Hi @Darren_O Great video and model! You’ve clearly done a huge amount of work and thinking about this and it’s really impressive. There is certainly a lot of areas for exploration and improvement in Skylark, and this continuous development is the whole point of WikiHouse, so I really welcome your efforts sweating through the details and how to best apply them to your design.
If you would find it useful, I’d like to offer some feedback and a bit of our own thinking and understanding about the system with respect to your model.
I’m very intrigued by the internal members within the wall cassettes and how they work. I appreciate the concerns about thermal bridging, but my understanding is that this is fairly minimal with narrow ply members. I would be more concerned about the issues your wall design creates for insulating the wall cavity effectively as well as the additional cost of material (and your time assembling each cassette). Depending on wind-loading, you may be able to have a 1.2m wide wall cassette with no vertical internal members (just a couple of horizontal ones), rather than the standard 0.6m wide Skylark wall blocks. But otherwise I would just go with the Skylark250 wall blocks and remove a couple of the horizontal ‘noggins’ if you’re very concerned about aspects of thermal bridging.
I also noticed your wall sections are bit deeper (300mm cavity). I wonder if you have done any thermal/PHPP calculations that are driving the additional structural depth? There’s always the option to have a thinner structure and use 40-60mm wood-fibre insulation boards externally mounted to the chassis which would alleviate any thermal bridging concerns.
Your spigot/socket connection here is actually something we tested as part of the predecessor to Skylark, the Blackbird system. Together with Pulp Build, we completed one project with this approach, which is a grid frame of posts and beams. Structurally however with Skylark a socket like this isn’t necessary, as the bowties do a really good job in terms of lateral bracing/racking. In terms of wind push-pull forces, the little notches we have at the top and bottom of each wall block are enough to grip on to the floor/roof, and greater concern is the front panel of the cassette pulling off (which is why we use the 9mm thick tabs down the sides, together with brad nails during assembly of the blocks).
I get your point that accessibility of bowties is a bit tight, but doable with the plywood wedge or a T-shape and a mallet. We have experimented with a similar approach to you where the bowties are raised up, but our experience has been that the current positioning saves a lot of extra material (and wastage) that moving them higher would cause.
The gap you mentioned is great as a service cavity but does make a trickier insulation detail on ground floor and roof. There are products like XPS or wood-fibre or even soft-fill that can be used to plug this on site without too much difficultly
Large opening spans.
Your thinking about a beam/lintel is exactly right, although we actually believe openings up to 2.4m wide are achievable with a plywood lintel rather glulam. We will be releasing a design for this with the SKYLARK250 v0.2 files, but essentially it’s 3x thickness of the plywood boards with approx. 450mm depth, that sits in the floor depth and notches over the adjacent wall blocks either side of the opening. However if you want to go much beyond 2.4m wide then absolutely, you can notch a glulam/solid timber lintel into the same space).
The exact size and location of these openings will need to checked over by the engineer for your particular design and site. As with any building system, the placement of structural openings has a potential impact for the stability under wind and gravity loads.
Using firrings like this is a good idea. Like you suggest it offers an extra layer in case of waterproofing issues with the roof. I think on shallower gradients this is probably more hassle than it’s worth (which is why we chose to integrate it on the 200 series), but if you want to use the same floor beams for the roof, then yours is a good approach.
You might be aware of this already, but you can also go longer on the floor span if space on your site allows. The Skylark250 series floor blocks go up to a 5.4m internal span (if you’re looking for more m2).
I hope this offers some helpful suggestions as a way to make your whole structure simpler to build, and use less material overall to achieve the same design. I really admire your in-depth investigation into Skylark joinery and for raising these queries and improvements. You’ve done some valuable thinking about thermal bridges, and I like your suggestion for creating window openings, rather than individual window blocks. I do wonder if sticking more closely to the structurally documented Skylark wall/floor block connections may help to engage with structural engineers (and potentially reduce consultancy cost).
I know Mel posted a list of engineering practices familiar with WikiHouse in the UK. If you’re still having trouble reaching out to these, please do shout and we will see if we can facilitate a meeting for you.