Liquid transport coefficient for suction

[WEI LI]

Hi experts,

WUFI help says the liquid transport coefficient for suction describes the capillary uptake of water when the imbibing surface is fully wetted. In the context of building physics, this describes rain on a facade or an imbibition experiment.
Does it mean that the liquid transport coefficient for suction does not need to be considered for the inner layer of the structure?
For example, in my case, I have the metal deck as the outermost layer of the roof, and the EPS board as the insulation layer. Obviously, the EPS board would not affect by the rain at all. In this case, I only have to consider the liquid transport coefficient for redistribution of the EPS board?
Please advise if I am right or wrong.

Thank you so much

[Thomas]

Hi WEI LI,

since EPS is not a capillary-active material, the liquid transport coefficients will be zero anyway and the question does not arise.

In general, the following applies:

If you want to define a material data set for a capillary-active new material and you are sure that in your simulations this material is never directly exposed to rain - because it never rains in your simulation, because the construction is water-tight, or because this material is located so deep in the construction that it is not directly exposed to the rain - you do not need to determine the liquid transport coefficients for suction. It is sufficient to have the liquid transport coefficients for redistribution.

However, the coefficients for suction must not be set to zero; set them equal to the coefficients for redistribution instead.

The reason is: When the component takes up rain, WUFI uses the suction coefficients for all materials in the component. (Because it would be difficult to determine where to use suction coefficients and where to use redistribution coefficients.) Since in WUFI's material data sets the suction coefficients and the redistribution coefficients become very similar for low water contents (as is usually the case deeper in the component) this switching back-and-forth does not really make a difference for the materials which are not directly at the surface.

However, if for a material the suction coefficients are set to zero, capillary transport in this material will stop when it rains, which is not a realistic behavior. The attached example demonstrates this: It contains Baumberger sandstone whose suction coefficients habe been set to zero. The initial water content starts to distribute between the two sandstone layers. A non-capillary-active EPS layer makes sure the Baumberger sandstone is not affected by rain. A thin sandstone layer at the surface allows some capillary water uptake (the switching criterion for WUFI is: suction coefficients are used when there is water uptake at the surface, which would not be the case with the EPS at the surface). Rain is switched on and off via the climate file, and whenever rain is on, the redistribution in the Baumberger sandstone stops. To avoid this behaviour, never set the suction coefficients to zero if the material is capillary-active.

NoSuction.png (6.32 KiB) Viewed 6361 times

Best regards,
Thomas

[WEI LI]

Dear Thomas,

Thanks for your reply, it helps.
For capillary inactive materials, such as EPS, although the liquid transport coefficients are considered as zero, they can still store water liquid. In another word, the water content of the capillary inactive material is in the form of liquid, but they are not moving because of the liquid transport. There is only vapour transport, and subsequently, the condensation and evaporation.
Is that right?
Thanks!

[Thomas]

Yes, the materials have some water content, and the water content at a given location within a material can change because some water is transported to or from that location

• by capillary forces (described by the liquid transport coefficients) or
• by vapour diffusion (described by the µ-value), or
• because of the presence of a moisture source or sink (described by the source strength).

If the liquid transport coefficients are zero and the µ-value is very large and no moisture source or sink exists, the initial water content (with which the simulation started) remains constant during the simulation.

Best regards,
Thomas