Residential Window Walls

[medeek]

Yet another window wall, these things seem to love me, I don't love them. A couple issues with this window wall:

1. I've got 24" of wall at the corners and about 18" between the central door/window and side windows. I've also got a 7' to 8' pony wall underneath this window wall. Obviously I can't get any traditional shearwalls into the main window wall so my first thought was to use Simpson's (Wood) Strong-Wall (catalog C-L-WSW16) product which can be field trimmed to match the pitch of the roof and also is available in mult-story kits. Looking at the literature on the product it appears that it is always intended to bear on a concrete foundation and not other wood structures or members, hence the need to possibly use a multi-story kit or just balloon frame all the way up from the foundation to the inclined double top plate (ie. use a 20' strongwall).

The other option possibly would be to ignore the window wall entirely (as a shearwall) and do a three wall analysis treating the diaphragm as rigid and ignoring the window wall shear resistance. Section 4.2.5.1.1 of the SDPWS 2008 appears to allow for this since the great room is 25' wide and 16' deep with an L/W ratio of 0.64 and the diaphragm length (L) is less than 25'. Has anyone ever used this methodology in this type of circumstance?

2. The second issue appears to be a design flaw. The ridge beam needs to be supported at the gable end with a small beam that spans the polygon window above the door. How does one fit a beam in the space provided? I'm thinking we will probably just need to drop the window height until an appropriate size header can be inserted.

For reference the site criteria on this job is:

Roof Snow Load: 50 psf
Wind Speed: 85 MPH (110 MPH Ult.)
Exp. C
SDS: .739

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[XR250]

Why don't you want to nail the laminations together?
I never spec LSL beams except as bands. I have enough problems with way over-designed LVL's sagging on me. Don't wan't another animal to deal with.
Also, they simply don't stock them at the local yards.

 

[medeek]

I always worry about the side loading if you have to hanger a beam from another nail-lammed beam. A single beam eliminates this problem. I'm probably worrying about nothing.

Here are a few numbers so far.

I have a 50 psf flat roof snow load -> sloped roof snow load of 50 psf

My glulam beam spans 16' and a 5-1/2 x 15 beam seems adequate for the job. My end reaction is 7464 lbs (2194 D + 5270 S).

Half of this reaction is picked up by each cantilevered beam, 3732 lbs.

With the cantilever beam the deflection and the bending control. I end up with a 3 ply (5-1/4" thick) LVL beam with the down force at the PSL column of 7179 lbs and an uplift at the end of the wall of 810 lbs.

I'm very nervous about the hanger at the rake beams. My SDS 1/4" screws attaching my HGU5.62 hanger are going to be quite closely spaced with 18 per rake beam and about 3" from the end of these beams, splitting may be an issue. It might be better to cut the inner ply at a scarf and then splice the two rake beams together and then hanger from this section. I probably need to draw a diagram to illustrate the point.



A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[EngineeringEric]

As for the side loading. There are published calculations for LVL with side loads to determine what type of fastener and spacing is required

 

[medeek]

I know I've got my copy of TJ-9000 open right now, looking at this page. I'm not saying it can't be done its just one more thing to hassle with and then spec out on the drawing so that the contractor doesn't mess it up. A single ply beam is more convenient.

Simpson calls for 2-1/2" SDS screws for the HGU hanger series this will really only penetrate the first LVL ply.

My idea might not have any merit but below is a quick sketch of the two rake beams spliced together and then the hanger attached to the splice:

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

I'm actually starting to think Kootk's idea of running two beams instead of one would be a much better solution. I've already spent way too much time mauling this whole thing over today but I'm still not satisfied with the solution.

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[XR250]

Sounds like a better idea. I would not try that connection with that kind of load. Mine ridge reactions typically max out at about 2500 lbs. when I use that system.

 

[medeek]

So the architect likes the idea of the dual beam design with cantilevered roof joists, time to rerun all my numbers. I am still going to keep the LVL rakebeam (1 ply only) to use as part of my moment frame, however it will no longer be carrying any significant gravity loads and I will use a 14" deep LVL to match the I-Joists of the vaulted roof.

The one issue now is the architect wants the PSL post centered in the 18" wall section, how exactly the additional studs and trimmers frame out around it is a bit of a mystery...

I've also never encountered a cantilevered roof joist system like this before. I'm planning on installing a strap across each joist to the opposing side but beyond that is there anything else to look out for?

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

I think I am going to go ahead with the wood moment frame system on this wall. The wind loads are fairly low as are the seismic (comparatively speaking) so I think I can make the numbers work. However, I am interested in broadening my horizons so that I can do a window wall design in steel in the future. What are some resources for doing this type of residential design with steel? I have a number of steel text books and the code books already but these really don't provide much direction as far as what a typical design should look like.

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[rn14]

If you go steel you can go with an ordinary moment frame and then the detailing is easier but you have to take a hit on the R value and that will carry over to your wood shear walls too unfortunately so we usually go with a SMF. The problem with the special is that there are requirements for lateral bracing of the top and bottom flanges of the beam and this isn't very possible with wood framing. It seems to be conveniently ignored to some extent around here. Here we would probably have a two story moment frame with a bolted flange plate (BFP) connection at the upper sloped beam and a reduced beam section connection (RBS) at the lower horizontal beam.

 

[medeek]

I understand the terms OMG and SMF but the rest is Greek to me, I might as well stick with wood for now, at least I understand what I'm dealing with.

