Partition Load In Multi-Story Residential Wood Structures 1

[abusementpark]

I see where most engineers only require wood floor trusses in these type of structures to be designed for 40 psf LL without any explicit consideration for the non-load bearing partition load within residential units. I am aware of the TPI 1 exception that allows non load-bearing partition load to be neglected for the wood floor truss design, which I assume is the basis for this.

Does this mean most engineers are neglecting it for the overall building design (e.g. load-bearing wall studs, foundation, etc.)? I could definitely see the logic in doing that if we are saying the floor trusses don't need to be designed for it, then why would elements further down the load path with greater tributary area need it included? Also, I assume you'd still be accounting for the self-weight of load-bearing walls, party walls, corridor walls, etc. I am just talking about neglecting the partition walls within residential units.

 

[msquared48]

Where the partition walls are transverse to the floor joists, I would not worry about it.

Where the walls are parallel to the joists, dpoble up the joists under the wall where the wall is 50% or more of the joist span.

For seismic though you could check the wall weights of the story and add the weight to the story weight in your calls - probably effectively 5 psf or less though.

Mike McCann, PE, SE (WA, HI)

 

[XR250]

Its nice to have something under a partition if your joists are 24" O.c.
Over time, the subfloor likes to sag under the sustained load.

 

[KootK]

I've struggled with this myself on and off over the years.

1) Can you point me to this TP1 exception? I wasn't aware of that.

2) Among engineers that predominantly do residential, I do see a lot of folks ignoring partition loads entirely.

3) Among engineers that predominantly do ICI work but dabble in large scale residential, I do see partition loads specified, mostly in the 10-20 psf range. Often, there's already floor toppings in play which tend to dull the effect of the extra load.

4) I feel that 20 PSF is too heavy with respect to a) reality and b) my ability to remain competitive. So here's what I've been doing lately:

a) apply 6 PSF partition weight throughout the units, citing the HUD seismic recommendation below.

b) adding the calculated weight of partition walls because, obviously, that tends to be denser partition weight relative to that within the unit.

c) adding the calculated weight of exterior and corridor walls.

I feel that this is a pretty good blend of:

d) responsible

e) competitive

f) computationally easy.

I'll not lie, (f) factors in pretty heavily.

 

[abusementpark]

KootK - see the excerpt below in TP1.

 

[dik]

KootK:

Part 4 of the Canadian code requires that for non-fixed partitions that a 20 psf partition load be used. My drawing notes stipulate that partiton locations are fixed and that to change them, an engineer has to be involved. I then include a partition load of 3 to 5 psf... and have written a little SMath program for Partition Loading...

Dik

 

[gte447f]

Personally, I would err on the side of conservatism and assume that the TPI exception would not apply because the interior walls would exceed 60 plf. I like 10 psf for framed walls with 2 layers of gyp (ASCE 7-10 Chapter 3 commentary says 8 psf). Either 8 psf or 10 psf exceeds 60 plf for any wall 8ft high or higher.

For building design (i.e. load bearing walls and such), I am comfortable with including a uniformly distributed load allowance (e.g. like 10 psf) with the floor dead load. To get an idea of what allowance to include, you can multiply weight of wall in psf times height times total length of all interior walls and then divide by floor area. I did this on a few projects many years ago, and if memory serves, 10 psf additional floor load was reasonable.

For delegated component design like floor trusses, I would "specify in the structural design documents the non-bearing partition loads that need to be applied to the trusses", in order to conform with TPI 1. Ideally, these would be line loads (plf), and the truss designer would locate and distribute them as needed to design the trusses, but that probably rarely occurs in real life. More likely, they would prefer to apply an additional uniformly distributed load as I describe above for building design.

 

[Ron247]

One reason I would err on the side of being conservative is the defining of a "non-load bearing wall". If the 3rd floor truss can span 25' without an interior support but you place a 2nd story wall directly under the middle of the 3rd floor truss for example, it is a load bearing wall to the 2nd floor trusses. It is not an essential load bearing wall, but it is a load bearing wall.

Ignoring the load for truss design probably is not that severe as it would be for ignoring it overall. The walls are present and the weight can add up to a considerable amount. Therefore, it appears the code reference appears to only apply to the truss design.

6.2.2.1.2 seems to address some of the earlier responses to the post.

 

[phamENG]

I'm in gte447f's camp and typically use the same procedure. On the drawings (at least for large single family and small to medium multifamily), I'll key trusses at parallel walls and tell the truss designer to align trusses with walls parallel to truss span. In my general notes, I direct them to either provide two standard trusses under the wall OR design those trusses with an additional dead load of XX plf. I then require the proof of the latter in their submitted calculations.

 

[abusementpark]

Remember that I am mainly talking about the partition walls within units, not party walls or sound walls, or corridor walls, which typically have multiple layers of sheathing on each side. The ones within units are usually two layers of gyp board total and a minimal amount of stud framing, about 6 psf, and as long as the ceiling height doesn’t exceed 10’, which it usually doesn’t in this type of construction, then the wall self-weight would be less than 60 plf.