Deposition and erosion process

suizo · June 11, 2014, 05:39

Hello,

I am currently working on a "2D" turbidity current problem going down a flume in CFX v15.0. I am running a two phase flow case (water and particles) using the inhomogeneous model in order to have deposition and entrainment of particles at the bottom of the flume (No slip condition). Everything seems to go well, but I have a doubt on how CFX solves the deposition and erosion problem presented by Gary Parker as:

$\partial$ CH/ $\partial$ t + $\partial$ UCH / $\partial$ x = Vs*(Es - Co)

C: Average concentration of the turbidity current
H: Height of the turbidity current
U: Average velocity of the turbidity current
Vs: Settling velocity
Es: Entrainment rate
Co: Concentration at 0.05*H

I have read the User's Guide, but nothing is mentioned there.

So, my question is: What equation does CFX solve in order to integrate the deposition and erosion process presented above into the simulation?

Thanks in advance for your help.

ghorrocks · June 11, 2014, 06:31

CFX does not support the model you describe. You might be able to implement it as a user variable - is this what you intend? It would not be simple as variables like H would require some thought to implement.

suizo · June 11, 2014, 08:03

Thank you for your fast answer ghorrocks. One more question. By saying CFX does not support the model, you mean it does not erode nor deposit?? Or are you specifically saying that Parker's equation is not implemented? In other words, does the inhomogeneous model consider ANY DEPOSITION AND EROSION?

And what is the definition of deposition according to you, sediments would have 0 velocities or very low velocities? In my model, visually, it looks like the current deposits. But when checking velocities, they exist, but with very low values.

Thank you for your kind help and answer.

ghorrocks · June 11, 2014, 08:14

No need to limit yourself to one more question. Feel free to ask as many questions as you like, that is why the forum is here. When you ask good questions we all learn.

CFX has no deposition or erosion model for a eularian phase. It has models for fluidised beds, maximum packing factors, particle to particle collisions and many more - have a look in the documentation to see what is modelled.

I do not know what is causing the essentially zero velocities you are seeing - it is your model and you have not described what you are modelling. If you want help in understanding what you are seeing please post an image of your geometry and the results, and the CCL file.

suizo · June 11, 2014, 09:25

Thanks again for your response. There is something I do not understand well though... Is there a settling velocity calculated by CFX for the particles? I tried modeling a small 2D recipient with dimensions: 30x30x0.5 cm^3. The inhomogeneous model was used and a 2% initial volumetric concentration was given with no inlets or outlets just to see if particles (with a mean diameter of 80 micrometers) settle. A high drag coefficient was used (Cd=2500).

The result showed an increase of concentration at the bottom of the ''reservoir'' and a decrease at the top after 39 seconds, however with particle velocities around -0.000120746976 m/s vertically and -0.00000443535691 m/s horizontally. So settling velocity seems to exist, but how is it calculated? And what is the meaning of this rise of concentration at the bottom (see the link below) if it is not a deposition?

PS. In the homogeneous model, no settling took place and the model was in stable conditions throughout the simulation.

https://www.dropbox.com/s/hcy8c2pymnhkdd0/image.png

ghorrocks · June 11, 2014, 18:52

The settling velocity is a function of the particle mass, gravity, particle drag and the density of the continuous phase. You do not set the settling velocity explicitly, it is a function of these parameters and CFX models those parameters directly.

Your comment about the homogenous model not modelling settling suggests you do not understand the basics of multiphase modelling. Of course you do not get settling with a homogenous model as it assumes there is no inter-phase slip, and settling is a form of inter-phase slip. I recommend you read the CFX documentation to understand homogeneous and inhomogenous multiphase models.

As for your "settling", I bet that is just particles accumulating in the bottom control volume. I suspect it will not build up a bed any thicker than 1 element (until you put in a maximum packing factor model). If you change mesh size your "bed" thickness will change as the thickness of the element will change. So it is settling, but it is not building up a bed.

