[shotta287]

Hi all,
I have a simple model in which I have a tube with blood flowing through. At the outlet I have applied a pressure (suction). The goal is to simulate the suction of a blood clot. The clot will be patched in as a region. The clot has the same properties as the blood but just higher viscosity and density. Blood is primary phase and clot secondary. When I try to patch in a region for the clot, there is nothing under 'registers to patch'. See the attached image. Any help would be much appreciated.

[jpina]

You need to have different a part for each zone in ANSYS Design Modeler. And once in Fluent in Mesh Interface you'll have to generate a mesh interface with "match".

[shotta287]

Originally Posted by jpina

You need to have different a part for each zone in ANSYS Design Modeler. And once in Fluent in Mesh Interface you'll have to generate a mesh interface with "match".

Thanks for the suggestion. Could you elaborate on that please? What do you mean by different part for each zone? Please excuse my amateur questions, I am a beginner to the software!

[jpina]

Have you used design modeler for creating the CAD?

Create the model for what you want initially to be air.

Freeze it (look at the options)

Create the model for what you want initially to be blood.

Freeze it.

Down you should have two parts.

At the afternoon I could send you a snapshot

[CeesH]

You can also go to region > adapt > set the extent of the region > mark

There should then be a message 'X cells marked for refinement' or so. Go to initialize, patch and it should pop up there. Of course this way of patching is limited to simple shapes (box, sphere), so if you want to initialize a complex shape, you are better of following the instructions above.

[shotta287]

Originally Posted by jpina

Have you used design modeler for creating the CAD?

Create the model for what you want initially to be air.

Freeze it (look at the options)

Create the model for what you want initially to be blood.

Freeze it.

Down you should have two parts.

At the afternoon I could send you a snapshot

I made this simple model on designmodeler in order to see if the patching region method works. I have more complex geometry waiting to be imported in from Solidworks!

[shotta287]

Originally Posted by CeesH

You can also go to region > adapt > set the extent of the region > mark

There should then be a message 'X cells marked for refinement' or so. Go to initialize, patch and it should pop up there. Of course this way of patching is limited to simple shapes (box, sphere), so if you want to initialize a complex shape, you are better of following the instructions above.

This is the method I am currently trying. After marking a cylindrical region, when I try to patch it, nothing shows up under 'registers to patch'.

[CeesH]

Did you click 'mark' and not 'adapt'?
Second, does the interface show you a message that a certain number of cells has been marked?

[shotta287]

Originally Posted by CeesH

Did you click 'mark' and not 'adapt'?
Second, does the interface show you a message that a certain number of cells has been marked?

'Mark' first, then 'adapt'. Yes a message does come up saying 'X number of cells marked for refining, 0 marked for coarsening'. For some reason this region is not showing up under registers to patch.

[CeesH]

You don't have to click adapt, only mark. Does it work then?

[shotta287]

Originally Posted by CeesH

You don't have to click adapt, only mark. Does it work then?

Thank you, that works! Can't believe such a simple thing was causing me to stall. I have another question which I hope you may be able to answer; I have geometry waiting to be imported into designmodeler for some cfd analysis. Is it better to replicate this geometry in designmodeler or is it fine if I just import it over? Many thanks once again

[CeesH]

both should work; as long as you don't have to make major changes, I'd say import is the quickest way.

[shotta287]

Originally Posted by CeesH

both should work; as long as you don't have to make major changes, I'd say import is the quickest way.

Thank you for your help, much appreciated. Is there any way I can set up FLUENT to tell me the removal time of the clot?

[CeesH]

I'm not aware of any direct method to calculate that. What you can do is write a UDF that calculates the droplet center-of-mass (execute-at-end macro should work, calculates it at the end of each timestep); from the center of mass vs time you have the position, and can also calculate velocity and acceleration.

(if you don't want to use UDF, you can also export the phase-field every x timesteps and do the calculation in post-processing, i.e. MATLAB. but for large 3D simulations this becomes time- and memoryconsuming really, really quickly)

[shotta287]

Originally Posted by CeesH

I'm not aware of any direct method to calculate that. What you can do is write a UDF that calculates the droplet center-of-mass (execute-at-end macro should work, calculates it at the end of each timestep); from the center of mass vs time you have the position, and can also calculate velocity and acceleration.

(if you don't want to use UDF, you can also export the phase-field every x timesteps and do the calculation in post-processing, i.e. MATLAB. but for large 3D simulations this becomes time- and memoryconsuming really, really quickly)

Is it possible to set up a phase 2 volume fraction monitor against flow time?

[CeesH]

yes, if you run transient, you can use surface monitors (for point, line or surface-average monitoring), or volume monitors (to get the volume integral or average over a volumetric section, or the whole volume).

Set up either of them on the locations you want, set them per timestep instead of per iteration, and plot flow-time or timestep on the x-axis, and you are good to go!

[shotta287]

Originally Posted by CeesH

yes, if you run transient, you can use surface monitors (for point, line or surface-average monitoring), or volume monitors (to get the volume integral or average over a volumetric section, or the whole volume).

Set up either of them on the locations you want, set them per timestep instead of per iteration, and plot flow-time or timestep on the x-axis, and you are good to go!

Thanks for your reply I am running transient with non-iterative time advancement. What volume monitor settings would be ideal to see the clot removal time? I don't understand the meaning of volume integral and so forth. Also is it possible to save the graph for viewing later on? At the moment it is only printing to the console so i see the graph being produced in real time as the solution is being calculated. Please see the attached screenshot. Is there a way to plot volume fraction phase 2 from 1-0 against flow time? Thank you again!