Learning Forum

[akban]

Hi  
In my system, background density varies linearly from bottom (high) to top (low). Thereafter a jet of medium density is opened from bottom. So it should move up and at a neutral layer, the jet fluid should spread horizontally.
 
Could you please suggest which model should be used ? And how should I approach ? 
 
 

Regards Ayan

[Amine Ben Hadj Ali]

Please add some screenshots to understand the application. If only one Fluid is included you just need to model the whole thing as single phase flow with variable density.

[akban]

background fluid is water with varying density. At bottom density is 1016 kg/mt3 and at top density is 1001 kg/mt3 .

Water of density 998 .3 kg/mt3 is impinged from bottom as round jet.

[Amine Ben Hadj Ali]

Water density is treated baratropically or are working with different "waters"

[akban]

Hi,

I am working with two different waters.

It is like we have a tank full of water (density stratified as mentioned earlier). Then a constant density water jet is impinged from bottom.

[Rob]

You're using the multispecies model? As the "new"  water enters it'll mix as well as lose momentum so may not spread out as you initially described. 

[akban]

Hi,

I am confused whether I should use species transport model or not. 

Yes, as the "new"  water enters it'll mix with background fluid due to entrainment.

[Amine Ben Hadj Ali]

You need to more clear so that we can help you. How did you defined both waters? Have you done that already?

You can use the species model and define water-1 and water-2 as components.  If you want to discard diffusion at all you can even work with a multiphase model (The mixture model) and assign phase-1 to water-1 and phase-2 to water-2. Or you just use only one component where the density varies due to pressure (Barotropic).

[akban]

Hi,

I am adding two images to clarify my system.

1st image shows a tank full of water (density stratified as shown by different colours). Then a constant density water jet is impinged from bottom (shown by arrow).

Initially I tried working with VOF model which was erroneous as VOF is for non-miscible fluids. Then I chose Mixture model. I used a udf for linearly varying density of the background fluid. This was phase-1. In experiment, water-salt solution was in a way that at bottom density is 1016 kg/mt3 and at top density reduces to be 1001 kg/mt3.

Then water of density 998 .3 kg/mt3 is impinged from bottom as round jet with a velocity 2 cm/s.  This was considered phase-2.

However, it did not produce any expected result. Moreover, there should be entrainment what I am unable to model. The expected result can be seen in 2nd mage (from lab experiment).

Hopefully, now I am able to explain my problem properly.

[akban]

[Amine Ben Hadj Ali]

Does the density of background water depend on its salinity? I would define the density if water in dependency of its salinity and use species model or track the salinity with additinal scalar. Check this thread for more information /forum/forums/topic/how-to-model-salinity-in-fluent/

 

I would go for species model to account for diffusion without additional modeling which is required for the more general Eulerian approaches.

[akban]

Yes.....the density of background water depends on its salinity. The density stratification is created by varying salinity.

[Amine Ben Hadj Ali]

Check the verification manual in ANSYS help there is reference which has been validated by our Fluid codes.

[akban]

Hi,

I am not able to solve the problem. What is the problem if I use an UDF having height-varying density as Phase-1 and water with constant density as  Phase-2  (The mixture model) ?

[Amine Ben Hadj Ali]

Check the limitation of defining user density in the manual.

[akban]

Hi Amine,

Yeah.... I saw that it is not possible to address every application using UDFs. May be this is one of those applications.

So what do you suggest ?

 

Regards,

Ayan

[Amine Ben Hadj Ali]

I guess all suggestions have ben provided already. Just go through and tell us where you now stuck and details.

[Harri]

Dear Amine,

I am Harri working on the same project. I have been regularly following this thread and suggestions has been implemented. But, I am facing problem in creating a linear stratification(density variation with height) without UDF.  if possible can you please elaborate the suggestion. it would be helpful.

I hope to read your reply soon

Best regards,

Harii

[Amine Ben Hadj Ali]

You need to provide more details: Why is the density stratified? Due to salinity or temperature? As I said DEFINE_PROPERTY for density is actually only for temperature dependency.

 

Your question is now: You want to have that stratification without UDF?

 

[Harri]

Thanks for the kind response, Density is stratified due to the salinity (note-temperature is constant in our experiments). 

Yes, I would like to create a linear density variation without using UDF.

Regards,

Harii

[Amine Ben Hadj Ali]

Use a mixture for your salt water if you can formulate the density of mixture by means of salinity. You fix then the salinity mass fraction the domain (not solving) and patch it along the height with the required values.

 

Or

Switch energy equation, define density depending on Temperature piece wise linear for example or anything suitable, patch whole domain with a temperature profile allowing to have density stratification and do not solve energy equation

[Harri]

Dear Amine, 

I am able to get density stratification without UDF. However, I have one more query.

Since water is injected into the density stratified background, therefore, there must be two phases. In model option, I am using 'mixture model'. However, the solution does not look to be correct.  To be more clear, the fluids are not mixing and getting trapped as expected in the experiments. Can you suggest something?

Best Regards,

Hari

[Amine Ben Hadj Ali]

Using Eulerian model with dispersion.