Learning Forum

[Simo]

Good morning, 

I am trying to model a capillary water absorption in a solid porous media, such as construction materials, by Ansys Fluent. I did not understand how to insert the two phase fluids (water and air), and to implement the (concrete) like material in the mean that the water will come through. My version is the one for student (Free Version), and the mean is a simple cube 2D 150 mm x 150 mm. Thanks for your attention. 

 

Best regards, 

Simone.

[Rob]

Have a look at the Corey model in Fluent's porous media; turn on the Eulerian Multiphase Model. However, you may find as it's almost a diffusion process that there's a hand calculation that may tell you what you need to know. 

[Simo]

Good evening, 

Thanks for your answer. Where may i find the explanation of Corey model? About that, on internet i could not have found nothing. Yes, what I am trying to look in for is the rising of the wet front. 

[Rob]

I'd start with the documentation (User's Guide > Solver > Cell Zones > Porous Media). Click on Help in the Fluent Solver to bypass the security steps. 

[Simo]

Good morning, 

Yes, i looked for the model in the user guide, and I've assimilated my concrete sample like a packed bed by my experimental data. The fluid still rising a lots, and seems there's no materials in porous media. I calculate the both coefficients alpha (and then 1/alpha) and C2 and i put them in the porous schedule. Maybe the problem is the hydrostatic pressure i give: to simulate that i put a slow velocity to 0.001, could be that correct? My sample is just in contact with water. No pressure applied. 

 

Thanks for the answer. 

Simone. 

[Rob]

You need water to beable to flow, and have some means of getting it into the domain. So, a block sat in liquid may be a good start, possibly with an open channel bc to replenish the water. 

[Simo]

https://www.youtube.com/watch?v=6DREG-OpB5k this example could be suitable with mine? 

[Rob]

Looks reasonable from a quick skim - try it and see! 

[Simo]

I tried to insert a tray with water filled within 5 mm, but with Brooks and Corey model the water front doesn't rise. Whereas if i erase the Brooks and Corey model, the water front rise a bit, more on sides (the model is on the right). No water absorption. Moreover, i set contact angle 0. 

 

[Rob]

What's the mesh resolution like? 

[Simo]

0.0005 m. Otherwise, if i do not consider brooks and corey model, the water front dont rise again but at least it curves beyond the sides.

[Simo]

Hello, 

I have tried instead to make just the sample and to set the water front starting from 5mm. I changed the initial theory model to Brooks and Corey and capillary pressure by Leverette-J by setting Leverette angle to 30°. Actually it does not match with my experimental data, but works. Otherwise, if i set Brooks corey also for capillary pressure, the front does not rise. It keeps still. 

Best regards, 

Simone. 

[Rob]

It's an extra resistance, are we missing something in the model?

[Simo]

I set as upper condition wall (outflow is not allowed for Eurelian flows), and I set two-phases and i modeled air as same as the water. Is that correct? Upon I have found packed bed values C2 and alpha, I inserted that into both air and water phases' schedule. Should air have the same values than water?

 

 

[Rob]

Wall? If it's a wall where does the gas go? The packed bed coefficients (Ergun Equation) are covered in the User's Guide, and doesn't contain fluid properties https://ansyshelp.ansys.com/account/Secured?returnurl=/Views/Secured/corp/v242/en/flu_ug/flu_ug_bcs_sec_cell_zones.html%23flu_ug_using_porous_media

[Simo]

Good morning, 

Yes, I followed the User's guide of packed bed and i calculated from my experimental data alpha -> 1/alpha and C2. And i have inserted that as viscous resistance and intertial resistance, respectively. That for the phase water and air too. should the same values been inserted for air?

[Rob]

Yes - as there isn't any effect for density & viscosity in the documentation. If you were modelling a filter with Eulerian it's a different answer! 

[Simo]

Yes, I am modeling with an Eulerian model. What should it change?

[Simo]

I tried to make different analyses with others materials i got in the lab, and Eulerian model does not work. The front rise equal for every materials I proposed. Nonetheless, Volume of Fluid (multi-phase) contrary works quiet good and reach convergence with my experimental data. How is that possible? Am i missed some values in settings? 

[Rob]

VOF will rise based on the wall, Eulerian multifluid VOF should also respect the porous media. Checking with a colleague. 

[Simo]

I checked a little bit and seemed that Brooks and Corey does not fit with my work. A few examples on YB explained well how to implement Brooks and Corey but far rising in a straw or low section rise where the angle of contact is predominant. Is there any forum page open for a case like that?

Regards. 

[Rob]

Possibly not, having double checked you may need a UDF. 

[Simo]

Hi, 

How can I see Help user guide? 

I did not solve any problem, yet. In my case, concrete is compared with other ligher material, but as soon as the Dp is higher than others, the permeability is higher, and then, the water front does rise more than ligher material (made by finer particles). How can i solve that problem? I checked in a internet guide I have found, but no idea what it missing. Is there a guide into the forum I may take advices from? 

[Rob]

If you click on Help in the solver you'll get the "licenced" version, or use https://ansyshelp.ansys.com/public/account/secured?returnurl=/Views/Secured/main_page.html?hl=f  as we've put most of the documentation outside the security firewall.