Could there be a beam continuously underlying/supporting a "bearing wall" in timber construction

[Ricyteach]

SUMMARY: What I am imagining is- instead of a bearing wall just sitting on joists which then sit on another bearing wall below- perhaps there a style of construction where a wall sits on top of joists, and those joists then sit on top of a beam that supports the weight of that wall plus the joist loads, and transfers those loads over to the exterior walls and down to the foundation. Does such a thing exist? If so, is there a way to determine this without cracking open floors and walls?

DETAILS:

I'm looking at sizing some footings and the W beam for a typical 3-storey row house in Brooklyn, NYC. 30 ft deep x 20 wide floor plan.

The basement lumber beam is being replaced with a W beam. The existing beam is 3.875"x6.5" and has a 8 ft span between two freestanding masonry columns. It continues past that span with another shorter span and goes into a beam pocket in the rubble wall on one side, and on the other it has another shorter span and then rests on an existing CMU wall (parallel to that CMU wall) and then goes into the beam pocket in the rubble wall (at the location the CMU wall meets the rubble wall).

The goal is to remove the masonry columns, and replace most of the lumber beam with a 16 ft span W beam supported at each end. The portion of lumber beam on the masonry wall will remain (the masonry wall will be filled with grout and we're going to check if there's a foundation under it; will also check that existing CMU wall based on TMS 402 to make sure it's ok to support that partial leftover beam).

Reconnaissance claims there are 3 bearing walls stacked on top of this beam, with all 3 floors transferring their loads to this beam.

I'm trying to figure out if this passes the smell test. Could this 8 ft beam be supporting all that load...? If I assume 50 psf floor loads (tributary width for the beam is reported to be only 7.7 ft; not sure why- chasing that one down) and 20 psf for 10 ft tall two-sided plaster walls, and another 10 psf dead load for the attic, the beam load is 1900 lbs per foot.

Assuming the (very old) lumber beam has an E between 1,000 ksi to 2,000 ksi, in the middle of this 8 ft span (bearing in mind the beam continues past this span) it should be seeing a deflection of something like 1/2" to 1" under full DL+LL, and about 1/4" to 1/2" for dead load only. But it doesn't have anywhere near this amount of deflection. The deflection appears to be so little it's not distinguishable.

Because of this I am wondering if this existing basement beam really is carrying the loads from the upper floors. I am imagining it could be only carrying the 1st floor load + 1st floor wall. So I imagined a way this could be true by inventing the idea of a beam carrying the weight of a wall and floor at each storey, and transferring that load to the outer walls and down to the foundation. Could this be the case? For each floor to be sort of "self supporting" in this way?

 

[kipfoot]

When does an owner or builder say, "I want to spend more of my money today for the benefit of a future owner who I don't know." I think the answer is relevant to your situation.

I expect that you're overestimating the actual live load (by a lot) and dead load (by some) and underestimating the natural arching of the plaster walls (which we always ignore). I do not expect that you'll find secret beams.

 

[LittleInch]

Is the supporting wall a common wall or is there in fact a beam like this one underneath each wall?

Or maybe a beam and a timber column which lines up with the brick piers?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Ricyteach]

Nope, no indication of columns lining up with the brick piers.

It's a good point about it being unlikely to have been built this way strictly for economic reasons!

Also walls acting like deep beams seems to be a big part of the answer to my riddle since the plaster and lathe probably wants to behave like a big diaphragm. And as you said, the actual loads are much less than design.

So all of this seems to suggest this is either a common wall or bearing walls stacked on top of each other, and you wouldn't expect to see much deflection in the middle of the 4x6-1/2 beam on an 8 ft span.

Thanks for helping me understand what's going on here!

 

[SWComposites]

i've owned houses of that vintage, more or less. there is no way there are beams in the upper floors. and based on what I saw in those houses, its highly unlikely the framing in the house will show good using current codes.

you had better be sure the new beam is very stiff (not just designed for strength) or there is going to be lots of plaster cracking in the floors above; over design the beam for a very tight deflection criteria. and don't even think about trying to jack things up from the basement.

 

[Ricyteach]

Yup, using deflection criteria for this beam based on modern day load assumptions! 50 psf combined floor load, 20 psf for the plaster walls.

Does L/480 make sense for the 16 ft span in a home like this? Should it be ramped up much higher to L/720 or something? Building code says L/360 but I don't think that's stiff enough.

 

[SWComposites]

I'll let one of the CEs here respond with specific recommendations on criteria.
Us old time aircraft stress types have a saying: "when in doubt, make it stout".
And having owned several houses with plaster and lath, you really don't want those walls moving.

 

[Tomfh]

Walls in these situations end up acting as deep beams. They arch to supports. They often crack at relatively modest deflections, due to them being so stiff, and so brittle. I jacked up my floor 20mm midspan within an 8m long wall, and got plaster cracking. That’s L/400, failing to control cracking of finishes.

Don’t skimp on beam stiffness. You really don’t want it moving more than a few fractions of an inch.

 

[Eng16080]

I think that you're likely overestimating the actual loads getting to the beam. Also, as mentioned above, there is probably a lot of strength due to the walls above acting as deep beams. This isn't something that would normally be accounted for in a design, but I suspect there's significant strength especially with the span only being 8 ft.

To answer the original question, I doubt there are beams or some type of structural system above which transfers the loads to the outside walls. There would be seemingly little reason to construct it that way. But, if it's important to know for sure, there's no point speculating. You should instead do some selective demolition to expose the framing.

Going from an 8 to 16 feet span is a big change to the structure. For the same beam section, doubling the span amounts to something like a 16x decrease in stiffness. Using a really high deflection limit like L/480 or L/720 probably makes sense here. I would probably try to use a steel beam section with the greatest depth that will still be acceptable for head height.

 

[Mike Mike]

Agreed
Walls arch
Plaster cracking bad
Make beam stiff
Check footings for new concentrated loads

 

[Ricyteach]

Great suggestions thanks all.

Yes I immediately recognized the 16x increase in deflection! I was able to get the span down to 15 ft (by suggesting we support on steel columns about 6 in from each wall) for a 25% reduction. So that helps. Will anchor the tops of the steel columns to the existing walls. Haven't decided how to do that yet.

But the beam we need is not the beam that we can use unfortunately. Need a 260 in4 section so that's a W8x31. But its dimensions are too big and would interfere with lots of plumbing and electrical that can't be moved, and it would kill the project.

So getting creative. I think that I've decided on two W6x20s- about 3 ft apart on each side of the existing beam- with 5/8" reinforcement plates top and bottom. Sister the continuous 2x8 joists to stiffen and deliver the load from bearing walls to the W beams. This is based on L/480. I've explained to the contractor the desirability of going stiffer but project budget and other considerations won't allow it. Out of my hands.

Serviceability criteria is by others!!!!!!! will be all over the drawings and report.