Framing of residential home cathedral ceiling

[71corvette]

A contractor friend of mine has asked me to evaluate a conceptual roof framing idea he’s come up with for a single family residential home he’ll be building in Vermont. Most of the roof framing will be very conventional, but one portion of the home will include a cathedral ceiling that spans the full width of the house (25’-6”). Rather than purchasing trusses he’d like to stick frame the entire roof. To do this he’s come up with a framing system that I’ve not seen constructed before (detail attached); essentially he’s proposing to use the rafters and ceiling joists together to form a scissor truss of sorts.

I plan to run some numbers on this concept tonight, but in my experience I’ve seen mostly trusses or structural ridge beams used in these types of applications. Given this I thought I’d solicit opinions and feedback from those of you here who may have dealt with similar circumstances in the past.

Detail:

http://files.engineering.com/getfil...1483-46f6-abad-307f70feb517&file=00000001.JPG

Any thoughts would be appreciated.
Thanks!

 

[ishvaaag]

Well, at least is not an "invent". It is a simple variant of what is called "scissor" truss ("cerchas de tijera", in spanish). You will see that some other scissor trusses add more members and so you may choose to do, but at 25 ft the complication may not be warranted.

 

[Ron]

That's do-able....I'd use bolts instead of nails, though. A single bolt through the cross-connection would allow a bit of movement without inducing undue stress in the members.

 

[skibum2b]

How does he anticipate attaching the ridge beam to the rafter?

 

[msquared48]

I would add a king post at each truss to help with the "Ridge Beam" support. Other than that, looks very dooable. Go for it!

Mike McCann
MMC Engineering

 

[IRstuff]

A more conventional scissor truss:

http://www.askthebuilder.com/B348_Scissors_Truss_Design_.shtml

TTFN

FAQ731-376

 

[BAretired]

The structure can be made to work, but it is not strictly a truss because it is not composed completely of triangular panels. A vertical member at the center would make it a truss in the conventional sense.

With or without the vertical member, the structure can be analyzed and appropriate connections can be made so that it is structurally capable of carrying all loads applied.



BA

 

[JAE]

And I would think the critical issue would be the connections, not the members. As Ron suggested, bolts might be better than nailing.

 

[RatherBeSurfing]

Very do-able. However, be very careful to account for horizontal thrust at the bearing wall. Without a true "ridge beam" you will have large horizontal forces at the bearing walls. Even with a true ridge beam, you will still have significant forces causing the walls to bow outward as the ridge beam deflects if you do not design the ridge beam stiff enough. Two points to pay special attention to: 1- design for a very stiff ridge beam, and 2- pay careful attention to the connection details (as was metioned before). Other than that, looks like duck soup!

Aloha bruddahs,
Keawe

 

[BAretired]

You will not have large horizontal forces at the bearing walls because the bearing walls are not capable of resisting horizontal forces.

What you WILL have is horizontal displacements at the top of the bearing walls. The magnitude will depend on the stiffness of the trusses.

BA

 

[71corvette]

Hey guys, thanks for your input. I figured those would be the responses, but it's nice to get a second opinion or two!

I ran some numbers on this last night and was finding some rather large horizontal deflections at the truss supports. In addition, the connection design would be a bear. In the end I think it may be easier and cheaper to either buy trusses or purchase a structural ridge beam.

Thanks again for your thoughts.

 

[Stillerz]

Framing that by hand would be a waste of time and money for the homeowner.
Trusses would be, cheaper, 100 x's easier to construct, and make for a much neater job since making a ceiling like this is not as easy as it looks.
Getting a nice flat ceiling using less than perfect lumber can be a real bitch.
It's just a waste of material.

 

[Splitrings]

71corvette,
I think you will find that you will need a larger member for the bottom chord than the top chord. I quickly modeled your "truss" in RISA 3D. It indicates (if my model is correct) that a 2x6 SPF #1 BC won't be anywhere near adequate. I used a balanced snow load (Ps) of 42 psf which is come for VT. Another thing to watch is there are some pretty substaintial loads at the joints. It will take more than a handful of nails or single bolt to resist.

I quickly added a web member from the peak to the intersection of the BC's and that reduced the bending moment considerably in the BC. I can't spend anymore time to convince myself why this is.

I also happen to have a scissor truss design from a company out of Canada sitting in front of me. It has a 2x8 BC and a 2x6 TC. Some of the joint loads are huge, 6000# plus. 32' bldg, 4' o.c. truss.

 

[BAretired]

splitrings,

It is not obvious to me why the bottom chord should have large bending moments with or without the suggested king post. Except for ceiling load, the bottom chord is a tension member. I would expect the top chord to be subject to significant bending moments and thus larger than the bottom chord.

BA

 

[Stillerz]

I agree with BAretired here...
I still contend that any time spent on this will ultimately be seen as a waste.
Get the trusses!!!!

 

[rday]

I'm with BA on the chord size issue. I agree the connections will probably need to be bolted.

I will say, I see truss companies model scissor trusses with pin-pin support conditions regularly. I think they should be modeled closer to a pin-roller configuration unless we account for the thrust load. Tough to do on top of a wood frame wall with a vaulted ceiling. As BA pointed out.

For the detail shown in the OP, I think in a pin-pin configuration the bottom chord will load in compression and maybe be larger than the top chord. Check the support reactions, can you truly resist that horizontal load?

 

[71corvette]

I too am unsure why the bottom chords would need to be larger than the top chords. Based on the quick model I put togheter last night the snow load on the roof was requiring the top chord to be larger than the bottom chord (in my model the bottom chord acted primarily as a tension member).

In any case, my contractor friend called a lumber yard today and they designed a structural ridge beam for him. The total price was $250 for the beam (2-14" deep LVL's). In the end this was a much simplier solution than engineering the full roof system.

 

[Splitrings]

If you analyze this assmebly as a truss, all members are broken at the joints and are pin connceted, I agree the TC will need to be the larger member and the BC becomes solely a tension member.

But your assembly isn't a series of pin connected members. Both the top and bottom chords are continuous through the joints. This is also the reason for the moment in the BC.

 

[BAretired]

If all members are pin connected at every joint, the structure as depicted is a mechanism and will collapse. To prevent this, either the top chords must be continuous or a vertical member is needed at midspan.

BA

 

[Stillerz]

I was going to suggest just sizing a ridge beam. The 25'-6 spans is not very wide...just wasnt sure how long (in/out of the page on your sketch) the ridge would have to be.
Use a nice deep rafter for insulaating and you'll have a nice wide open space.