Rafter Thrust 2

[Lake06]

I am involved with a repair for a newly constructed home. The home has a valuated ceiling area and it was determined that a ridge board was utilized with collar ties located in the upper 3rd height. I realize that this should have been constructed with a ridge beam and not a ridge board. The home has all its finishes completed.

Owner would not like to utilize a ridge beam for the repair because this would involve ripping up the top of the ceiling and the height would make it difficult to install. I am thinking of utilizing 3 lvl boards (14”x1.75”) laid on their side and 2 lvl (9”x1.75”) running vertical to form an inverted T shape. This beam would run parallel to the top double plate of the house were the rafter meets the exterior wall. (I will post a picture of this detail later). I would install 2x12 joists which would tie the rafter into this lvl beam at 24” o.c. The double top plate and proposed beam would run parallel but would be aprox 1.5’ apart. There would be a proposed beam on each side of the home.

The home has a 12 on 12 roof slope with 2x12 rafters at 24” o.c. The valuated ceiling area is 28 feet wide by 20 feet long with the ridge running in the 20 foot dimension. The rafter is currently nailed into the double top plate with 5, 8d common nails and has a hurricane strap with provides 160 lbs of lateral resistance. I have calculated the lateral thrust due to dead load to be 210lbs per rafter. The lateral thrust due to dead and snow load was calculated to be 760 lbs per rafter. The exterior walls are 2x6 construction with ½” osb sheeting.

I have seen this detailed before to utilize this lvl beam but not sure of how to connect the beam at the ends. The lateral reaction would be very high at the ends of the beam. I was thinking of utilizing a 3/8” dia cable. This cable would be located at the ends of the beam and tie the beam into the opposite proposed beam located on the other side of the room. The cable can be hidden in the wall. Does anyone have any thoughts or comments on this? I guess the alternative would be installing rafter ties spanning the room located 4 feet above the top exterior plate spaced at 4 feet on center with ties nailed or bolted to both sides of the rafter.

 

[BAretired]

What do you think of running the proposed beam as shown in my attached detail above?

1) I don't understand the purpose of the vertical member.
2) If the horizontal beam is capable of carrying the thrust, it should work but you still have the problem of connecting it at each end.
3) It is not very attractive in appearance.
4) I think Mike's idea is simpler and more economical.

BA

 

[Lake06]

BAretired

1) The vertical members are not needed for thrust but help with deflection of the beam
2) I was thinking of maybe utilizing a 3/8" cable at each end and connecting it to the proposed beam on the opp. side of the room.
3) I was envisioning this beam would be boxed in with sheetrock and maybe some recess lighting located below.
4) I agree that Mike's idea is simpler and more economical.

 

[msquared48]

Lake06:

It is my professional preference to have all blocking full depth. However, there have been other discussions in the structural forums here supporting partial height blocking. Depends on the force seen. I believe in the other threads though we were talking about the edge shear force of the diaphragm, not the chord force - two different animals.

Partial height blocking can work, but it does put the rafters into cross-grain bending. How much you tolerate is your call if you use this route. I don't.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[Lake06]

Msquared48

If I were not to use full depth blocking then I would assume the blocking would be fastened to the adjacent rafters and not the osb. Is this correct? I am also thinking that If full depth blocking was used then how would the air flow travel from the overhang to the ridge. I would think you would want to leave at least 1.5" between the top of the blocking and the bottom of the osb roof diapragm. Also how thick would you recommended making the blocking? As wide as the double top plate (5.5"). If you wouldn't mind I'd like to sketch up a detail and have you take a quick look. Thanks

 

[msquared48]

Holes are provided through the blocking as needed when full depth.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[dhengr]

Lake06:

When I read the OP, I thought, that’s pretty typical McMansion sexy design and construction, as long as the granite counter tops look good, everything must be O.K. They use the prescriptive rafter tables in the IRC but don’t bother to read the footnotes, verbiage, and code sections on the need for rafter ties; and the difference btwn. rafter ties, raised rafter ties, and collar ties, and what they do and how they work. And, I wondered, where and how is the distress showing up; and you say there isn’t any, and I add, not yet. To fix this problem now is going to involve some demo and heavy lifting however you cut it, and it’s likely going to show. So the question becomes aesthetics. And, since you asked, your solution ain’t pretty, unless you call it a light trough around the room. It can be made to work, but you see that the end reactions become the big structural problem with this scheme. And, steel rods rather than cables would be better because cables stretch too much on loading.

I agree with MikeMc’s general discussion of diaphragms and chords tied into the double top plate. And a 20x20 diaphragm on a 12/12 pitch is certainly a manageable diaphragm, not very large. It is very good at distributing lateral loads to the four walls (wind, EQ, etc.), but much less good at preventing wall bowing, because of its high pitch, and smaller horiz. resisting component against this action. It will inhibit some of this bowing, but I suspect the wall will bow, and this is accompanied by out-of-plane distortion of the roof diaphragm, it doesn’t take much. You don’t know how the sheathing is nailed to the rafters, you have difficult blocking details at both chords (walls and ridge and demo. to install), and you don’t know how it is nailed to the gable ends. So, now design your diaphragm. The damn ridge board can deflect down, which causes thrust and spreading, and some rotation a joints, and the roof plane just flexes out-of-plane to accommodate this movement. Mike’s argument, as well as his diaphragm, would be stronger in resisting this action if it were on a 5/12 or 6/12 pitch. And, obviously, the roof diaphragm would be even less effective at a higher pitch. Many times it works fine, and probably would here too, but I have always had trouble analyzing and rationalizing this condition. The way I think of it is that the whole roof system just relaxes a bit and tends to push the walls out a bit at mid length. This diaphragm is probably stronger and stiffer than we can give it credit for in our designs, given its proportions. It’ll likely act as a diagonal tension field btwn. the gable ends.

