# Viscosity between continuous and discrete phase in DPM simulation.

• 74 Views
• Last Post 5 weeks ago
anshs posted this 05 February 2020

I'm coupling the discrete phase and continuous phase, but the volume of continuous phase which is going out of the bucket with the discrete phase is too high! Maybe because the friction (viscosity) between the 2 phases is large! What to do in here?

abenhadj posted this 06 February 2020

All is related to drag. Check your boundary condition at top 2

Best regards, Amine

• Liked by
abenhadj posted this 06 February 2020

And surface tension resistance is missing at free surface. You need to code that

Best regards, Amine

• Liked by
anshs posted this 06 February 2020

What is the relation between drag and boundary conditions at the top! The top is Open, i.e Gauge pressure = 0.

Where do I put in drag in the DPM solver? Isn't it inherent?

anshs posted this 06 February 2020

Can I not implement the surface tension by using phase interaction -> Surface tension -> water- air = 0.072 n/m, (because water is inside the container and air is present outside!)

Thanks

abenhadj posted this 06 February 2020

Just asking to check perhaps you have issue with backflow volume fraction. Drag to be selected under DPM panel.

Best regards, Amine

• Liked by
abenhadj posted this 06 February 2020

No you are not accounting for surface tension effects on the particles trying to penetrate the free surface. This is different to what you wrote.

Best regards, Amine

• Liked by
anshs posted this 06 February 2020

Yeah, I didn't find any option of the viscosity of the inert gas, neither the surface tension property in the material panel. may you please help!

Thanks.

anshs posted this 06 February 2020

I have got one more doubt about this simulation, While the plume is going upwards, the upper part of this plume is relatively very large and it becomes smaller if I turn on to larger value in the  'DPM iteration interval'DPM Interaction -> DPM iteration interval in the DPM panel. Why is this so? And what should be the ideal value of the DPM iteration interval?

I've seen posts with relations between the under-relaxation factors and DPM iteration intervals, Is that the only way to judge? What if it does not matches the actual consition?

abenhadj posted this 06 February 2020

No option for surface tension resistance felt by particle. You need UDF.

Your other question you logged another thread. I won't comment on that.

Best regards, Amine

• Liked by
anshs posted this 06 February 2020

No, the other question was about the time-scale constant! Here, I'm referring to something else. Please check this once, if you have time!

abenhadj posted this 06 February 2020

This question is now closed?

Best regards, Amine

• Liked by
anshs posted this 06 February 2020

No, if you would help me with the question on the 'DPM iteration interval'. I'm very thankful for help anyways!

The question is about, the top part of the plume and why it's having a large volume, while in the original experiments we hardly see anything like that! Thanks.

rwoolhou posted this 07 February 2020

You need to look at the flow field. The image a few posts up looks sensible for a jet into slower moving fluid.  Please post an image of the panel you're talking about re DPM iteration interval.

• Liked by
anshs posted this 07 February 2020

You need to look at the flow field. The image a few posts up looks sensible for a jet into slower moving fluid.  Please post an image of the panel you're talking about re DPM iteration interval.

You're right the actual problem is the continuous phase flow velocity only, that's what causing the oil phase at the top to be dispersed more than the usual experimental results!

The above are the kind of results which I expect, but I'm getting very different results!

And also it's looking a bit decent only because I have reduced the Turbulent Dispersion's -> Time scale constant to 0.003, but the default value recommended by fluent is 0.15 while using k-e turbulent model. Don't know where I'm not getting it right! I feel it has something to do with the tracking only!

Had I used the recommended value my results would've been very different! You may check them here: https://studentcommunity.ansys.com/thread/effect-of-time-scale-constant-in-stochastic-tracking-of-dpm-particles/

anshs posted this 11 February 2020

Any help? @rwoolhow

Thanks

abenhadj posted this 11 February 2020

What does happen with the bubble after hitting the free surface in the reference you are showing?

Best regards, Amine

• Liked by
anshs posted this 11 February 2020

The bubble actually, penetrates the slag layer and moves in the slag layer! And finally, the bubbles have to be deleted by using custom made functions as soon as it reaches the free surface!

Thanks!

abenhadj posted this 11 February 2020

And are accounting for all this?

Best regards, Amine

• Liked by
anshs posted this 11 February 2020

Could not get what you are saying?  They have shown that the bubbles are to move up and penetrate the slag region, after that the bubbles move towards the top region ( above the slag region ) where there is ambient air. And at the end of the simulation, the bubbles seem to escape the container!

I don't think they are deleting the bubbles, but they have a 3rd phase above the slag layer where the bubbles would finally reach its fate!

anshs posted this 21 February 2020

My problem is that the water (continuous phase) is getting dragged with the liquid for too long (check figure 'a' )! The gas (modeled by Ar DPM particles) should leave the water after some time and thus creating only a small bulge kind of thing!

anshs posted this 21 February 2020

Micro-level picture! Considering a single bubble! [ Expected results ]

abenhadj posted this 21 February 2020

Are you accounting for surface tension effects? Does the case converge? What about the DPM source terms. Sorry but I cannot provide more guidance here before digging deep and this is not feasible on this platform.

Best regards, Amine

• Liked by
anshs posted this 21 February 2020

Where may I get more in-depth discussions?

abenhadj posted this 21 February 2020

This can be helpful for you:

COUPLED DPM AND VOF MODEL FOR ANALYSES OF GAS STIRRED LADLES AT HIGHER GAS RATES

Best regards, Amine

• Liked by
anshs posted this 21 February 2020

Yeah, but this paper doesn't talk about the interaction of DPM particles with a second phase!

abenhadj posted this 24 February 2020

Which second phase? The paper clearly describes what you are looking for. Of course it does not tell how the particles are eliminated or removed from the domain.

Best regards, Amine

• Liked by
anshs posted this 24 February 2020

Please check this comparison between the two: (a) Represents the paper given by you, (b) My work. Thanks

abenhadj posted this 24 February 2020

Okay I do not see a huge difference: you have three phases they have two phases.

Best regards, Amine

anshs posted this 24 February 2020

Okay, so now you need to look at these results , which I'm getting. I want to disturb the oil phase with my DPM particles!

You need to look at the flow field. The image a few posts up looks sensible for a jet into slower moving fluid.  Please post an image of the panel you're talking about re DPM iteration interval.

You're right the actual problem is the continuous phase flow velocity only, that's what causing the oil phase at the top to be dispersed more than the usual experimental results!

The above are the kind of results which I expect, but I'm getting very different results!

And also it's looking a bit decent only because I have reduced the Turbulent Dispersion's -> Time scale constant to 0.003, but the default value recommended by fluent is 0.15 while using k-e turbulent model. Don't know where I'm not getting it right! I feel it has something to do with the tracking only!

Had I used the recommended value my results would've been very different! You may check them here: https://studentcommunity.ansys.com/thread/effect-of-time-scale-constant-in-stochastic-tracking-of-dpm-particles/

abenhadj posted this 24 February 2020

We are moving along a circle: Surface Tension Force for the particles as resistance force is not in-built. The default value for time scale constant are just default values and so tunable.

Best regards, Amine

• Liked by