MortarThePoint

2-ply trusses are a difficulty as they have a thin gap between the plys. I dribbled so PVA glue into such gaps as well as smearing sealant along the grove at the top and bottom of the truss bottom chord. It won't be perfect, but should help. Perhaps the ideal would be to apply sealant all the way along top and bottom, but I didn't.

Some pictures below. I don't seem to have one that shows how the horizontal OSB laps onto the wall plate by about 50mm and is screwed down onto the VLC that is double sided tape taped to the top of the wall plate. That leaves a stable joint that can then have sealant applied. It's a slow process, but one a became reasonably proficient at and should be good and airtight. Having sealed the OSB box itself, if there is a leak through the VCL joints then there is a second line of defence. The VCL will be continued along the rafters later. There is enough VCL for it to extend about 1ft beyond the top of the vertical OSB I think.

Yes, the half landing frame is very solid and has lateral timbers bracing against the wall opposite I was mainly thinking about the splitting risk you identified, but that wouldn't be due to loads

This is why I was thinking it would be a good idea to glue (CT1) the nail plate to the stringer rather than using any nails. The glue would push through the holes in the nail plate making for a very string bond That's a nice arrangement as it avoids any screws into end grain. Given I already have the screws into the end grain, I am less inclined to drill / screw any more holes into the stringer. The rotation caused by any loading on the stair (or the stairs own wait) is to push the stringer into the angle that sits inside the bottom flange of the UB so it feels I figured end grain attachment would be OK. Our American cousins do this a lot and one of their common approaches is to place a piece of Plywood onto the face of the the trimmer joist and then screw through that into the end grain of the stringer. I have four 45mm wide stringers so hopefully there is a fair degree of redundancy at play.

@Gus Potter What do you make of the logic here? Do you think it would be beneficial to have the pillar less rigid, so not use the I-Studs? (Note: image says 146mm M.F. but its going to be 90mm)

Sorry, I should have quoted more of your original post. The nail plate would be to mitigate the chance of the stringer splitting at the screws at the top. I think that the screws at the top don't really perform much of a load bearing job as the stair would still stand without them since the steel angle is rigidly bolted to the steel beam so couldn't rotate out of the way you allow the stringer to fall. Here is a mock up that shows where I was thinking I could glue the nail plate. The three screws per stringer are shown as arrows. The nail plate would just be in an attempt to stop the timber from splitting. Outermost stringer removed for visibility.

The image below shows an area of metal frame under a steel beam that includes two wide doorways. The blockwork opening is 3.6m wide. The door openings are something like 1.5m each meaning the Pillar is just 600mm or so wide. I'm planning to use 90mm metal frame and can probably only glue (CT1) the track at the bottom to the concrete floor due to pipes and membrane. I had thought to make the Pillar as strong as possible so ordered some I-studs as well as plenty of C-studs. Thinking about it again, I'm now thinking that making the pillar really rigid may be a bad idea as that would increase the force transferred to the small area of track at the bottom when the pillar gets knocked or a door slams. What do you think?

I'm avoiding plywood and assume OSB wouldn't be a good alternative. Could I glue a flat nail plate in the area of the screws instead? I'm thinking of the Simpson style that is just like a drilled sheet of 1.5mm steel. Perhaps NP15/140/180. I'd use CT1 which would squeeze through the nail plate holes and leave it well stuck. I'd want to apply anything only to the inboard surface of the outside string as my dimensions there are critical

I used OSB and battens to make horizontal and vertical surfaces: 1. Stick (50mm double sided tape) VCL to top surface of timber wall plate draping down wall 2. Add battens to follow with OSB 3. Horizontal OSB from top of wall plate 4. Vertical OSB up from free end of horizontal OSB up between joists Battens along OSB joint. 5. Apply double sided tape to OSB edges 6. Pull VCL onto surface of horizontal OSB 7. Cut VCL to allow sections to pass up vertical OSB 8. Wrap where VCL touches joist with polythene tape 9. Apply sealant It required carefully measuring the joist gaps and thicknesses and pre cutting the VCL before attaching to wall plate. I'll try to dig out some photos I applied sealant the the OSB box before attaching the VCL to it but that's me

Can't get it in one length and prefer the look of it stopping at each post

What's the normal method of attachment between the intermediate post and the two sections of handrail? I'm familiar with the approach at a post that only takes a single handrail section where you can drill through the post and screw into the handrail end, but that wouldn't work with two inline sections of handrail.

The plan is to use two 90x25 and two 40x25 timbers glued to make a 90x90 hollow square section. It's going to be painted and the narrower width timbers will be inline with the handrail and spindles so somewhat hidden. I'm thinking the 90x25 facing the landing will get screwed to the Unistrut at 200mm intervals to couple the timber to the metal. Alternatively, perhaps the 40x25 pieces as they don't stop at the top of the beam.

This metal post will be down the middle of a timber post that will add the vast majority of the strength of a solid timber post too.

