Problem with External Wool Insulation in Lightweight Construction

[apgtechnical]

Hello,

I'm trying to understand an issue I am having with a lightweight permeable construction in WUFI 6.1.

I have been looking at a wall construction using 2 layers Gypsum Board, 150mm Rockwool insulation between studs, OSB 3 sheathing, breather membrane, 150mm Rockwool external insulation, then a ventilated cavity and rainscreen.

Having simulated this assembly omitting the exterior insulation layer the calculation runs without error, however including the Rockwool externally brings about a lot of convergence failures and I am having difficulty understanding why this occurs since the presence of exterior insulation should reduce the temperature and humidity of the internal layers.

I've attached my project file and hope someone can offer some guidance. I have been back and rebuilt the assembly and such but keep finding the same results. This seems to occur regardless of how thick the Rockwool insulation is as even at 5mm I find the same problem. Could this be related to the permeable nature of the whole assembly? I understand WUFI is more commonly used in solid constructions but this is the first real issue I've come across with more permeable assemblies.

A colleague has since tried running an identical simulation using WUFI 5.3 and this appears to be much less problematic. Are there new parameters taken into account with this version that weren't previously? I haven't come across anything in the update notes that might affect this.

It won't let me upload my project file but I have added a Dropbox link containing the file and the weather file we used for the climate Aberdeen, UK.

https://www.dropbox.com/sh/ugc01mby79q1 ... YtRya?dl=0

Thanks

[Christian Bludau]

Dear Paul,

one problem are the "air layers without additional moisture capacity. They show a really bad convergence behavior in case of condensation in that layer.
Further the air layers contain effective values for µ and lambda. So you must not add them to get a thicker layer than given in the database. They have to be extrapolated. See program help and material information for further information.

I checked the behavior in WUFI Pro 5 and 6. There is a noticeable difference in the results, but in both cases the balances are really bad, so you can not
Reason for the differences are the different calculation kernels of WUFI Pro 5 and 6. In case of convergence they show identical results but if there are problems in the calculations the results are more or less inaccurate and this leads to different results in the versions.

You can try the following:

• Adapt the air layers
• Use the air layers with the short names, at least for the surface layers of the cavity
• The rockwool you used has a very low moisture storage function with a strong increase at the end. That also can cause problems in convergence. Try to approximate that on the existing values.
• Use the adaptive time step control in Control -> Numerics, that calculates much better.

I also noticed, that there is a mistake in your adapted thermal conductivity of the rockwool 035 materials. You only changed the the dry value in the basic values. As there is a tabled thermal conductivity, moisture dependent, that is not affected. You have to adapt this table as well.

Christian