Concrete Fire Wall Design

[Said the Sky]

Hello all,

I have been requested by the architect to provide a detail for a concrete fire wall that spans about 37' ft above slab on grade (43' to top of footing)

This will be a non-load bearing wall in terms of gravity loading and will not be connected to adjacent structure on either side (free standing), so I will be designing it for cantilever loading of 0.5kpa factored lateral load or seismic loading.

According to CSA A23.3 walls exposed to fire on both sides need:
1. 200mm thickness
2. aspect ratio width/thickness of not less than 4.0. (I assume width is the length of the wall?)
3. Min 50mm cover for reinforcements
4. Reinforcing is required to be placed in vertical and horizontal in accordance with clause 10 and 14 of A23.3. Do I need to conform to Lu/30 for non load bearing walls to determine minimum thickness within clause 10&14 or do I just follow the requirements for reinforcing? if not I'll have to use a 475mm wide wall?
5.effective length of wall k*Lu is not more than 3.7m, in our case k = 2.0 for cantilever, Lu = 14m, way exceeds 3.7m.

i've attached an elevation of the fire wall in question, I have about 3 around this building, its a 3 storey residential townhome.

I've also attached the code im referring to in A23.3

looking for some advice, should I consider a different type of fire wall design, tied wall, weak-link wall, or stick with cantilever wall (more reinforcing bigger footing etc)

thanks!

 

[Said the Sky]

1.4.5.1 Masonry and concrete wall provisions

 

[Said the Sky]

1.4.5.1 Masonry and concrete wall provisions

 

[BAretired]

The firewall should be laterally supported on both sides such that it won't collapse if there is a fire on the opposite side.

A wall exposed to fire on both sides is not a firewall.

 

[Said the Sky]

Hi BA,

my understanding is that if you laterally support your firewall on both sides of the structure, and if one side of the structure collapses due to the fire then the remaining undamaged side of the building is providing lateral support to the firewall, and also some lateral load due to the collapsed building. This is a condition for a "tied firewall", which I am open to doing, if there is a reliable way to transfer that lateral load imposed on the undamaged side by the collapsed side, however I am more keen on bolts that melt away from the damaged side but remains intact on the undamaged side thus still providing the lateral stability of the wall

for a cantilever fire wall, the wall is not connected into either side of the building and I believe typically has a 1" air gap between the concrete wall and the wood walls on either side. So when one side falls down the concrete wall is exposed to lateral loads from explosion or heat, and wind which I believe is accounted for in the 0.5kpa requirement.

I guess I am looking for advice if anyone done a firewall this high before? I don't think I need anything more than an 8" wide wall in terms of a cantilever design, however if I am to follow clause 10/14 of A23.3 then my minimum width of wall will need to be alot more than 8" (min width Lu/30 for non load bearing wall), more like 19" wide which I believe is overkill.

 

[BAretired]

The best firewall is a double wall, with each wall tied to the structure on just one side. The problem is high cost, so that solution is rarely if ever used.

I believe that the next best solution is a single wall tied to the structure on both sides such that the wall is prevented from collapsing in case of a fire on either side. Melting bolts, if used, would have to be installed on both sides of the wall. There is no guarantee that the temperature on the fire side will be sufficient to melt the bolts, so it would be better if the ties on the non-fire side are adequate to resist tension from collapsing structure on the fire side.

A 43' high cantilever seems offhand, to be totally impractical to me, but correct me if I'm wrong.

 

[Said the Sky]

if the cantilever wall is only required to be designed for the 0.5kpa (factored load) as per my attachment from the NBCC commentaries. Then I believe a 8" wall cantilevered is still sufficient (although it might not stay plumb), as my seismic load does not exceed the 0.5kpa (low seismic area). By my quick numbers I only need 15M@10"O/C verts, footing will need to be big enough for overturning, but I think would still be cheaper than a double wall. If the cantilever walled is required to be designed for regular exterior wind forces then it won't work. If I am reading the clause correctly from my attached image, then the 0.5kpa lateral load includes wind pressure.

The double wall idea also requires a unique connection into the floor system which, where I am from I don't think I can trust them to install it properly. I will probably have to be a ledger with post installed bolts, thru-bolted to the rimboard at each floor to drag the load into the diaphragm.