close volume too right? and remove the rod.

I dont really understand what do you mean by Box where simulation is run
@Retsam


#15

Box == closed volume. Place where fluid circulates.

Here is the link to simple example:

https://www.simscale.com/projects/Retsam/karlomrftest/


#16

@Retsam so what i need to do now is

  1. remove the rod
  2. make my design a closed volume
  3. create an internal mesh

#17

Study my simple Internal MRF exammple and mind that you need to do boolean subtract of your blades from your box.

When importing the geometry to SimScale, you will have only two solids: Box and MRF.

Mind that I suggest to simplify your simulation to the core, as you need quicker meshing and simulations in order to progress in understanding of your setup.


#18

what is boolean substruction?


#19

Sir I really dont understand, can I send you my geometry in solidworks and you make an example with it for me to refer? I am really anxious right now… I dont know what went wrong, I am so close to giving up. sigh :cry:


#20

Boolean subtract, I misspelled. You remove solid which is in another solid. That creates a void in that solid. Look at my above mentioned project. Blades are voids in that Box, called Pipe.