DWS and D(θ)

[GlennDeMeersman]

Hello,

I'm trying to convert material data for input and I have the diffusivity in the form D(θ)=D0 exp(Bθ), with B and D0 a constant and θ the volume fraction liquid content, as in Hall and Hoff (water transport in brick, stone and concrete).

Can I put this into DWS directly or will my calculation be inaccurate? (as they have the same units)

Glenn

[Thomas]

Hi Glenn,

the definition of this diffusivity D appears to be same as the definition of what we call the liquid transport coefficients, so you just need to tabulate the exponential formulas you have for your materials.

Since these formulas use the water content expressed as a volume fraction θ but WUFI uses the water content expressed as a moisture density w [kg/m³], you'll have to convert the water contents for a proper tabulation. w [kg/m³] = 1000 [kg/m³] *θ [-] = 10 [kg/m³] * θ [volume-percent].

Regards,
Thomas

[rozumjak]

Hello,

I have a similar one..

I am using material input data from Czech technical university and they use for their simulation "moisture diffusivity (denote by greek letter KAPPA) (m2 s-1)"

see https://www.allbeton.ru/upload/mediawik ... erman_.pdf (section 2.2.2)

From my understanding this is identical with DWW in WUFI. However, when I use A-value to generate the "liquid transport coeff. function" I am getting rather different values than the ones measured in the lab of CTU. See the figures (ACC material)

https://1drv.ms/f/s!AhweJe89QAhLis8PxxBc_72iqbD16g

My question is: 1) Do you agree that the moisture diffusivity from the CTU lab eaquals to DWW? if yes, is it possible that the estimation of "liquid transport coeff. function" by A-value is that inaccurate?
2) What would be better approach for DWS. Estimation of "liquid transport coeff. function" by A-value or using the same values as for DWW?

Thank you very much
Jakub

[Thomas]

From my understanding this is identical with DWW in WUFI. However, when I use A-value to generate the "liquid transport coeff. function" I am getting rather different values than the ones measured in the lab of CTU. See the figures (ACC material)

My question is: 1) Do you agree that the moisture diffusivity from the CTU lab eaquals to DWW? if yes, is it possible that the estimation of "liquid transport coeff. function" by A-value is that inaccurate?

Hi Jakub,
in their description of the measurement of the moisture diffusivity kappa (section 2.2.2), the authors describe that they evaluated 'moisture profiles', but they do not explicitly say whether these profiles were obtained in an absorption experiment (where one face of the test specimen is exposed to a supply of liquid water for absorption) or in a redistribution experiment (where the supply of water is cut off and only the redistribution of the absorbed water is observed). Depending on which of the two possibilities applies, the measured kappas are either DWS or DWW.

Later, in section 3.2, they compare the measured kappas to separately measured water absorption coefficients A. The fact that they directly compare the kappas and the A's seems to imply that the kappas have been obtained in an absorption experiment, similar to the measurement of the A's. If that conclusion is correct, the kappas are DWS.

The diffusivity curves shown in Fig. 2 reach maximum values between about 2e-7 and 5e-7 m²/s. In the "aerated concrete" data sets in WUFI's Fraunhofer-IBP material database (with bulk densities 400, 500 and 600 kg/m³), the DWS reach maximum values between 1e-7 and 2.7e-7 m²/s, and the DWW reach maximum values between 2.2e-8 and 4e-8 m²/s. This suggests that the kappas in Fig. 2 may be DWS, although this argument is not strictly decisive, because we do not know how comparable the materials in the database and in the paper really are.


The authors themselves acknowledge (page 355) that the A-values they measured do not agree well with the detailed moisture diffusivity curves kappa they measured (they speculate this may be caused by the different orientations of the specimens in the two experiments). It is not surprising then that WUFI's DWS or DWW curves estimated from these A-values don't agree well with the measured kappa curves, either.

You can easily determine the A-value which corresponds to a given DWS curve by simulating a water absorption experiment in WUFI.
* Use the attached climate file 'saug.kli' which applies a constant climate of 20°C, 100 % RH and a large amount of rain (that is, liquid water for absorption) to the left side of the material.
* Apply a large sd-value to the right side of the material to exclude any moisture exchange across the right-hand surface.
* Set the initial conditions to 20°C and 0 % RH (i.e. to a completely dry material)
* Run the simulation for maybe 24 or 48 hours or so to determine the amount of water taken up during that period.
* It may be useful to adjust the thickness of the material and/or the simulated period of time to make sure that the water front travels through a large part of the material but does not reach the right-hand side. (Look at the film to see what happens.)
* Determine the absorbed amout of water in kg/m² by consulting the "Status of Last Calculation" dialog, or by looking at the Quick Graphs, or by looking at the Result Graphs, or by using the ASCII export to write the total water content to a text file.
* Divide the absorbed amount of water by the square root of the number of elapsed hours to obtain the A-value in kg/(m² h^0.5), and divide by a further factor 60 to obtain the A-value in kg/(m² s^0.5).

For example, a 0.1 m thick specimen of the "Aerated concrete (density: 600 kg/m³)" from the Fraunhofer-IBP material database absorbs 23.57 kg/m² in 24 hours, the A-value of this material is therefore 23.57/sqrt(24)/60 = 0.080 kg/(m² s^0.5).

You can use this to check the quality of WUFI's DWS coefficients estimated from the A-values, to determine an unknown A-value from a given set of DWS coefficients, or to find a set of DWS coefficients which correspond to a given A-value.


So, in the case of the paper you cited, both the A-values and the kappas appear to refer to water absorption conditions and thus to WUFI's DWS. I'm not sure which of the two should be preferred for a simulation, because we don't know the reason for the disagreement between these two. If you decide to use one of these for the DWS in WUFI, you still need the DWW. In such a case, and if no further information about the redistribution coefficients is available, we simply fill the DWW table with the DWS values, divided by ten.

Kind regards,
Thomas

[rozumjak]

Thank you very much Tomas.

This was amazingly helpful