I have been struggling for a number of years with composite wall junctions, and am very careful about setting skin priorities correctly. However, in many instances I’ve had to resort to using patches - mostly where 3 different walls form a T-junction.
For me, this a laborious solution, and almost outweighs the benefits of using composite walls. For example in a current project, although I have really only 5 different external wall types, and 3 different party wall types, I have no less than 59 different patches to cover (literally!) all the various configurations that occur at wall junctions. To be clear, all these patches are unique, and many occur several times in the project; so I have developed patches which are actually modules, in an attempt to reduce the editing load if a composite wall requires to be changed.
I have posted on this issue previously – but it appears that nobody suffers from the same problem, which made me think there might be an easier way round it.
Recently I have tried splitting each composite into 2 parts longitudinally, so that at a T-junction each part composite only forms an L-junction with its counterpart. Beyond the junction, the wall reverts to a full composite for ease of inserting openings. The difficulty has been in determining where to make the longitudinal split – not necessarily along the cavity, which you would think would be the most obvious location. Even then, I have found the odd stray line in the result, but these are minor and not too distracting. It’s still not a particularly elegant solution, because now for each wall type I have up to 3 composites all requiring to be managed, i.e. the original full wall plus the 2 component parts
Hopefully the snapshot illustrates what I mean.
I would be most interested to hear how others work round this problem. I can’t believe that I’m the only one!