COSMOSFloWorks

[KlaasDS]

Hy,

I'm trying to simulate an closed container with on the bottom two slow turning screws (See picture). Are there any suggestion how I get the best result?

I've tried to with rotating regions but then I get an maximum velocity of 400 m/s (I've measured a maximum of 1 m/s on the real model).
I've read a lot of information about the rotating region and I'm sure I've applied it correct.

The problem is, I think, rotating regions is only for fast rotating objects.

Some one an suggestion for a better and correcter result ?

Should I take my basic mesh finer or doesn't matter that ?

Kind Regard
Klaas

 

[cbrn]

Hi,
are you absolutely sure the units are consistent? I.e. did you input by mistake "degree/mn" or "radian/mn" instead of "rpm", for example?

Regards

 

[KlaasDS]

Hi,

The screws turn at a speed of 17 rpm (=1,78 rad/s) in the correct direction. The Inlet Volume flow is 3 m³/hr (=0.00083333 m^3/s). Outlet is environment pressure.

Kind Regards
Klaas

 

[cbrn]

Hi,
also consistent density? viscosity? gravity acceleration?
I know my questions sound stupid, but the only reason I see for the strange velocity field results is some mistake in the units. In fact, I know FloWorks has indeed been used for "slow-motion" and not only for pumps etc.
You could also build a very simplified model for which you can also compute, by hands, the theoretic results (though simplified, but that's what you get with FloWorks anyway because as far as I know it doesn't have rotating mesh capability, so the rotor/stator interaction is simplified).

Regards

 

[KlaasDS]

Hi,

There are no stupid questions

If I simulate the model without rotation of the screw (without rotating regions etc.) than the results match the measured results. So I assume gravity, viscosity and density are correct.

An other way I've simulate the flow is by setting the screw as a rotating wall. Could that be an better way to solve the problem ?

Kind Regards
Klaas

 

[cbrn]

Hi,
if the screw is "enclosed" by the appropriate "enclosure" (!), i.e. a cylindrical wall I presume, then it should be OK, just like in the "pump runner" example. In fact, in reality FloWorks only computes the interface BCs (velocity / pressure fields) on the enclosure wall, using the Navier equations including the rotational terms, inside what has been defined as "rotating region", then applies the field BCs back to the stationary region. There is no interaction between static and moving parts.
I now suppose that, in your case, the enclosure of the rotating region is not effectively "closing" the region you want, thus propagating the velocity field outside the screw using the same law, thus giving absurd velocities "far" from the screw.
If you substitute the entire screw with a cylinder of same external diameter, on which you define a flow condition based on vectorial definition, then what happens? It's a try I'd give, just to undirectly check if the problem is the screw or something else.
I'm sorry not to give more direct and precise hints, but I stopped using FloWorks back in 2004...

Regards