Load path of unusual roof truss configuration 2

[Montreal.eng.]

Hi, I have come accross this type of roof truss layout 2-3 times while inspecting attics in mid-sized residential light wood framed buildings.

I'm wondering if anyone is familiar with this type of roof truss configuration where there seems to be a truss beam supporting 2 spans of roof trusses. I have a feeling that the roof trusses are supported on the exterior wall on one side and the beam on the other. The beam being a 4ply-truss itself carries the load to the exterior walls as well.
It would be great if someone can confirm or explain the load path on this roof and why they wouldn't use trusses that span from end to end. Thanks!

 

[phamENG]

Well that's pretty nifty. Haven't seen one of those before.

Answer is fairly simple - to span trusses that far would be impractical. Too much stress in the members, too large to ship, handling difficulties, etc. That looks a pretty large roof system.

 

[Greenalleycat]

I'm a bit confused by the pictures to be honest but, from the description, I think you're saying there's a deep 4 span truss running one direction with smaller trusses framing in perpendicular on both sides?
If so, that's fairly common here, we refer to them as girder trusses

 

[Montreal.eng.]

I think you're saying there's a deep 4 span truss running one direction with smaller trusses framing in perpendicular on both sides?
If so, that's fairly common here, we refer to them as girder trusses

Basically, there's what seems to be a girder truss as you say, composed of 4 seperate trusses (see pic 4&5) and the perpendicular pitched roof trusses seem to sit on the girder truss and tie together with truss plates (pic 1,2&3). I wish I had taken better pictures, I was up there for a different reason.

My guess is that the roof system still bears on the exterior walls like a standard truss roof but instead some of the load gets passed on to the truss girder that in its turn also bear on the perpandicular walls creating a point load on them.

The reason why I'm asking is because we had a client ask us if one of his walls on the top floor was load bearing and we weren't sure whether it was or not since it almost alined with the girder in the attic and we had never seen a layout like that before.

 

[phamENG]

If it "almost aligned with the girder" then I'd say no, it's not load bearing. That's a beefy girder truss. It's what, 4 or 5 feet tall? At 4-ply, I'd be surprised if it had an interior bearing point if the span is less than 50 feet or so.

 

[Montreal.eng.]

Yea it was pretty massive about 4 ft tall. Thanks for the input, it confirms what I was thinking as well.

 

[phamENG]

Hmm...my 50 ft number might be a bit long. It does look like there's a splice/interior bearing point in that first picture now that I look at it again. The broken pieces of plywood are an interesting feature...

 

[Montreal.eng.]

If my memory serves me right, I believe at that distance from the exterior wall to the left where the splice is located, we were in the hallway not in the appartment. It could be 2 spans of a girder truss.

 

[KootK]

Having spent some time in the metal plate truss industry, here are my observations:

1) This is a pretty rare setup and, in fact, I've never seen one quite like it. The closest analog that comes to mind is that, sometimes, similar things are done with upset LVL or timber beams. In those situations, however, one normally splits the supported trusses such that you've got pairs of trusses coming in from either side with notched bearings rather than single, long trusses with pockets.

2) Some things that would normally steer a truss fabricator away from doing what we see here:

a) I would think that erecting truss over such a tall pocket would be awkward.

b) The absence of a true flexural splice across the pocket might impact normal lifting procedures.

c) This makes for a bit of a weird condition with respect to bracing the top chord of the girder which will have oodles of compression force in it. Yeah, there's something there for bracing in the form of strong axis bending of the truss verticals. I wouldn't normally call that a very positive or stiff bracing setup, however, without some numerical justification to back that up. Moreover, the lateral bracing of the truss verticals themselves is a bit dubious. This is one of the reasons that girder truss profiles are usually extended up to the sheathing plane. All this said, this is an existing structure with history. So far, so good.

It does look like there's a splice/interior bearing point in that first picture now that I look at it again. The broken pieces of plywood are an interesting feature...

3) I believe that to be a location where two, independent, simple span girders abut. Per 2c above, I believe that the plywood may have been installed as a way to take up the slop in the rotational bracing "system" for the ends of the two girders that meet at this location. 2 parts clever; 2 parts scary as it were...

4) Save the business of sorting out how many girder spans we have, I agree with OP's general assumptions regarding the load path.

 

[jayrod12]

) Some things that would normally steer a truss fabricator away from doing what we see here:

a) I would think that erecting truss over such a tall pocket would be awkward.

b) The absence of a true flexural splice across the pocket might impact normal lifting procedures.

I believe it is two distinct simple span trusses on each side sitting on an upset girder truss. Most of your point 2 is moot.

I've seen this sort of arrangement where long clear spans were desired. In fact we did it at the cottage to get the Roof look and the open floor plate we were after. There isn't a interior wall that supports roof. Makes remodeling in the future easier.

 

[phamENG]

I believe it is two distinct simple span trusses on each side sitting on an upset girder truss.

Agreed. If you look at the third picture, there's no truss plate crossing the joint where they abut.

 

[Montreal.eng.]

I believe it is two distinct simple span trusses on each side sitting on an upset girder truss

If you look at picture 3, there isn't a single truss plate connecting the two spans together untill you get to picture 2 where a plate connects the 2 spans right above the girder creating a pocket.

 

[jayrod12]

I'm still confident it's two distinct trusses. In the second third picture it's actually 3, two clear span, and one piggy back on the right side of the joint.

 

[phamENG]

Montreal.eng. - that's not a truss plate crossing the whole pocket, that's two truss plates that are almost perfectly aligned and the resolution of the picture makes them blur together a bit.

 

[RontheRedneck]

I have designed truss systems like that on large commercial buildings. But never on a residential structure.


KootK mentioned the TC bracing of the flat girders. That's an issue which sometimes gets overlooked. In this case the trusses on both sides would brace the TC, so I don't think it's an issue.

Somewhere in the thread someone mentioned the possibility of a bearing wall. It's impossible to say what may or may not be bearing since I can't see plate sizes and web configurations.

It looks like the trusses on each side of the flat girders are fairly long, and the spans of the flat girders appear to be fairly long. So the reactions of the flat girders could be 10,000# or more. Hopefully they have adequate bearing where needed.

When people ask me about what is or is not bearing, I always tell them to look below a wall in addition to looking above. In most cases there will be a beam or bearing wall in the basement.

Of course I'm used to the typical midwest basement, and they're not common everywhere.