[namandoshi]

I am simulating desert cooler in cfx,-steady state

i have a desert cooler inside a room (just a plane section of 0.5*0.5m), (there is no cooler body)

and a vent as opening with the ambient condition of air (temp, mass fraction)

Q1. now how should i give the condition of the cooler?

Q2. how do i give the source for the cooler (to account for the mass of water added in cooler to air)?


velocity needed =5m/s (normal to plane direction)
temp of air(output of cooler)- known
mass fraction of air - known

I have calculated mass flow of total air(dry air+moist)= velocity (5m/s)*area of cooler(0.5*0.5m) *(density of moist air at that output condition of cooler)


please help

[Gert-Jan]

This is completely unclear to me. What is a desert cooler?
Do you want to cool the Sahara? Or an icecream? Then it has to a dessert cooler.

What question are you trying to answer using CFD? Velocity profile? Temperature? Condensation? Moisture concentration?

And a sketch would help a lot...

[ghorrocks]

In Australia we know this as a Coolgardie safe - https://en.wikipedia.org/wiki/Coolgardie_safe

If you want to model the cooling action by simple empirical relations then this is a simple simulation where you can use a source term (or thermal conditions on a wall) to model the cooling.

If you want to make it a little more complex, but still reasonably straight forward you can also track the water vapour in the air.

But if you want to model the evaporation process directly, the water flow in the fabric and all those details this will be a very challenging simulation to do in CFX. Definitely not for beginners.

So back to Gert-Jan's question: "What question are you trying to answer using CFD?"

[Gert-Jan]

I only know dessert coolers. And then call them fridge.

[namandoshi]

Originally Posted by Gert-Jan

This is completely unclear to me. What is a desert cooler?
Do you want to cool the Sahara? Or an icecream? Then it has to a dessert cooler.

What question are you trying to answer using CFD? Velocity profile? Temperature? Condensation? Moisture concentration?

And a sketch would help a lot...

Ans1.
An evaporative cooler (also swamp cooler, desert cooler and wet air cooler) is a device that cools air through the evaporation of water. it os used to cool room. like air conditioner but uses water to cool air and then this cool air is sent out inside room.



ans2.
i want to get steady condition inside room velocity at each point, water mass fraction, and temperature distribution.

[namandoshi]

Originally Posted by ghorrocks

In Australia we know this as a Coolgardie safe - https://en.wikipedia.org/wiki/Coolgardie_safe

If you want to model the cooling action by simple empirical relations then this is a simple simulation where you can use a source term (or thermal conditions on a wall) to model the cooling.

If you want to make it a little more complex, but still reasonably straight forward you can also track the water vapour in the air.

But if you want to model the evaporation process directly, the water flow in the fabric and all those details this will be a very challenging simulation to do in CFX. Definitely not for beginners.

So back to Gert-Jan's question: "What question are you trying to answer using CFD?"

Yes sir, i want to keep it simple,i am not interested in what happens inside cooler (even i haven't made cooler body), i just want that ambient air to absorb water from the cooler (source to be added) and output of this is sent to room.

[ghorrocks]

That's easy then. Just use a source term to absorb some heat and another source term to add some humidity to the air (which I presume is modelled as an additional variable or maybe a multicomponent mixture).

[namandoshi]

Originally Posted by ghorrocks

That's easy then. Just use a source term to absorb some heat and another source term to add some humidity to the air (which I presume is modelled as an additional variable or maybe a multicomponent mixture).

Thank you sir,
used -ve energy source idea.worked

[namandoshi]

Originally Posted by ghorrocks

That's easy then. Just use a source term to absorb some heat and another source term to add some humidity to the air (which I presume is modelled as an additional variable or maybe a multicomponent mixture).

sir,
Value of mass fraction is coming much more than it can be.
for eg at 100%relative humidity,40'C ,mass fraction of water(Humidity Ratio or specific humidity) is 0.0488.
But I am getting around 0.9 which is impossible.

fluid used air and water vapour (both ideal gas)

[ghorrocks]

Then you are not setting the simulation up correctly. But without any details of what you are doing there is no way for us to know what is wrong. Please attach your output file, and image of what you are modelling and what you expect the results to look like.

