Many convergence failures in roof analysis.

[evrodrig]

This is time sensitive so any help you can give in a timely manner would be much appreciated.

I am analyzing a residential roof section to confirm a moisture failure in the plywood. The roof section consists of asphalt shingles, bituminous paper, plywood, 7" open-cell spray foam, 8" air layer (I've been modeling it using the thickest available in WUFI - 6"), and gyp. bd. interior. I first used air without addit'l moisture capacity and was getting hundreds of convergence failures. I was advised to change the air layer to the type with addit'l moisture capacity and that brought the convergence failures down to 28. I'm not sure how reliable these results are with 28 convergence failures. I even reduced the run time to one year, but ideally I'd like to run it over 3 or 5 years. Do you have any advice on how to bring the number of convergence failures down?

I am also running a model with the same components but with a 1-1/2" vented air layer added between the plywood and spray foam and am having the same issue.

I know modeling an air layer as a 'material' is tricky and not always super accurate to what may be happening in real life, but how can I get the closest approximation of what's going on in this roof section related to moisture in the plywood? Or am I expecting something that cannot be achieved with this program? Any help would be much appreciated. Thanks in advance!

[Manfred Kehrer]

Using Adaptive Time Step Control would.
As long as the balances (Balance1, Balance 2) in the calculation results are about the same you should be fine.

Manfred
Official North American WUFI Collaborator
Enjoy WUFI® .... It is easy and complex.

[Christian Bludau]

The number of convergence failures is not very important. More important is that there are acceptable differences in the balances. The differences should not be to high depending on the kind of construction. For a lightweight roof they should be close to zero.

For more inforamtion see also the program help by pressing the "help" button in the state of last calculation window.


@Manfred: There is no Adaptive Time Step Control any more in WUFI Pro 6.

Christian

[evrodrig]

Thank you so much for the quick response. Balance 1 is -0.14 and Balance 2 is -0.16. Would you consider those close?

[Manfred Kehrer]

That should be OK, I think.

[Christian Bludau]

Yes, thats 20g difference over the calculation time. Thats quite good.