Could you get this to work?
Could you get this to work?
(OP)
This is an interesting one. Yesterday I was asked to look at a house built in the 1970's and have attached the basic frame section. The frames are constructed of 4x6 lumber @ 42" oc and 2x6 T&G decking for floors and roof. The loft floor stops about 10' from the front wall of the structure and there are interior walls in the plane of the frame at 10'-12' spacing. The longitudinal beams shown at the edge of the loft are (3) ply 2x10's with offset butt joints. I could not find any locations with 2 plies butted at the same location so I assume they are 16' long with 5'-4" offset.






RE: Could you get this to work?
RE: Could you get this to work?
Dik
RE: Could you get this to work?
Cal91. Yes, the interior walls are providing lateral support.
dik. I agree, it looks like it should not work but it has been there for 50 years. It was a windy day and I did not perceive any wind induced motion while upstairs. I also did some rhythmic bouncing and the loft was stiffer than the main level (and stiffer then most floors in my own house for that matter.
To answer your second question the connection at the loft/rafter does not have any moment capacity. in fact, I would say that it has limited gravity capacity. See attached photo.
RE: Could you get this to work?
RE: Could you get this to work?
Assume that it is stable (given 50 year history).
RE: Could you get this to work?
RE: Could you get this to work?
It's analogous to a simple supported beam on two pinned columns is vertically stable, and laterally stabilized by diaphragm / lateral force resisting system.
RE: Could you get this to work?
I agree that the structure works, that was the intent of the comment. If the structure functions as intended, the design is acceptable. It is like it had a fifty year load test, even though the load path is not apparent it has one.
RE: Could you get this to work?
RE: Could you get this to work?
Mike McCann, PE, SE (WA)
RE: Could you get this to work?
RE: Could you get this to work?
The main rafters are bending at the loft connection, which ties each side together. (the photo you attached)
RE: Could you get this to work?
Mike McCann, PE, SE (WA)
RE: Could you get this to work?
RE: Could you get this to work?
And, the only stabilizing triangles I see are above the loft... nothing below that.
I still do not like it.
Mike McCann, PE, SE (WA)
RE: Could you get this to work?
RE: Could you get this to work?
I suspect that the frame derives much of its gravity load "stability" from the fact that the gravity loads appear to be distributed pretty symmetrically.
My spidey senses lead me to speculate that something is missing from this picture though. If stability is really as fragile as it appears, I would have thought this thing pretty tough build in the first place without resorting to shoring etc.
I want a picture from the outside. Is it red like Papa Smurf's hat?
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
Lateral stability - floor and cross walls.
A trussed roof spanning between two side walls isn’t stable either without crosswalls. Are we going to fail that scenario too?
RE: Could you get this to work?
There is a small flaw in my argument here. The mechanism below would require the middle post to shorten. I've shown that below, greatly exaggerated. So the frame would be stable mathematically if the collar ties were sufficient rigid flexurally. But, then, I can't fathom a structural engineer worthy of the title putting any serious stock in that.
I also contend that the situation is quite different from that of a common house truss on plumb walls. With plumb walls, unbalanced gravity load does not produce a significant tendency towards lateral movement (racking). That's a pretty fundamental difference. Just ask the steel erectors that work on the Frank Gehry stuff.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
Msquared48. You state that it is unstable but that is only if there were no shear walls. With walls parallel to the frame at 12’-14’ on center it will not deflect laterally as stated by cal91.
Msquared48. There is a large window wall at one end and this is the end that has the cutback about 14’ for the loft. The lateral mechanism for these last two bays would be the roof diaphragms back to the 1st interior wall and to a pair of 6’ wall sections at the eaves. Given the short eave height these piers should be sufficiently stiff.
OldBldgGuy. I think you are heading in the right direction but I do not believe that the rafters have sufficient capacity in bending to support gravity loads. I.e. if you eliminate the walls and used a three-point support (fix center, roller at each tail) I think that it is likely that the tails would fail.
