Unmet Hours

Material Hygroscopic Properties - EMPD Moisture Transfer model

asked 2018-05-04 17:40:53 -0600

rafael.alonso
933 ●1 ●6 ●12 http://atmoslab.io

updated 2018-05-26 15:01:31 -0600

__AmirRoth__

4446 ●5 ●17 http://bleedinggreenna...

Hi all,

I'm working on the effect of varios materials on the humidity levels inside a room. Aside from the impact of their thermal conductivty and specifc heat, I'm looking into the hygroscopic properties of natural materials [adobe, compressed earth blocks] in comparison to concrete and bricks.

To account for the vapour absorption/release of the walls, I'm working with moisture transfer calculations between the walls and the indoor air in EnergyPlus, I'm using the EPMD model (MoisturePenetrationDepthConductionTransferFunction) and in order to assign the hygroscopic properties to the materials, the MaterialProperty:MoisturePenetrationDepth:Settings object.

MaterialProperty:MoisturePenetrationDepth:Settings,
Gypsum,                  !- Name
6,                       !- Water Vapor Diffusion Resistance Factor {dimensionless}
0.0068,                  !- Moisture Equation Coefficient a {dimensionless}
0.939,                   !- Moisture Equation Coefficient b {dimensionless}
0.0202,                  !- Moisture Equation Coefficient c {dimensionless}
12.2,                    !- Moisture Equation Coefficient d {dimensionless}
0.019,                   !- Surface Layer Penetration Depth {m}
0.073,                   !- Deep Layer Penetration Depth {m}
0,                       !- Coating Layer Thickness {m}
0;                       !- Coating Layer Water Vapor Diffusion Resistance Factor {dimensionless}*

Unfortunately, beyond the material properties specified in the MoistureMaterials.idf dataset that comes with EnergyPlus, I'm finding it very difficult to find the required properties for any other material (in my case earth-based constructions).

Does someone around here know any reliable material database where to find the hygroscopic properties of materials? If it contains earth-based construction would be fantastic! Unfortunately, I'm not an expert on physical properties and some of the terms are a bit abstract.

So far, the most complete databases I've found are below. However, there is no mention to earth-based constructions, and the terms they provide for Hygric properties do not seem to find with the inputs required in E+.

http://www.iea-ebc.org/Data/publicati...

Many thanks!

Rafael

answered 2021-09-15 12:31:33 -0600

GFlechas
247 ●1 ●6 https://www.linkedin.c...

Hi Rafael,

I ran into a similar problem trying to figure out where these coefficients come from. I have figured out how to generate the coefficients based on a material moisture sorption curve for a given material (as seen in the documents to which you link).

For example, if you look on PDF page 22 of IEA final report volume 3 task 3: material properties (your first link), you'll see a table that lists the moisture soprtion/desportion curve points for concrete at the bottom of the page (I've also put a screen shot of this below). If you take the relative humidity and soprtion moisture content and plot them, you'll see the sorption curve for concrete based on relative humidity. The EMPD coefficients are fitted from this data by running a regression that fits a curve to this data based on the function u= aphi^b + cphi^d where u is the moisture content (the vertical axis values in the sorption curve), phi is the space relative humidity, and a-d are the coefficients that fit the curve to the experimental data. I've expanded on the details found in the engineering reference here: https://bigladdersoftware.com/epx/doc.... table of moisture sorption data for concrete

So while you probably won't find EMPD coefficients for earth-based constructions, you will likely find moisture sorption curves for them. A regression to fit a non-linear curve to the data can be done in excel (there are several good tutorials on YouTube on the topic), though there are certainly better tools for the job such as Python, MATLAB, or R depending on your experience and familiarity with these tools.

As an example, I have run a regression on the moisture sorption curve for several different types of cross laminated timber (CLT) based on the moisture sorption curve in WUFI for these materials. I used MATLAB and generated the following coefficients for CLT made from Douglas Fir: a = 0.162 b = 0.6695 c = 0.1572 d = 0.9182

I have a plot as well where I compare regressions for different data cut offs. Wood appears to saturate to high moisture contents at relative humidities above 99% so I looked at a range of cutoffs. Those coefficients above come from the regression that ignores the sorption data above 0.3 kg-h20/kg-CLT. This provides a better fit for the sorption curve at relative humidities below about 99%, but is not going to be accurate above that. plot of regression to fit EMPD curve to Douglas Fir CLT moisture sorption data

So the light blue line is just a plot of the function u = aphi^b+cphi^d using the above coefficients. I hope this is helpful for the next person to encounter this question!

edit flag offensive delete link

Comments

Hi Gabriel,

Thanks for the detail answer - this is very helpful.

We rarely require this level of detail in projects, but it has come up a couple of times on research about earth constructions. I'm glad we now have a way forward with this!

I wonder why E+ incorporated such an intricate method to describe moisture absorption in the EMPD model - instead of using something closer to the tabulated values. Rafael