Water formation in water_tight materials

[Gregor2019]

Dear Delphin-Team,
I am still modelling underground power cables exposed to different soil conditions. Also I am going to investigate the effect air gaps around the protection pipes have onto the water and heat transport (image below).

The air gaps´ material properties can be found below.

Most importantly I assigned water_tightness to the air material. However, my results show the formation of water in the air gaps due to increasing water contents over time (see below).

I did not expect that as I assumed the air gaps to stay water free due to the assignment of water_tightness. The materials protection pipe, copper and insulation were also assigned water_tightness and they did not show any changes in the water content.

Is there any explanation why the air did behave different from the other materials that are water_tight? So far I could not think of any reason, so maybe I am missing something.

Thank you for any kind of help and have a good day,
Markus.

[fechner]

Hello,

WATER_TIGHT means there is no liquid transport. The vapor diffusion is still there. With this moisture can be transported into the air gap and condensation can take place. If you don't want to have any moisture transport you have to set the VAPOR_TIGHT flag too.

[Gregor2019]

Dear Mr. Fechner,
Thank you for your help first of all.

The problem that arises due to the WATER_TIGHT flag is that the condensating water will stay there without any further movement.
This means, under these conditions the water is going to accumulate in the air gaps (and won´t be sucked out by the adjacent soil material but only vapour diffusion). In my opinion this would result in an unsatisfying and unrealistic represenation of reality (especially as the air is even accumulating against gravity at the upper realms of the air gaps, e.g. the adjacent air gaps at the top corners of the protection pipes).

Today, I tried to fix this problem by assigning the air material moisture storage and transport information, which is easier said than done. Overall, I am not quite sure if I am on the right track and whether there is a solution to my problem. It would be great if condensating water would be able to be transported into the soil again via capillary suction.
The approach I though about is as follows:

1. Moisture Storage: I assign the air material moisture storage data by means of relative humidity, as water retention characteristics for air do not seem reasonable.

2. Moisture transport: Considering gravity, I could calculate a velocity which related to a specific length within the air gaps. This could then serve as water conductivity. This would be an quite arbitrary value but the actual uptake out of the air gaps would ultimately be governed by the soil´s moisture characteristics.

However, so far I was not completely successful, which is why I thought it might be reasonable to ask for an expert´s opinion whether this approach is feasible at all.


Thank you again for your help and have a good day!
Markus.

[fechner]

Hello,

I know thats a problem. Only the air elemnts direct at boundary can have a liquid transport to the adjacent elements.
In reality air has a liquid transport but no capillary system. That means the only driving force is a forced pressure. This can be created by gravity.

If you want to simulate this you need liquid storage and transport properties for air. Escpecially the storage function is important.
Such a storage function should have really low moisture contents over nearly the complete range of capillary pressures. The moisture content should start to rise in the area of capillary pressures near the values that can be created by gravity (some Pa). A sorption isotherm cannot handle such a behaviour. You need a moisture retention curve.
The liquid conductivity should be quiet high starting also forlow moisture contents. Here you have to play a bit with the values.
Such a material characterisation can create solver problems because of the possible big change of liquid fluxes in short times.

[Gregor2019]

Dear Mr. Fechner,
I have to apoligise for my late response. Your suggestion worked perfectly.

Thank you very much.
=)