Cantilevered Wood Floor Joists

[TRAK.Structural]

See sketch below. I have a situation where I'm trying to extend the interior space of a residential structure by cantilevering the ground level first floor joists (apprx 2' cantilever and 11' backspan). There are similar "bumpouts" to this on the home already, so the goal is to just make the exterior wall all in the same plane along this elevation and recapture some interior space.

Firstly, do these joists need to be pressure treated? I can't find anything specific in the code for this. Seems similar to a roof overhang where the rafters would typically NOT be treated; but it is obviously much closer to the ground which could be likened to joists in the crawl space that need a minimum distance from ground to NOT be pressure treated.

Secondly, I think an inverted hanger between the band and the joists makes sense to transfer the wall loads from above. Would you try to size the hanger to take the entirety of the wall loads, or some portion of them based on the width of the sill plate that is over the band, or something else?

 

[XR250]

I have never used PT for this. I don't think the hanger is needed unless the wall loads are hugmongous (which is unlikely if your cant. actually works) . Might want to block every other or third joist bay at the fdn. wall though.

 

[ResidentialStructGuy]

PT is probably not required for this case. The way I usually detail this would be to extend the cantilever floor joist to the edge of the wall and provide rim blocking between the joists. This allows the sole plate to bear on the joist instead of using a hanger. I would block every bay at the foundation stem wall (might be overkill depending on your lateral loads).

 

[KootK]

I wouldn't normally PT this but, then, I'd want the cantilever soffit to be inside of the building envelope.

I quite often do use inverted hangers for this and assume that all of the load travels through the rim member. My main concern with this is the inevitable built up posts that wind up existing within the walls for patio doors etc and landing on non-continuous single ply sole plates with joints between joists. I'm not overly concerned with truly uniform wall load in most cases. My concern for concentrated loads is also alleviated if I'm able to get double joists under all of the significant point loads etc. That can make a mess of a nice, clean joist run in a hurry though.

 

[KootK]

he way I usually detail this would be to extend the cantilever floor joist to the edge of the wall and provide rim blocking between the joists. This allows the sole plate to bear on the joist instead of using a hanger.

I don't love that.

Firstly, I don't think that you need to run the joist right to the outside of the wall thickness in order to be able to claim that the sole plate is substantially supported by the joists. Not unless your rim piece is very wide at least.

Secondly, in switching to blocking instead of a continuous rim member, you're losing the most important benefit of the rim: load distribution between joists.

In my mind, your approach gains little that is worthwhile and sacrifices much.

 

[XR250]

As KootK mentioned, I will use the upside down hangers at point loads. I am doing one now where I have a ridge load coming down. I used a cont. 9 1/4 LVL for the band and upside down hangers - except I put a dbl. joist under the ridge and used a right side up hanger to get the load into the rim and distribute it to the neighboring joists.

 

[TRAK.Structural]

My main concern with this is the inevitable built up posts that wind up existing within the walls for patio doors etc and landing on non-continuous single ply sole plates with joints between joists.

What about a continuous rim and blocking between the joists just inside the rim? Seems like that could alleviate some concern about point loads on sill plates that may or may not be continuous?

I don't love that.

Firstly, I don't think that you need to run the joist right to the outside of the wall thickness in order to be able to claim that the sole plate is substantially supported by the joists. Not unless your rim piece is very wide at least.

For a 2x4 stud wall above, how wide of a rim is too wide to consider some direct loading of the joists? 2x rim or 1 3/4" LVL is right around half the width of the sill plate.

As KootK mentioned, I will use the upside down hangers at point loads. I am doing one now where I have a ridge load coming down. I used a cont. 9 1/4 LVL for the band and upside down hangers - except I put a dbl. joist under the ridge and used a right side up hanger to get the load into the rim and distribute it to the neighboring joists.

That's a pretty nifty approach I hadn't thought of, thanks!

 

[TRAK.Structural]

I have never used PT for this. I don't think the hanger is needed unless the wall loads are hugmongous (which is unlikely if your cant. actually works) . Might want to block every other or third joist bay at the fdn. wall though.

Existing joists are 2x12 @ 24 (at the other bumpout sections I am trying to mimic). I can't get that size and spacing to work, which is always fun to explain to the owner and contractor.