June 11, 2014, 05:39	Deposition and erosion process	#1
suizo New Member Diego Join Date: Jun 2014 Posts: 3 Rep Power: 11	Hello, I am currently working on a "2D" turbidity current problem going down a flume in CFX v15.0. I am running a two phase flow case (water and particles) using the inhomogeneous model in order to have deposition and entrainment of particles at the bottom of the flume (No slip condition). Everything seems to go well, but I have a doubt on how CFX solves the deposition and erosion problem presented by Gary Parker as: $\partial$ CH/ $\partial$ t + $\partial$ UCH / $\partial$ x = Vs(Es - Co) C: Average concentration of the turbidity current H: Height of the turbidity current U: Average velocity of the turbidity current Vs: Settling velocity Es: Entrainment rate Co: Concentration at 0.05H I have read the User's Guide, but nothing is mentioned there. So, my question is: What equation does CFX solve in order to integrate the deposition and erosion process presented above into the simulation? Thanks in advance for your help.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problems in the UDF for DPM erosion	Messi	Fluent UDF and Scheme Programming	0	January 23, 2014 05:29
Sediment transport and deposition	Chaudhry	FLUENT	0	April 10, 2013 09:41

June 11, 2014, 06:31		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	CFX does not support the model you describe. You might be able to implement it as a user variable - is this what you intend? It would not be simple as variables like H would require some thought to implement.

June 11, 2014, 08:03		#3
suizo New Member Diego Join Date: Jun 2014 Posts: 3 Rep Power: 11	Thank you for your fast answer ghorrocks. One more question. By saying CFX does not support the model, you mean it does not erode nor deposit?? Or are you specifically saying that Parker's equation is not implemented? In other words, does the inhomogeneous model consider ANY DEPOSITION AND EROSION? And what is the definition of deposition according to you, sediments would have 0 velocities or very low velocities? In my model, visually, it looks like the current deposits. But when checking velocities, they exist, but with very low values. Thank you for your kind help and answer.

June 11, 2014, 08:14		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	No need to limit yourself to one more question. Feel free to ask as many questions as you like, that is why the forum is here. When you ask good questions we all learn. CFX has no deposition or erosion model for a eularian phase. It has models for fluidised beds, maximum packing factors, particle to particle collisions and many more - have a look in the documentation to see what is modelled. I do not know what is causing the essentially zero velocities you are seeing - it is your model and you have not described what you are modelling. If you want help in understanding what you are seeing please post an image of your geometry and the results, and the CCL file.

June 11, 2014, 09:25		#5
suizo New Member Diego Join Date: Jun 2014 Posts: 3 Rep Power: 11	Thanks again for your response. There is something I do not understand well though... Is there a settling velocity calculated by CFX for the particles? I tried modeling a small 2D recipient with dimensions: 30x30x0.5 cm^3. The inhomogeneous model was used and a 2% initial volumetric concentration was given with no inlets or outlets just to see if particles (with a mean diameter of 80 micrometers) settle. A high drag coefficient was used (Cd=2500). The result showed an increase of concentration at the bottom of the ''reservoir'' and a decrease at the top after 39 seconds, however with particle velocities around -0.000120746976 m/s vertically and -0.00000443535691 m/s horizontally. So settling velocity seems to exist, but how is it calculated? And what is the meaning of this rise of concentration at the bottom (see the link below) if it is not a deposition? PS. In the homogeneous model, no settling took place and the model was in stable conditions throughout the simulation. https://www.dropbox.com/s/hcy8c2pymnhkdd0/image.png

June 11, 2014, 18:52		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	The settling velocity is a function of the particle mass, gravity, particle drag and the density of the continuous phase. You do not set the settling velocity explicitly, it is a function of these parameters and CFX models those parameters directly. Your comment about the homogenous model not modelling settling suggests you do not understand the basics of multiphase modelling. Of course you do not get settling with a homogenous model as it assumes there is no inter-phase slip, and settling is a form of inter-phase slip. I recommend you read the CFX documentation to understand homogeneous and inhomogenous multiphase models. As for your "settling", I bet that is just particles accumulating in the bottom control volume. I suspect it will not build up a bed any thicker than 1 element (until you put in a maximum packing factor model). If you change mesh size your "bed" thickness will change as the thickness of the element will change. So it is settling, but it is not building up a bed.