I'm putting a steel frame in, to "open plan" the rear of a brick cavity-wall terrace house in the UK, which will then have a single storey extension added behind it.
I have already retained a structural engineer for earlier work on the project, and he seems quite sharp and pretty good. But I've also had considerable conflicting advice and have concerns on 2 key issues, so I want to ask for other people's suggested approaches in this situation, to inform me and let me think a bit before the fees clock starts ticking:)
The house is a pretty ordinary 1920 - 1940s brick London terrace house. It's got a 100mm cavity rear wall, pointing is good (or will be by the time we're done), and the foundations are quite wide (500mm front/800mm rear) and the subsoil is a mix of clay and large-ish pebbles, probably good for any likely increase in load despite the location (100kN+ bearing?). The oversite slab (about 350mm below finished floor level and 150mm above foundations) is solid with no cracks, and very dense - I killed a concrete saw blade just sampling through it, and had to use a breaker on the subsoil just to get 400mm down. The adjoining houses have a 200mm+ solid brick party wall on one side, and single leaf only (100mm) on the other. The adjoining house masonry is very poor and weak indeed and missing in many places.
The plan is to open plan the entire rear ground floor, with a symmetrical "figure 8" moment resisting 203x203x46 frame resting on the foundations, the bottom member then encased in C40 self-consolidating concrete for load spreading and protection. The "figure 8" would be about 3m high, and the upstairs floor is being replaced with dense block&beam running front-to-back (this is a settled point: despite the extra load and that the house wasn't designed for it, it's something we want to do for various reasons that don't matter here). The block and beam upper floor is being supported by load-spreading C40 on the oversite slab (calculated already) and so the rear concrete beams resting on the frame will only be ~ 2 - 2.5m long. The extension roof when complete will slope, and meet the external wall quite a bit above the top of the frame, so the top of the frame can be exposed outside the current external wall face - it won't affect water integrity or corrosion issues when completed. The engineer has indicated his default choice is to propose calcs based on the concrete beams sitting in the flange of the upper UC with stiffeners, and the splice in the upper UC joined by moment resisting join to the central UC.
There are 2 questions I keep having, that just keep coming to mind, and I want to see what suggestions exist before incurring fees or expressing any ideas to the engineer.
- 6m of 203 UC is heavy - just under 300kg. I'd like to get it as 2 parts with end plates and splice them, just to make it practical/safer to manage and place, but I'm not sure what sort of end plate/splice options I have. It'll be encased, so there will be uniform support along the underside of the beam, as well as the underside of the top flange too, as well as along the web, so torsion won't be an issue. But the load is at 3 points - ends and middle - and the middle is where any splice is likely to be.
What kind of splicing approach should I expect, and how easy will it be to make it rigid enough, if spliced close to the middle load?
- The torsion and eccentric load on the upper beam worry me. It just feels like weak designing, even if the calcss did stand up, because it doesn't nullify those issues like other similar-cost solutions probably could. The UC is 203mm wide, the upper masonry it supports is 300mm wide (2x100 brick + 100 cavity), so it'll need some kind of 320mm wide plate or channel welded/bolted onto it to provide full-width support to the upper walls. Yes both upper leaves would indeed be 100% supported, but they'd also be COG considerably further from the UC web, creating extra torsion.
Given those, I'm tempted to ask if he can design the upper frame from say 300wx200H RHS instead of 203x203 UC - same height, 2 webs one each edge, broader hence no additional top plate, much less of a torsion concern, and enough extra stiffness that I can cut 125Wx150H holes in one side every 520mm c/c, and insert the floor beams into the interior of the RHS. The concrete beams can then rest centrally spanning the entire inner bottom width of the RHS rather than resting eccentrically on a UC flange. It also avoids stiffeners that a UC would need, compensating for the extra material cost and there's a wider range of thicknesses than UC would have, for more or less rigidity, to avoid wasted over-speccing.
Does this make sense to propose, or are there issues I'm not seeing?