Here are some photos of the Unistrut in situ. It feels stiff. It extends 965mm above the top of the beam and to flush with the bottom of the beam. If a moment is applied to it equivalent to someone leaning on the handrail, the strut is pushed against the beam at the bottom and the top mounting is under tension. I used a top mounting that allows two M10 bolts (could upgrade to M12) to resist the shear and one M12 bolt in tension. All bolts are to be torqued, Loctited and doubled up. An exception to that may be the bottom mount's M10 shear bolt which doesn't really have room for a second bolt. The bolts are grade 8.8 so the M10 bolts have a shear strength of 22.3kN each and the M12 has a tensile strength of 48.6kN [https://eurocodeapplied.com/design/en1993/bolt-design-properties]. That means in terms of the bolts, the top mounting has a strength of 44.6kN. The strut acts as a lever to gears a force applied to its very top by a factor of (965 + 206) / 206 = 5.7. That means the bolt related strength would be 7.8kN before any safety factors are applied. @Gus Potter I know this isn't as beefy as your suggestion, but it feels solid. I could slip a solid rectangle section (e.g. 30x35) into the groove of the channel if I decide I want more strength. That could add up to another 79,000 mm^4 and achieve a total of 130,000 mm^4. However, I think the torsional strength of the 6.1m UC203x203x52kg may be similar to the bending strength of the 41x41x2.5 Unistrut. I could add a brace to the middle of the beam that couples it to the concrete HCF plank that is next to it which also spans 6.1m, but doesn't touch the steel beam. I think I'd rather not couple them though. I'm having two such posts which divide the balustrade up into three 1718mm sections. Discussions on another thread suggest 0.36kN/m UDL on the handrail which would make for a tip load of 0.62kN. That's well under the bolt strengths (~8%) and would create a Unistrut bending deflection of 17mm (L/57). It feels stiffer than that when I have applied a pretty large force. Any recommendation as to the required torques for the nuts?

I was assuming a UDL of 0.36kN/m so total load of 3m * 0.36kN/m = 1.08kN

Interesting, so adding that to the Stairbox 3000mm figure suggests each newel has to be able to take 1.08kN.

Other than height (900mm I think) I can't find any particular guidance on balustrades. It feels like there should be some maximum distance that a balustrade can be before it needs an intermediate newel post, but I can't find a regulation for that. Stairbox suggests 3000mm. Has anyone found a regulation that applies or has any guidance beyond what Stairbox say?

Looks like it's 6kN for a single M10 so if I used a bracket that allowed two M10 it would give over a ton of pull out strength What do you think @Gus Potter, could Unistrut be an option as it is super easy to work with and to source (e.g. TLC)? A lot less strong than 50x50x6 SHS (only 15% of the stiffness) but do you think it is 'enough'?

Unistrut steel channel looks like a possible half way house. It's 41 x 41 x 2.5 which isn't as stocky. It feels fair to model it as equivalent to a T section in terms of bending in a direction parallel to the red arrow (conservative since the parts circled in blue would add stiffness). The calculator gives a value for the second moment of area of 51,940 mm^4 which is 41% of the 40x40x5 SHS or 15% of the 50x50x6 SHS. That sounds like a massive downgrade, but the 1kN force at 600mm would give a deflection of about 6mm or L/91 which doesn't feel too bad for what is a pretty large force (100kg lateral force). If the mountings are 150mm apart, the force on the top mounting would be 5 times the load at the tip, so 5kN. I don't know how much force would pull out a T-piece from the channel.

Cutting a round hole in the timber post is easier than a square one. A 32mm piece of circular hollow section (CHS) with 4mm wall would have a second moment of area of 38892 mm^4 which is a quarter of the 40x40x4 SHS but would allow the use of a 32mm auger bit which is cheap. 1kN would bend it by about 10mm at 600mm cantilever. A 50mm CHS would be getting on for the 50x50 SHS but trying to drill a 50mm hole in a 90x90 post could be tricky to say the least. I may be able to outsource that though. I could mount a round tube against the flat surface of an angle bracket, but it would be nice to have something that mated to the curve.

Actually, these look beefy: 90 DEGREE RIGHT ANGLE 1X1 HOLE - TYPE (P1026) - HOT DIP GALV. 53X41X5MM https://directchannel.uk.com/90-degree-right-angle-1x1-hole 14mm hole is a bit bigger than ideal but that could allow adjustment with M12 bolts

Instead of welding plate onto the 50 x 50 SHS could I use any form of angle bracket bolted to the SHS and to the beam flange. I would want it to be substantial and may not find anything beefy looking enough enough. I guess an alternative is to make a 'bracket' out of a short length of SHS laid on its side along the surface of the beam flange. Two vertical holes to bolt that to the beam flange and one horizontal hole to bolt it to the vertical SHS which is what I'm trying to mount. I could have one such short length of SHS on each flange. I wondered about some arrangement with U bolts, but that wouldn't work being sleeved by the timber.

Cross post, but same conclusion.

Using this calculator: https://calcresource.com/statics-cantilever-beam.html suggests a 106mm deflection on a 600mm long cantilevered 16mm steel rod (210 GPa) if exposed to a 1kN load at it's tip. Wow! The 40x40x4 SHS would deflect just 2.7mm by comparison. I am amazed by the deflection of the threaded rod given that a product like the ZipBolt exists

40 x 40 x 4 is the biggest SHS section I can get at 40mm. That is still probably stiffer than a 16mm solid cylinder (core of M20 rod) Using https://amesweb.info/section/second-moment-of-area-calculator.aspx 50 x 50 x 6 SHS: 347072 mm^4 40 x 40 x 4 SHS: 125952 mm^4 16mm round bar: 3217 mm^4 [M20 rod?] 25mm round bar: 19175 mm^4 [M33 rod?] perhaps grade 8.8 has a higher Young's Modulus and Tensile Strength than Mild Steel though?