[namandoshi]

Originally Posted by ghorrocks

Then you are not setting the simulation up correctly. But without any details of what you are doing there is no way for us to know what is wrong. Please attach your output file, and image of what you are modelling and what you expect the results to look like.

problem description

https://drive.google.com/file/d/1NKc...ew?usp=sharing


cooler back-taken as opening with only normal velocity 5m/s(black arrows)

cooler body(small box) - sources as suggested by you (water (5kg/hour=.0014kg/sec ) + energy(2500 (latent heat of vapourisation of water) *.0014)(kJ/sec) (-ve))
wall- as ambient condition (313k) constant temp

cooler walls(sides walls top and bottom) -adiabatic
both vents as openings with opening temp and mass fraction to that of the environment (313k and .0187 respectively)(blue arrows)

cooler body is the part green highlighted

https://drive.google.com/file/d/19-m...ew?usp=sharing


contour of mass fraction
(updated)
https://drive.google.com/open?id=1v8...ZNMccJ10Xhb2eG

contour of temp
updated
https://drive.google.com/file/d/1skx...ew?usp=sharing

[namandoshi]

Originally Posted by ghorrocks

Then you are not setting the simulation up correctly. But without any details of what you are doing there is no way for us to know what is wrong. Please attach your output file, and image of what you are modelling and what you expect the results to look like.

sir, I think I m doing some mistake in modelling moist air,

I have made a new material
basic setting- variable composition mixture of air ideal gas and water vapour ideal gas.
thermodynamic state gas

mixture properties
ideal mixture

in default domain -
imported my new material
basic settings- continuous fluid

fluid models
heat transfer -thermal energy
turbulence- sst

transport eq -on for water ideal gas
air ideal gas- constrained

Should i change to water vapour 25'C?

[ghorrocks]

You did not appear to attach an image of the mass fractions.

Please attach your output file.

[namandoshi]

Originally Posted by ghorrocks

You did not appear to attach an image of the mass fractions.

Please attach your output file.

https://drive.google.com/open?id=1v8...ZNMccJ10Xhb2eG
mass fraction

[ghorrocks]

That image is showing a maximum mass fraction of about 0.1, not 0.9. Still, 0.1 is higher than saturation so I suspect that is higher than you are looking for.

Are you sure your mass fraction rate is correct? For instance are you sure all the water is evaporated in the fabric? Most of the coolers I have seen like this have water dripping off the bottom - which suggests not all the water evaporates. Also, could some water be swept away in droplet form and not evaporate?

[namandoshi]

Originally Posted by ghorrocks

That image is showing a maximum mass fraction of about 0.1, not 0.9. Still, 0.1 is higher than saturation so I suspect that is higher than you are looking for.

Are you sure your mass fraction rate is correct? For instance are you sure all the water is evaporated in the fabric? Most of the coolers I have seen like this have water dripping off the bottom - which suggests not all the water evaporates. Also, could some water be swept away in droplet form and not evaporate?

i am not sure that whatever mass source i have given will completely get evaporate i am assuming the rate to 5kg/hr of water is absorbed by the air inside cooler.

but still, air must not be able to hold mass fraction of .1 at 306K.

or should i solve my prb taking
1)water vapour at 25'C and air ideal gas?
0r
2)both air and water vapour @ 25'C

[ghorrocks]

The different materials will just change the material properties a small amount. So it will not matter too much much of those material models you use. But you should try both and see what difference it makes to check my comment.

You need to think about the physics of this device. If you put more water into the device than the air can evaporate then what happens? Does the air come out saturated and any left over water just drips to the bottom? Then you should make your source terms model this situation. You can define a source term to make the mass fraction 0.0488 (or whatever the saturated mass fraction is), and adjust the heat source term to remove the amount of heat based on the amount of water evaporated. And if you want to go further you should take into account the change in saturation mass fraction due to temperature changes.

As you can see, when you start looking into the details of an apparently simple system it starts getting complex.

[namandoshi]

Originally Posted by ghorrocks

The different materials will just change the material properties a small amount. So it will not matter too much much of those material models you use. But you should try both and see what difference it makes to check my comment.

As you can see, when you start looking into the details of an apparently simple system it starts getting complex.


i have started simulations with above 2 mixutres will post results soon.

[namandoshi]

Originally Posted by ghorrocks

The different materials will just change the material properties a small amount. So it will not matter too much much of those material models you use. But you should try both and see what difference it makes to check my comment.

simulation with air ideal gas and water vapour@25'c

mass fraction
https://drive.google.com/file/d/1AUF...ew?usp=sharing

temp
https://drive.google.com/file/d/1jaY...ew?usp=sharing

seems like absurd results

i have only change materials no boundary conditions were changed.

[ghorrocks]

You can't just try models at random and hope to get the correct answer. If you want to see the difference between the material models, first look at the density, viscosity, specific heat and thermal conductivities the different models generate. The difference models are just changing these properties.

As for simulation accuracy, that is an FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F

In short, you have to validate and verify your model if you want any hope of it being accurate.