Kootk. I completely agree that it is unstable save for interior shear walls. My involvement is to listen to floor squeaks. The couple that currently owns the structure just bought it a month ago and after moving in noticed that there are a lot of squeaks in the loft floor. They tell me that everyone who looked at the building with them commented on how unusual the framing was to they were starting to get nervous. Their first contact with me was a two days ago. The item that they called me out to look at (floor squeaks) was easily resolved. The 2x6 T&G deck is nailed to the beams and over time a few have experienced enough shrinkage that they have slight gaps. When someone steps on these planks, the gap closes and the nail squeaks against the board. Running deck screws through the board into the beam below should solve that problem.
My spidey senses were on edge when I first walked in as I said in post #4 of this thread but I didn’t run as my first inclination wanted. I started moving around (to be sure, cautiously at first) and using my body mass and inertia to load and deflection test the structure. ;>) I was very surprised to find it quite stiff. At that point and given that I knew that the building was about 50 years old, I had to admit to myself that the system worked even though I could not see how. I do agree that construction stability would be a major challenge because without the roof and floor plank diaphragms and the interior shear wall it is very unstable at all phases.
Tomfh. Lateral stability I agree but gravity stability by bending of the tails, I do not think so. More on that in the next post.
Kootk. I can assure you that there are no moment connections anywhere in this structure (other than the treads of the spiral staircase but that is not relative to this discussion.)
I have a problem with our diagram because it shows racking and the numerous and closely spaced interior walls prevent lateral movement. I doubt that they are plywood sheathed but even as double sided GWB walls they add up to a lot of lateral restraint. I do not disagree with the mechanics of the racked diagram or your conclusions, just that it is not apropos to the existing conditions. I would contend that from an analytical standpoint, the bare frame would be no more or less unstable if the stub walls were vertical. The structure is symmetrical about the ridge so the tipping force on the walls is opposite and EAQUAL. They offset. A fully pinned symmetric frame under gravity only will always have 0 net lateral force.
RE: Could you get this to work?
Yeah, I know. Here, I find that you need to carefully qualify your opinions in order to keep them from getting challenged needlessly on points of little consequence. A bit like lawyering I'm afraid. That's the only reason that I mentioned moment connections at all.
There is no discrepancy. My sketch was meant to address the case where in plane shear walls did not exist. I posted it in an attempt to dispel the notion that this thing would somehow be inherently stable without such help. As I've said a couple of times now, I'm in the shear wall camp, even if those shear walls would have been unintentional originally.
I take exception to this. What about unbalanced snow and live loads? What about wind and seismic acting concurrently those loads? Even if the gravity load is balanced somehow, it's still a problem in a stability sense. Remember, every stability problem is not about where the ideal structure starts off. Rather, it's about where it'll end up given imperfections etc. There're plenty of Euler columns that are loaded symmetrically on paper but are still altogether unstable.
This is only true of structures that are stable. That stability cannot be presupposed in this case as that's the issue at hand. You're right, you wouldn't develop net lateral reactions here. But that implies lateral movement as I suggested in my last post.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
RE: Could you get this to work?
This doesn't change anything in my opinion. Consider a structure that is:
1) Fully symmetric in terms of it's internal composition.
2) Fully symmetric in terms of it's external loading loading.
3) Fully pinned.
Such a structure will have a zero net lateral force but may still very much have a stability problem, as your structure here would have in the absence of the shear walls etc. This, because the imperfection that initiates instability neuters the symmetry, pretty much by definition.
With loading and geometry 100% symmetric, I would agree that the situation would be no worse off for having the sloped knee walls as compared to plumb framing.
Excepting stability issues, I actually believe this to be true for structure with asymmetrical loading.
Rest up your fingers and get back to work. We can debate further on the weekend if you're still interested.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
Well, that realization put everything into a different light! Once I came to grips with that I started looking to see what could be holding this thing up. I think that the basic gravity load path is 3 bearing points, center wall and the 3-ply 2x10 beams that are place above the loft floor joist/rafter joint. This beam runs the full length of the building down both sides and passes through all of the partition walls. My theory is that these beams are bearing at the partition walls down to the floor structure. When looking at it this way the rafter tails have a little bit of axial tension and relatively small amount of bending with support at the end provided by the inclined wall.
RE: Could you get this to work?
- There isn't a stability issue because the sloped walls aren't really load bearing.
