Absorption test and drying process

[ITeCons]

Hello,

I am trying to simulate the drying process of a test specimen of stone during two years (200x2000 mm). We considered the following initial conditions: the temperature is T=20ºC and the water content is the free water saturation. We considered the upper surface as a adiabatic border, lateral surfaces with a climate of T=20ºC and HR=60% and the bottom surface with a climate of T=20ºC and HR=90% during two years.

As you can see in the attached file, after two years of drying out (simulation), the outside part of the specimen is dry but not the inside inside part (figures on the right) (20cm_OL_v3.zip).

Then, we only changed the climatic conditions of the bottom surface to simulate that it is in direct contact with an infinite source of water using a .kli file (clima_hum_asc.kli is in each simulation folder):

$WUFI$ 1.1.1991_0.00 - 1.1.1992_0.00
Azimut : 0 º, Inclination : 0 º, Abgeleitet von ibp1991.wet
1000
8760 1000 0 16 1 16 1


But, as you can see in the attached figure, the drying process are very different because in this case the inside part after two years is completely dry (figures on the left) (20cm_OL_v1.zip), why? which is the reason? We just copy the simulation and change this surface boundary condition. There is something wrong in the .kli file?

Is it a simulation error due to material because the slope of the Moisture Storage Function is very high for values near to 100% of HR?

Thanks,

[Christian Bludau]

Hello,

the reason for the different drying process is the rain.
Wufi uses two different liquid transport coefficients for suction and redistribution. This coefficients are changed depending on if there is rain.
So if there is no rain, the coefficient for redistribution is used (is set to zero in your material - so there is only diffusion) and
if there is rain, the coefficient for suction is used (that is defined in your material, so the moisture is transported by suction and diffusion).

Your kli-file is working, you can see that on the water content in the grid elements close to your boundary condition, which is every hour quite high.

Christian

[ITeCons]

Hello,

Thanks, ok. But, what the coefficient of redistribution means? Because, I generated automatically this value in both simulations but the results over time continue to be very different.
The upper part of the test specimen (from 1552 to 2000 mm), which is not affected by absorption water from the lower boundary condition, in the simulation without rain (rain=absorption water) have a maximum water content 1158% higher than the simulation with rain at a certain time step of the simulation.
Our objective is to simulate/validate the drying process of the upper part of this test specimen with different climate boundary conditions with a simulation softwate (WUFI). Then, we want to perform an in-situ experiment, monitoring the weather conditions, simulate the experimental test with WUFI, and compare results.
However, we are not getting consistent results because in this case the drying process of the upper part of the test specimen must be equal in two simulations, with small numerical errors, because weather conditions are the same.

[Christian Bludau]

The liquid transport coefficients for redistribution express the speed of suction which happens during there is no liquid water on the surface. But you have to keep in mind, that the coefficient is switched for all materials in the whole construction if liquid water (rain) is present.

More informations about the two coefficients you can find in the PhD of Dr. Künzel:
Künzel, H.M.: Simultaneous Heat and Moisture Transport in Building Com­ponents – One- and two-dimensional calculation using simple pa­rameters. Dissertation Universität. Stuttgart 1994
http://www.ibp.fraunhofer.de/content/da ... -30731.pdf

Christian

[ITeCons]

Then, if there is rain, it is indifferent the value selected for the liquid transport coefficient for redistribution? Because the moisture transfer is mainly due to suction and not due to redistribution? I performed several simulations considering the .kli file (rain) and it is indifferent consider this coefficient for redistribution equal to 0 or the value automatically generate.
Therefore, I think that I can not automatically generate this value. So, there is a value for the liquid transport coefficient for redistribution that allows me to obtain similar values in both simulations?

I'll try to clarify my doubt with the document you send me.

[Christian Bludau]

Yes, if there is rain, in the whole construction the coefficient for suction is used otherwise the one for redistribution is used.
Maybe a kind of workaround would be to define different materials in x and y-direction. Try to use a material in x-direction where you enter the values for redistribution in the suction table while in y-direction leave the material as it is. This will cause a fast suction in the upper direction, while to the sides the drying will stay slow.
Christian