[Valpress]

Hello everyone,

I would like to ask a question corcerning a posibility to simulate drying process of a fiber material (i.e cotton towel). I am doing my PhD. and investigate this possibility in ANSYS Fluent 13.0

At the present I set the simulation this way:

1. geometry - there is a pipe and porous volume inside. Hot air (mass flow) is defined at the inlet. Porous volume should have represented i.e. cotton towel.
2. I used Multiphase-Eulerian model, 3 phases (air, water, vapor); there is mass transfer-evaporation set in phases interaction dialog box. It looks that evaporation works quite well.
3. water is PATCHed in porous volume cell zone as volume fraction after initialization

By this steps i would want simulate drying but I am not sure whether I can use porous volume (porous zone) for this. Problem occures when hot air is passing through porous zone - than the water escapes this porous zone after short time (circa 0.8 sec). Simulation is set as transient (time step 0,001sec). Porous coeficients was set the way reflecting real test (pressure drop in the pipe).

Has anybode some information about possibilities to simulate this case. i think that main goal is find out how to "hold" water phase in porous zone longer time and evaporate it- means do not allow to escape it immediately with passing air.

Unfortunatelly I can not use Ansys customer portal to ask the experts because our university doesn´t have maintenance.

Thanks a lot for any information.

[Cantstandit]

you could use "fixed value" option in the cell zone conditions, and set the liquid water phase velocity components to 0, this way it will always stay in the porous zone.

[Valpress]

Dear Cantstandit,

thank you very much for your advice. I should invite you for beer. This "fixed values" seems to be helping. For at the first sight it looks that I am simulating drying of a fiber (which is representing by porous zone) in Fluent.

Thanks, thanks, thanks

[jacek]

1. Geometry - simple pipe, part of it as porous, water inside,
2. Mulitphase Eulerain model, air, water-liquid, water-vapour,
3. Water is patched in porous volume after initizlization
4. Liquid water is fixed in porous volume U=V=Y=o m/s to prevent it from hovering with air

but when calculation is started I got error:

# Divergence detected in AMG solver: pressure correction -> Turning off correction scaling!
# Divergence detected in AMG solver: k -> Increasing relaxation sweeps!

Error: Divergence detected in AMG solver: k
Error Object: #f

Interrupting...
Done.

What is the problem with divergence?

[niloogh]

Originally Posted by jacek

1. Geometry - simple pipe, part of it as porous, water inside,
2. Mulitphase Eulerain model, air, water-liquid, water-vapour,
3. Water is patched in porous volume after initizlization
4. Liquid water is fixed in porous volume U=V=Y=o m/s to prevent it from hovering with air

but when calculation is started I got error:

# Divergence detected in AMG solver: pressure correction -> Turning off correction scaling!
# Divergence detected in AMG solver: k -> Increasing relaxation sweeps!

Error: Divergence detected in AMG solver: k
Error Object: #f

Interrupting...
Done.

What is the problem with divergence?

hi jacek
i have a problem such as you.(Error: divergence detected in AMG solver)
did you solve your problem?
if it,s possible for you help me in this case

[jacek]

My problems with divergence stopped when I remove boundary layers from the pipe.
But more a important thing was setting a correct temperature of boiling in normal conditions - 380 K, slightly over boiling point. Then there were no problems with convergence when boiling really runs.

But this gave me suspicion if one has to to with boiling or evaporation in Fluent.
If it should be evaporation of water to surrounding air, evaporation take place in every given temperature. But there must be saturation point - max water vapour partial pressure of vapour in given temperature in humid air (relative saturation = 100%).

For now I can't see saturation in Fluent, however this temperature is in phases mass interaction so named. If it were evaporation Fluent would have saturation curve P-t and humid air balance implemented.

[niloogh]

Originally Posted by jacek

My problems with divergence stopped when I remove boundary layers from the pipe.
But more a important thing was setting a correct temperature of boiling in normal conditions - 380 K, slightly over boiling point. Then there were no problems with convergence when boiling really runs.

But this gave me suspicion if one has to to with boiling or evaporation in Fluent.
If it should be evaporation of water to surrounding air, evaporation take place in every given temperature. But there must be saturation point - max water vapour partial pressure of vapour in given temperature in humid air (relative saturation = 100%).

For now I can't see saturation in Fluent, however this temperature is in phases mass interaction so named. If it were evaporation Fluent would have saturation curve P-t and humid air balance implemented.

tnx jacek
i think my problem is related to my periodic bcs.
and still i didn,t fix it

[sanmi]

weldone sir.....m also working on simulation of fish drying....pls can you give me guidelines to follow. thank you sir, i will be glad if get your reply as soon as possible

[patilyogesh]

Originally Posted by jacek

My problems with divergence stopped when I remove boundary layers from the pipe.
But more a important thing was setting a correct temperature of boiling in normal conditions - 380 K, slightly over boiling point. Then there were no problems with convergence when boiling really runs.

But this gave me suspicion if one has to to with boiling or evaporation in Fluent.
If it should be evaporation of water to surrounding air, evaporation take place in every given temperature. But there must be saturation point - max water vapour partial pressure of vapour in given temperature in humid air (relative saturation = 100%).

For now I can't see saturation in Fluent, however this temperature is in phases mass interaction so named. If it were evaporation Fluent would have saturation curve P-t and humid air balance implemented.

Hii Jacek,
please tell me how to remove boundary layer.