[fraabe91]

Hi All,

I'm an aerospace student from UWE Bristol and the following problem has to deal with my dissertation.

The overall problem:
Commercial airliner suffer from the problem of condensation within the double skin. I'm going to, finally, analyse the effect of the Zonal Drying System that blows dry air into this gap to prevent condensation.

But right now I need to get the condensation right. What I've done so far is basically that I specified a certain temperature for the cabin wall and the outer skin. This then establishes a flow due to the temperature difference. Up to now I'm running it as multiphase homogeneous model with air as ideal gas and water as dispersed fluid.

heat transfer --> thermal energy
turbulence model --> Shear stress transport

Initialization parameter:
- Cartesian velocity components are all automatic with value and chosen to be zero
- temperature was set to half the difference between both walls
- relative pressure 0


My supervisor told me that I probably need three materials: Air ideal gas, water liquid and water gaseous. But she wasn't quite sure and wanted to think about it.

Is there anybody that can help me on that?

Thanks in advance,
Florian

[ghorrocks]

I would talk to ANSYS Support and see if they have a demonstration condensation simulation for you to use as a starting point.

Another thought: Condensation is formed when the temperature drops below the dew point. So if the purpose of the work is to determine whether condensation forms or not, can't you simply simulate a flow and see if the temperature drops below the dew point? Then the simulation is single phase and you do not need to model all the complex phase change physics. This is MUCH simpler.

[fraabe91]

The point is that I'm supported by germanwings and the company that is producing the zonal drying system. They finally want me to model also the insulation blankets and they know that some water will condense inside of them and stay there. But that will be another problem in the future how to model this. Right now I would like to model the complex phase change as i hope it then is better to visualise the areas of condensation.

[ghorrocks]

I cannot count the number of times I have heard "I'll just model everything right from the start" as a project plan. In my experience that usually leads to months or years of work and no firm outcome at the end.

In my experience, CFD projects are much more successful when they start simple and add complexity as they go along. That is why I suggested starting with a simple single phase model. It should only take a little while to get that working well and means you have something useful to show the client early on (so you have some runs on the board). Quite often you find that these simple models actually tell you everything you need to know to develop a better product. So don't forget that the outcome of the project is a better product, not a fancy CFD model.

Then you add complexity to the model where it adds value. This keeps the project manageable, gives the client a steady stream of results which keeps them happy and allows you to shift the focus of the project as it progresses (this always happens in big CFD projects).

Anyway, that's my 2c worth.

[fraabe91]

Oh sry by the way I've already done simple single phase simulations with a simplified geometry last year. So yes it is a ongoing process.

[ghorrocks]

Good, I am pleased to hear that.

So then can you explain why you are doing a condensation model? What information is it going to tell you that a single phase model cannot?

A quick google search found some links which look interesting:
http://www.cfd-online.com/Forums/cfx...ation-cfx.html
http://www.diva-portal.org/smash/get...FULLTEXT01.pdf

[fraabe91]

Because i would like to see the effect of the zonal drying system in the end. The system basically removes moisture from the air and then blows this into the gap. And I would like to see the differences with and without this system being used. And I would like to get an idea of how much the weight can be reduced as the water is removed.

[ghorrocks]

Your description sounds like it could be done with a single phase model with a humidity model. Where does the condensation fit into that?

Sorry for being a little off your chosen topic with this line of questioning but in my experience it is a very common problem to choose an excessively complex physical model when a much simpler model will work just as well. Did you see the links I posted? They both looked pretty relevant to me. Other than that your best bet is to get condensation examples from ANSYS support.

[Opaque]

Well said by Glenn. It is worth more than 2 cents, perhaps millions. I cannot imagine the amount of effort thrown down the drain just because we want to solve the most complex problem out of the gate.

Let us see what nature teaches us: turn, crawl, stand, walk, run.

Originally Posted by ghorrocks

I cannot count the number of times I have heard "I'll just model everything right from the start" as a project plan. In my experience that usually leads to months or years of work and no firm outcome at the end.

In my experience, CFD projects are much more successful when they start simple and add complexity as they go along. That is why I suggested starting with a simple single phase model. It should only take a little while to get that working well and means you have something useful to show the client early on (so you have some runs on the board). Quite often you find that these simple models actually tell you everything you need to know to develop a better product. So don't forget that the outcome of the project is a better product, not a fancy CFD model.

Then you add complexity to the model where it adds value. This keeps the project manageable, gives the client a steady stream of results which keeps them happy and allows you to shift the focus of the project as it progresses (this always happens in big CFD projects).

Anyway, that's my 2c worth.