'proj1' simulation project by SSHage


#1

I created a new simulation project called 'proj1':

airflow


More of my public projects can be found here.


#2

Very nice model @SSHage!

Can you tell me what kind of simulation this is going to be? Would love to see the end result :slight_smile:

Best,

Jousef


#3

Hi @jousefm,

I’m trying to simulate the airflow within a mask due to breathing by a person. I’ve successfully created a mesh of my model. However I can’t seem to figure out how I should go about defining the boundary conditions. I plan to use a sinusoidal function for the inhaling and exhaling of the person but I cant seem to select the nose as an inlet for the breathing. Do I have to make another surface that covers the person’s nose in my CAD program before I can select it as an inlet surface?
Furthermore, for material used, how do I just select the region between the mask and the person to be air. I can only select the whole model to be made of air, which is definitely not what is intended.

Thanks!

Yours Sincerely,
SSHage


#4

Hi @SSHage!

Indeed you could use a sinusoidal function for the inhaling here and maybe another one shifted by Pi for the mouth to see how the air behaves inside the respirator mask if you like. For the inlet boundary condition I would recommend “closing” the nostrils with faces to define the inlet boundary condition there and for the mouth I would do the exact same step to define the outlet boundary condition. What you can also do is to vary the angle in your study by splitting the velocity components to see the impact. For the material definition this just defines the property of the modeled fluid. For the nostrils/mouth or lips we can employ a no-slip boundary condition. For the turbulence model we can use the k-omega SST model.

Let me know if I could help a bit! Would love to see the end results - and if you need any help feel free to contact me :slight_smile:

Jousef


#5

hi @jousefm,

sorry to trouble you once again. I have a lot of questions this time.
Firstly, I’ve set a sinusoidal function for an inlet at the nostrils and another sinusoidal function shifted by pi for an outlet at the mouth (using velocities for both). However, I get an error message that the continuity equation cannot be resolved. What modifications do I have to make to resolve this error?
Secondly, is there a way I can set my mask to be an outlet for excess air as well? I’ve been encountering errors when I try to set the mask as an outlet. I know there’s a permeability function but I’m not really sure how to use it. Is it possible to set the mask as an outlet without using the permeability function?
Thirdly, the airflow for many of my cancelled (for exceeding simulation time) seems to indicate that the airflow is restricted within the nostrils. Is there an error in how I’ve set up the simulations?
Fourthly, it seems like many of my simulation times are extremely long even for short 4 second simulations. Is this because my human model is overly complicated and leads to longer simulations times? If this is the case, would it be better if I simplified the model?
Lastly, I’ve also tried using a simplified model. However, when I attempt to use a hex dominant automatic mesh, only one of the components gets meshed whereas if I used tet-dominant mesh, the solid is fully meshed. Is there a problem with my CAD model?

Once again, sorry for asking so many questions at once. Thanks once again!!

Yours Sincerely,
SSHage


#6

Hi @SSHage!

Can you tell me which simulation you are referring to? I saw that you uploaded several new geometries. For the permeability we can think about using porous media option if that’s valid in the first place.

Tagging @Get_Barried here to get some ideas from his point of view.

Best,

Jousef


#7

hi @jousefm,

sorry for the confusion. I’ve deleted a few of my meshes to avoid confusion. My first 4 questions are regarding my 2nd model (where I’ve covered the nostrils with a surface) whereas the last question is referring to the 3rd simplified model.
I believe there might be an error regarding how I have set up the CAD model. If I want to simulate the airflow within the mask, am I right to say that I have to create a solid that encompasses the entire internal volume of the mask so that the CFD software is able to compute what is happening within?
My current CAD model involves creating a human model and covering it with a mask. However, the mask just consists of just a single layer. The area between the mask and the face is empty. As such there are no meshes made in that region, thus I do not think computation is possible in the region between the mask and the face. Could this be the reason behind all the errors which I’ve asked in the previous post?

Thank you!!

Yours Sincerely,
SSHage


#8

Hi @SSHage,

I think I might have some ideas but do give me a bit of time to formulate a proper response.

Will post it here!

Cheers.

Regards,
Barry


#9

Hi @SSHage,

Your case is quite unique so I’ll try to give some ideas as best as I can.

Your equations need to basically ensure that whatever flow you’re inputing into the domain, is equally being output especially for incompressible flow. So I guess if you want to find out where in your equations where it is “lacking” per say, just equate them. Maybe also post your equations here so others can take a look.

Unless you want your mask to be like an open outlet, it is rather unrealistic if not difficult. For the case of your simulation certainly using the porous function would be more realistic. Another way would be to set the mask as a “wall” of sorts and have small gaps as outlets, but again, quite unrealistic and difficult to assign due to the non-simple geometry.

Your simulations take a long time because it is transient, this is normal. Simplifying the geometry to decrease the number of nodes will help in this but it is subjective and dependent on what you want to find out, how simple your geometry can be, how coarse your mesh can be and how you need to achieve convergence criteria.

For CFD I would recommend the Hex-Dominant meshes and in particular the parametric one as you have the greatest degree of control over your geometry which is important for this complex and unusual case. Your CAD model looks fine as it is watertight but again, its not so simple and you may want to look into simplifying some parts of it to achieve at least a baseline result before moving on to an actual human face.

Hope I provided some insight! This is a very unique case.

Cheers.

Regards,
Barry