Wall loads are not insignificant. The wall supports a 2nd level, walk-up attic, and roof. All center bearing so about 7' of trib for 2nd floor and attic, and 14' of trib for the roof.

 

[ChorasDen]

On your drawing you indicate less than 2ft. From a code standpoint, I am going to buck everyone else's opinion on this one and say that these should be pressure treated. In unconditioned crawl spaces, IBC requires at least 18" for untreated wood, this is exterior with likely less than 18" airgap to grade, and thus, is going to be more subject to moisture issue than in an unconditioned crawl.

 

[TRAK.Structural]

On your drawing you indicate less than 2ft. From a code standpoint, I am going to buck everyone else's opinion on this one and say that these should be pressure treated. In unconditioned crawl spaces, IBC requires at least 18" for untreated wood, this is exterior with likely less than 18" airgap to grade, and thus, is going to be more subject to moisture issue than in an unconditioned crawl.

Yes, this is my concern. It's probably going to come down to what the code official says.

 

[SWComposites]

Existing joists are 2x12 @ 24 (at the other bumpout sections I am trying to mimic). I can't get that size and spacing to work

then sister two 2x12's together @ 24" spacing.

 

[SE2607]

On your drawing you indicate less than 2ft. From a code standpoint, I am going to buck everyone else's opinion on this one and say that these should be pressure treated. In unconditioned crawl spaces, IBC requires at least 18" for untreated wood, this is exterior with likely less than 18" airgap to grade, and thus, is going to be more subject to moisture issue than in an unconditioned crawl.

You need only 8" of separation between earth and wood to avoid PT. The 18" interior gap is to provide a crawl space (for someone other than myself).

 

[ResidentialStructGuy]

Secondly, in switching to blocking instead of a continuous rim member, you're losing the most important benefit of the rim: load distribution between joists.

In my mind, your approach gains little that is worthwhile and sacrifices much.

How many joists do you distribute the load from the walls to? I am not familiar with detailing a continuous rim joist to distribute the loads to the cantilever joist. TIA for the insight.

 

[TRAK.Structural]

You need only 8" of separation between earth and wood to avoid PT. The 18" interior gap is to provide a crawl space (for someone other than myself).

Excerpt from 2015 IRC (this is the adopted version in my area).

 

[XR250]

Existing joists are 2x12 @ 24 (at the other bumpout sections I am trying to mimic). I can't get that size and spacing to work, which is always fun to explain to the owner and contractor.

Wall loads are not insignificant. The wall supports a 2nd level, walk-up attic, and roof. All center bearing so about 7' of trib for 2nd floor and attic, and 14' of trib for the roof.

No way you are getting that to work in real life.. Even if it calcs out, you will eventually have a sag at the cant. and a hump in the middle of the floor. Ask me how I know

 

[TRAK.Structural]

No way you are getting that to work in real life.. Even if it calcs out, you will eventually have a sag at the cant. and a hump in the middle of the floor. Ask me how I know

Somehow it is working already adjacent to the area I am trying to extend, see pic.

I'm leaning towards using LVL joists at a tighter spacing than the 24" oc for the other areas.

 

[SE2607]

View attachment 7356

Excerpt from 2015 IRC (this is the adopted version in my area).

The proximity of the soil to wood on the outside of the footing is always going to govern.

 

[TRAK.Structural]

The proximity of the soil to wood on the outside of the footing is always going to govern.

Right. This code section however only explicitly addresses wood INSIDE the periphery of the building foundation, which this is not.

This is also not a deck/porch/other unconditioned area. So to me it is somewhat of a gray area.

 

[gte447f]

By the "letter of the law" it doesn't have to be pressure treated, because it's not inside a crawl space of the periphery of the foundation. If you want to use pressure treated for best practice, I don't think anybody could argue to strongly against you. If the owner and/or contractor want to go ignore your advice, let it be on them.

Regarding the joist size and spacing of the original versus what your new analysis, just design whatever you have to to meet the building code requirements. If it differs form the original construction, so be it. If anyone questions you about it, simply explain that the existing doesn't meet current building code requirements, so the new construction has to be different.

 

[XR250]

Somehow it is working already adjacent to the area I am trying to extend, see pic.

I'm leaning towards using LVL joists at a tighter spacing than the 24" oc for the other areas.

View attachment 7377

Are there existing dropped or flush beams across the bump outs that may be taking some load off?