- This nicely addresses the construction issue that tripped my spidey senses earlier.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
RE: Could you get this to work?
I disagree that the rafter tails are in tension. They are primarily in bending, but slightly in compression. The short sloping walls are also in compression.
Without the benefit of shear walls, this structure is unstable.
BA
RE: Could you get this to work?
RE: Could you get this to work?
You are assuming a hinge in the rafter at the junction to second floor. Actually, the rafter is continuous. If the sloping wall is removed, the rafter would be a cantilever which would deflect downward. The sloping wall prevents that, so the sloping wall is in compression. That compression can be resolved into components parallel and perpendicular to the rafter. One causes axial compression while the other causes bending in the rafter tail.
It is possible that the 3-2x10 beams carry some of the second floor load into the partitions as you have suggested but I don't know how much roof load they could carry. That would depend on their connection to the rafters which is not specified.
But whether or not the beams carry roof load into the partitions, the structure relies on the shear walls for lateral stability.
BA
RE: Could you get this to work?
RE: Could you get this to work?
I bloody bet. Just imagine how concerned you'd be about the security of the investment that you'd made as the owner. Frankly, when I look at your initial sketch with fresh eyes, I feel a bit foolish for failing to see this in the first place. Darn that hindsight.
I agree with BA on the rafter tails not being in tension. I left that alone as it would seem to be of little consequence at this point, other than as an aid to clarifying general understanding (which always has some merit).
While there's little doubt that the shear walls are the lateral resisting elements in the conventional sense (W/EQ/P-Delta), I think that it's still a pretty big win to have some assurance that unbalanced gravity loading is no longer prone to making substantial contributions to lateral instability. Although you may have expressed it a bit differently, I'm sure that was what was nagging at you originally as well.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
RE: Could you get this to work?
Yes it can fall over if there are no cross walls, the same as any frame/beam supported on parallel walls can fall over.
I don't think it's fair to criticise this system on the grounds that it's not stable without the shear walls when we don't make such criticisms of other systems.
Harder to erect, and more unstable during erection, no doubt...
RE: Could you get this to work?
I believe we do make a similar criticism of other buildings. A rectangular building which relies on diaphragm action of the roof and end walls for lateral support is unstable without end walls whereas a building with frames may not need end walls. This building has frames, so it needs to be stated that shear walls are required for stability.
BA
RE: Could you get this to work?
RE: Could you get this to work?
Speaking only for myself, I wasn't concerned for this system because it's not stable without shear walls. I was concerned for it because its nature is such that unbalanced gravity loads would induce lateral displacements. And that's not generally true of typical systems. In fact, it's a situation that requires a greater than typical amount of designer attention to lateral load paths and P-delta effects. Of course, this is moot given recent revelations about beam support.
And rightly so. Initially there was speculation that the frames themselves might provide the lateral stability. And, of course, now we're debating issues that are somewhat independent of your original question.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
BA
RE: Could you get this to work?
Any wall that goes from floor to floor or floor to roof can be used as a shear wall.
I made a point to call out the 3-ply beams so that they were not missed.
RE: Could you get this to work?
which have been introduced above.
BA
RE: Could you get this to work?
BA
RE: Could you get this to work?
I understood your OP. What I am questioning is the connection between the beam and structure. How are the longitudinal beams
fastened to the frames?
BA
RE: Could you get this to work?
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
BA
RE: Could you get this to work?
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.
RE: Could you get this to work?
RE: Could you get this to work?
To get around the problem of the long lines, I am inserting my own line feed at a point where I think it ought to be.
My eyesight is a bit problematic, so my preference is to use larger text than most people which makes the long lines
difficult to read.
The typical rafter to beam connection in your latest photo seems to be fairly substantial, so I would agree with your
earlier suggestion that the longitudinal beams are probably transferring most of the roof and floor load to the partitions.
Presumably there is a column in each partition under each beam, possibly a triple stud.
The partitions are still acting as shear walls and, even if they are not lined with plywood, should still be adequate
with drywall properly fastened on each side.
BA
RE: Could you get this to work?
By the way, I am with you on the large text, 150% for me.
RE: Could you get this to work?
BA
RE: Could you get this to work?
Dik
RE: Could you get this to work?