With the two beam design the 14" deep TJI Joists will be resting on the glulam at a 7:12 pitch. Based on the TJI documentation you cannot birdsmouth cut I-joists on the upper bearing points, only lower bearing. What do you typically see done in this situation, beveled glulam, beveled bearing plate, variable slope seat connector?

Based on TJ-4000 I will also need to nail (2) 8D nails from the I-joist bottom flange into the glulam and also install a backer block and twist strap (slope greater than 3:12).

The architect also wants to use flat 2x4 outlookers (gable overhang of 18") however I don't see how one can notch around the flange if they are oriented this way (detail O, page 30).



A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[XR250]

With the two beam design the 14" deep TJI Joists will be resting on the glulam at a 7:12 pitch. Based on the TJI documentation you cannot birdsmouth cut I-joists on the upper bearing points, only lower bearing. What do you typically see done in this situation, beveled glulam, beveled bearing plate, variable slope seat connector?

Quit using those god-awful TJI's and your problem is solved!

 

[medeek]

What's wrong with TJI's? I like their product better than LP or BCI, in my opinion its a better product. How else are you going to span these floors and roofs? Floor trusses are too expensive and in my opinion are not stiff enough, the joints have a miniscule amount of play in them but it all adds up so floor performance suffers.

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[jayrod12]

I also prefer trusses like xr250. I've seen too many rotten tji joists from minor water infiltration.

My hands are tied a lot of the time.

 

[medeek]

Here is what I have so far. I think I have all my load paths covered but I will maul it over during the weekend and gladly consider any feedback from the board. Somewhat of a non-standard portal frame (moment frame) going on, I'm still figuring out how to analyze it fully. My strap and holdowns are still pending my final lateral numbers. The ambiguity is caused by assuming a rigid diaphragm so the lateral load to this wall is somewhat governed by its rigidity compared to the other three walls of the room. How to assign a rigidity to this construct is where I'm scratching my head at the moment. More rigidity will attract more load which will drive up my straps and holdowns...

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

Forgot to add the pad footings for the PSL posts

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

I created a simple moment frame model in RISA2D, to look at the forces involved. The worst case scenario is only about 1,500 lbs lateral from wind with a flexible diaphragm, rigid diaphragm gives a much smaller lateral load. The results are shown below:

Axial Loads:

Shear:

Bending:

Numbers:

The problem with the model is that it is a little to simplistic. I don't have a member type that I can use to simulate the shear wall panels (RISA's wall panel aspect ratio limitation did not let me use a high aspect ratio wall panel), so I just used the same 3.5 x 11.875 LVL for all the members. Varying the stiffness of the shear panel will affect the moment distribution at the top and bottom of the shear panels.

One interesting thing that popped out was the zero moment at the peak. If the load from the diaphragm is balanced this is the case and there is really no need for a moment resisting connection at the peak. Granted the wind loading will not be balanced but overall the moment at the peak will probably not be significant.


A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

The largest moments are at the shear panel base as expected. I have 24" of width at my shear panels so my worst overturning force will be: 2376 lbs + 385 lbs = 2,761 lbs

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[medeek]

I've created a SketchUp model to help identify any further problems with the design. I just realized that the floor joists will not span the 25' of the room and a floor beam will be required at mid-span.

View model and download SketchUp model here:

https://3dwarehouse.sketchup.com/model.html?id=cb424537-7ad2-4afb-a53b-1ab547feec26

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com

 

[XR250]

What's wrong with TJI's? I like their product better than LP or BCI, in my opinion its a better product. How else are you going to span these floors and roofs? Floor trusses are too expensive and in my opinion are not stiff enough, the joints have a miniscule amount of play in them but it all adds up so floor performance suffers.

I actually meant I-joists in general. We use 2x10's and 2x12's here for most roofs . I-joists if we have to. The framers and contractors prefer the dimensional lumber.

For floors, I use mainly 2x10's or 2x12's with steel I-beams or LVL beams if needed.
Sometimes I-joists and floor trusses, if there is no other practical option.
Here is my problem with i-joists (based on my field experience)

1) difficult to frame with if sloped or skewed
2) The flimsy OSB bands suck for point loads and do not like water. Deck attachment is sketchy at best. I usually spec LVL bands behind decks.
3) Make renovations difficult as they are difficult to modify
4) They don't hold up well in fires (this is anecdotal but believable)
5) Blocking point loads thru them is difficult and is rarely done well or at all. I have seen numerous I-joists buckled from point loads. Never seen a 2x10 do that.
6) The loads from the long spans joists and their skinny profile cause them to dig into dropped beams. The blocking between them does not, so it causes a hump in the floor and will pop floor tiles
7) They do not hold up to water (as Jayrod said)
8) I have seen too many instances of them sagging alot more than they should under sustained loads - such as heavy kitchen islands.
9) They are expensive! - which is why tract builders try to span them out in length and spacing.
10) They are not designed to be used in typical roofs without structural ridge beams. This requires the expense of the beam and the expensive hangers.
11) Cannot notch them and use ledger strips - like dimensional lumber (further adds to the cost)
12) Another source of off-gassing in a house (I am guessing)
13) They seem to shrink in length just as much as dimensional lumber.

I could go on and on, but I am tired.

 

[medeek]

Added the pad footings and a few more straps:

A confused student is a good student.
Nathaniel P. Wilkerson, PE

http://www.medeek.com