We have some walls! For the ground floor at least and the first floor should be complete in a couple of weeks. There's a detail in the roof construction that is still unresolved ...

You may remember (or not!) that we're having a barrel-vaulted roof with large (600 mm) overhangs to help protect the straw bales. So at the verge there is a curved glulam beam at the top of the wall and outboard of that are some cantilevered sprockets to support the projecting roof (18 mm ply plus aluminium roof).

The unresolved detail is how to connect the sprockets to the curved glulam. Since the sprockets are only supported at one end there is a moment. Thoughts have included long, crossed screws at 45° or proprietary connectors through-bolted on either side of the glulam. Finding a solution that an engineer will sign off on is the question.

We can't be the first to have this situation! Any ideas?

Posted By: djhFinding a solution that an engineer will sign off on is the question.

What about asking the engineer who is going to sign it off

I think bot is right, we need to know a little more about the construction Dave. Wind loading under the deep soffits on a large roof resting on straw bales however securely it's fastened, sounds a bit insecure to me too.

Looks awfully flappy at the verge to me, why so wide an oversail? Rain will hit the wall anyway!

Does the fact that there is a continuous internal air barrier mean you don t need a vapor barrier between the plywood and the aluminium roof sheeting?

The verge glulam is tied to purlins that run at right angles to it. The inboard area is the main roof timber structure, filled with warmcel or similar. The galv strap is fastened to the purlins. The ply skin has some strength; it supports the ali skin and provides the racking strength of the roof.

There's no vapour barrier between the plywood and the roof sheeting; there's a 'metal'-type breather membrane. Apparently not labelled on this section.

Does anybody have any ideas for connecting the sprocket to the glulam?

Posted By: fostertomWith the steel straps (fixed along the tops of 'main roof structure' purlins?) providing top tensile resistance to drooping, the bottom only needs minimal location, where it thrusts against the glulam. Lateral restraint by the ply skin. Neat.

Upwards wind loading?

the only way I could think of doing it then would be a metal connector plate screw bolted onto the glulam, as long as the glulam is sized to take the load and the sprockets have sufficient depth it should be ok:

I think a big gust of wind would rip the ply clean off any sprocket and hinge over breaking the ply and buckling the roof sheets, depending how the sprockets are fixed taking them too or leaving then behind.

625mm is very funny centres! I would put one under each join in the bottom sheet of ply then Soave at 2' centres about 610mm

The top strap could be builders band, cheaper and would do the same job without lifting the ply or needing housing in. They seem way too long too,

"Epoxy is an interesting idea but the engineer doesn't trust site made glue joins. He ignores them for structural purposes."

O well just a thought. Don't really blame them as there many ways to mess up using epoxy especially on site.

Posted By: tonya big gust of wind would rip the ply clean off any sprocket and hinge over breaking the ply

What, even though the hinge line is curved? Don't think so - the 3D geometry makes the ply very strong and resilient, and as Dave says

Posted By: djhThe ply is not primarily reliant on the eaves sprockets

so why the fuss about making them into cantilevers, if just decorative?

So what's the curve Dave? It's done in 2 layers, presumably break-joint, lots of glue/screws, so should act monolithic. I wd most certainly design it that way.

First I've heard of eave sprockets - and sounds like they're sorted.

6.5m radius is enough to be v stiff, no danger of fold-back, and fine to carry the outboard load unaided, I'd say. The verge sprockets are purely decorative, hanging from the ply (do it - visually).

Actually not hanging from the ply but self supporting, with the steel straps in tension at top edge, and bottom edge compressively against the glulam. In fact, don't think you even need the steel straps - the ply will do the same tensile job.

djh

The detail doesn't actually indicate the curved glulam beam. You say it's near the wall. Assuming the purlins are spanning from this to the gable wall, you could then extend these to form the sprockets. Intermediate ones would be decorative, bonded to the u's of the ply. I would also increase the ply thickness to 24mm.

If the glulam is directly at the gable line, I would use the brackets suggested by pot de baille. If these are bespoke, then it'll work. I've had the exact detail designed by an engineer with around a 1m overhang at eaves and verge on a flat roof. The tang on the t-plate will have to be quit long to resist the uplift on the sprockets.

Thinking on this more and from what Tom has pointed out out about the overhangs strength through the ply's radius I am thinking epoxy is going to be fine. Not sure ply would be better than solid wood though. The advantage to using say some 30mm larch or other suitable species is it would be expanding and shrinking in the same plane as the glulam beam also could match the beam. The sprocket can be glued and screwed to the overhanging ply and if you're concerned about the Epoxies strength you could screw and plug at an angle on the lower corner where the sprocket meets the glulam. This would also keep the sprocket steady while the epoxy sets. The main issue with West epoxy that is mentioned in the link are that the surfaces are grease free and the wood is down to 10% moisture content but 12% would probably be fine. If you can achieve these two and work in temperatures above say 10c can't see there being any issues. West sytem is Lloyds approved for boat building so should be more than capable for your application.