Learning Forum

[srk2757]

Hi everyone,

I am new to ansys fluent. I am doing my thesis on ansys fluent simulations. My task is to model a continuous fiber moving in water with particle injections. I am thinking of DPM model for simulating my problem. 

I need suggestions and future tasks or problems and guidance from you...It would be really helpful for me to finish my thesis...

Any idea or discussions are warmly welcomed..

Thank you.  

[Rob]

What do you actually want to get from the model? 

[srk2757]

Sorry to ask you this question. I don't know how silly it may sound. My task is to analyse the results. But I don't know what does that mean??

Could you please let me know what actually can I get while simulating a dpm Model in fluent.

[Rob]

Do you need to show the fibre move, how is the motion defined? What do the particles do, are they dragging flow with them, or causing the fibre to move or grow? 

CFD is used to answer a "question". If the question is vague or not well defined you can't build the model. Consider how you'd build an experimental rig and apply similar techniques to the CFD simulation.  Once you've planned it all out you're ready to open software, build the geometry, mesh and set up the solver. 

[srk2757]

Yes,I need to show the fibre movement.  In reality the particles immerse in water (its a slurry form of water and powdered thermoplastic polymer). For that I want to a create water medium first and then will start to inject particles from three or four sources. I don't know whether it is correct or not. Right now that is my thought. 

 

[Rob]

Someone's asked a similar set of questions before: have a look on here as that'll give you some ideas for an approach.  Also read up on the Discrete Phase Model (DPM) and Eulerian Granular models: you need to decide which is most suitable. 

 

[srk2757]

Thank you...I will read up the DPM and Eulerian Granular models. You told me that somebody has asked similar set of questions before. Could you please forward the thread of that post??

 

[srk2757]

I will give you the picture of my  model. I will go with DPM modelling in the set up.

I need you to evaluate whether  my model is right according to my assumptions.

I want to build a continuous moving fiber in a fluid of discrete particles.

For this I considered a rectangle, 'base' being the fiber. In boundary conditons, I have given moving wall conditions to the base thinking it will act as a moving fiber.

The other three boundaries are considered as left side being the pressure inlet, right side being the pressure outlet and top side being the outbound.

I just want to know whether I can give the left and right edges of the rectangle as pressure inlet and outlet or not.

I don't know any pressure details.

# in below picture 

H1- top outbound (fluid which is far away from fiber)

V4- right end of the fluid medium

L5- radius of the carbon fiber( taken as axis-symmetric)

Please guide me how to proceed further...

[Rob]

Assuming you know the flow rate I'd suggest a velocity boundary, it'll be easier to control and most likely more stable.  Now you've read up on the models you can decide which best suits your application, consider model approach & volume fraction. 

I can't remember when the other thread was posted. Use Google to search the community, key words are probably fibre or fiber (in case the originator was American and/or can't spell properly). 

[srk2757]

Thank you for your suggestions...Could you please tell me what are the velocity values should I give for a stored fluid in a beaker where the fiber is moving. I got the velocity and diameter of the fiber...nothing else...and according the model I created, can I assume the fiber is moving continuously ?? If you clarify these details I will jump into particle injection...

Thank you..

[Rob]

If there's no motion then pressure is the best option.  Remember to set gravity & the operating density to that of water. 

[srk2757]

Flow around the moving parts such as rotating blades, impellers and moving walls can render the problem unsteady when viewed from a stationary frame. A moving reference frame simplifies the model by converting the flow around the moving part to a steady-state problem with respect to the moving frame.

 I have come across this in ansys material. Could you please explain me that moving reference frame simplifies my model or complicates it.

My model too has similar situation where my fiber is moving in the flow around it and we will introduce particles in it.

and from the old post, I will consider your suggestions. but could you please let me know what would pressure inlet and outlet values.

Thank you

[Rob]

Reference frame is used if we want to average the effect of motion. In this case it will depend on whether you want to actually move the fibre: I suspect you'll need moving deforming mesh. What may help in 2019R2 (and improved in R3) is the availability of a limited capability Mechanical solver within Fluent so we can do FSI within Fluent. 

You need to figure out boundary values: it's entirely dependent on your set up.  

[srk2757]

Hi all,

It's been a lot of time..Here I want to give you an update on my thesis problem wanted to know your valuable suggestions.

Your previous suggestions were really helpful. 

Looking into the problem,

now we are considering a wall which is fiber and creating a particle cloud around it. Here the fiber is grounded and the particles are charged so that particles stick on fiber because of polarity.

Now, I am thinking of two approaches.. Multiphase modelling(which is of 2 phase model, air and particles) and Eulerian granular approach.,

I will give the fiber some velocity and let particles sit on fiber with gravitational force.

Here, I need to know your suggestions about boundary conditions and how to set up this kind of model

and another one is I will consider your moving deforming mesh in which, the fiber is constant wall and the total mesh around it is moving with some velocity.

Please share your thoughts on it and it will be of great help.

 

Thank you

[Rob]

Both Eulerian & DPM can be made to work: what particle loading do you have, and how do the particles interact with the solid when they hit it?  For particle build up, moving mesh may be useful. 

Note, we can't give detailed "how to" answers, but we can give you things to go and work on.