Learning Forum

[srath9]

Hello,

 

I am simulating a closed domain (no inlets and outlets) with 3 phases using mixture model. The whole domain is rotating at a constant RPM. The simulation is converging well and the flow profile is established.

 

Objective: To open 2 outlet surfaces and measure the volume flow rate of each phase

 

Question: How do I fix the mass flow rate of mixture (this is defined by a pump used at the outlet)? In Fluent, I can't fix the mass flow rate of the mixture but for individual phases (which I don't know).

 

Thank you!

[Rob]

It's not even that simple as the pump generally works by sucking volume and not mass. An older approach is to use UDFs and sink (source) terms, but that's not overly simple. How well separated are the phases?

[srath9]

Thank you for the reply! 

The pump flow rate is known, so I was thinking to use density of mixture and calculate approximate mass flow rate. 

The 3 phases are: 

1. Buffer (like water) 
2. Air
3. Solids (<1% of total volume with density of ~1077kg/m3 and 12microns in size)

Notes: 

1. The buffer and air are well separated (surface tension ~0.05N/m). 
2. The solids are mostly in the buffer phase, but I haven't specified any surface tension with air. 
3. As the whole domain is rotating (like centrifugal chamber), solids having higher density are pushed to the walls.

[Rob]

I'm assuming the solids concentration is higher at the perimeter as they're pushed to the walls? I think a sink term may be your best option, and you can set that per phase based on what's in the local cells. A UDF is likely the safest option, but Expressions should also be available in the last couple of builds. 

[srath9]

That is correct. The concentration of solids is higher close to the walls.

I haven't used a UDF to define a sink term before and am willing to investigate the approach to know more. Would you be able to share any tutorial/document with regards to this? 

[Rob]

UDFs aren't something I can fully support as they're "specialist knowledge" and therefore fall under things I can't go into detail. However, have a look for DEFINE_SOURCE in the UDF manual; if you use an expression you're going to need to find the volume (mass) fraction of each phase and then create a sink term based on the various fractions and total flow. 

Expressions may be easier but UDFs allow a deriviative term for stability purposes.