The big question is, are the rafters adequately fixed to the top plates, and the plates to each other, etc., to transmit ever other rafter thrust. MiketheEng. suggested the cleanest, simplest fix, 1" dia. wrought iron rods at 4' o/c, and those make nice mistletoe hangers, etc. I’d try to sell them on fake trusses at 4' o/c, with raised bot. chords as high as your rafters will allow, assuming you know what grade they are, and you can hang a light or fan from the bot. chords of those. These two schemes really address the problem most directly. Then, the design problem becomes the connection, in tension, to the rafters. Rafter ties are a good thing, we can actually understand them and how they work. If it’s a new house, back charge the builder.

 

[msquared48]

dhengr:

you said above that "Mike's argument, as well as his diaphragm, would be stronger in resisting this action if it were on a 5/12 or 6/12 pitch."

I have to strongly disagree with this as exactly the opposite will be true. The higher the pitch, the more the diaphragm will act as a ... deep beam ... instead of a horizontal diaphragm. This is in respect to the ability to pick up the lateral thrust of the vertical loads, which will decrease with increased pitch, and transfer it into diaphragm shear and chord forces.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[Lake06]

Now I’m really stumped about which option to pursue. If I went with the interior beam option I would change the use of cables to a steel rod at each end. I do agree the cable would expand under heavy loads. I like the idea of the wood blocking to transfer the load down into the double top plate and would look the best but I don’t know if I would feel comfortable with recommending this. Rafter ties are always a good idea and are recognized as a means to prevent lateral thrust. I guess it wouldn’t be the end of the world if a few ties had to cross the ceiling. Then the question is whether to use iron rods or dimension lumber. Being a 28 foot span at the top plate I would have to use a glue lam board, due to the length requirement. I think the longest dim lumber they make is 24 feet. However I could locate the ties about 4 feet above the top plate then I would need a 20 long board. I would imagine I would need two, one nailed or bolted to each side of the rafter with spacer’s in-between to satisfy the unbraced length. Then wrap the ties in pine to look like a beam or possible put small lights in-between or hang below.

 

[BAretired]

Lake06,

It is your call. You cannot expect to get answers from this forum because none of us know precisely how the existing building was put together.

The structure will try to behave like a folded plate. If it does, the ridge board cannot deflect far. If it had been a ridge beam, it too could not deflect much because the diaphragm action prevents it from doing so.

For that to happen, a few conditions must be fulfilled:

1) The deck needs a reliable tension chord at the outside. This could be provided by a continuous fascia board attached to the deck. Seems pretty straightforward.

2) The deck must be adequately nailed to the gable walls at each end to carry the in-plane forces.

3) The gable walls must be capable of resisting the sloping reaction of the diaphragm as a shearwall OR the top plate of the gable wall must be connected at the ridge to develop the necessary tension from each diaphragm and must be adequately supported from below to resist the vertical component of that force.

If you can satisfy these conditions, the problem is solved but we can't because we have not seen the structure.

BA

 

[4thorns]

Would it be possible to run a beam on each side, in the 20' direction, under the ends of the collar ties where they meet the rafters? Collar tie connection and end support for the new beams would be issues to look at (among others) but might be an option.

Similar to this thread but using Glulams for example instead of walls.

http://www.eng-tips.com/viewthread.cfm?qid=297098

 

[BAretired]

4thorns,

Yes that would be a structurally acceptable solution. It may not be an architecturally acceptable solution, however.

BA

 

[4thorns]

BA. The OP mentioned wrapping collar ties with pine to make them look like beams. After almost 30 years of working with wood and having a boss many years ago say "Keep throwing pine at it until it looks like it's done", it just seemed like a good option.

 

[csd72]

BA,

we seem to be thinking along the same lines though I did not explain it very well. The rafeters will bend between the eaves and the ridge under transverse loads and the reaction at the ridge will create axial load in the opposite rafter (and vice versa) the reaction of this is transferred to the end walls via diaphragm action.

Action under wind e.t.c will be a little different.

 

[BAretired]

Okay, csd. Sorry if I misinterpreted your earlier post.

BA

 

[Lake06]

The client would like to install rafter ties, he thinks the ties will provide a nice look with recessed lighting located below. However he would like to place them at 6'-0" o.c. and our existing rafters are spaced at 2'-0" o.c. Local butilding codes state that rafter ties shouldn't exceed 4'-0" o.c. My question is if i design the rafter tie connection to handle the thrust of a 6 foot spacing would this be acceptable or should i restrain the rafters between the 6'-0" spacing, open to suggestions. Thanks

 

[ToadJones]

maybe design the collar ties to take the load from the 6'-0 span and then design bridging of sorts in the 2'-0" spaces to get the load to the tie

 

[Lake06]

ToadJones

I just want to make sure were on the same page. When you say collar tie are u referring to the tie located in the upper 3rd or lower 3rd of the rise above the double top plate? What i'm calling the proposed rafter ties are located right above the existing double top plate (rest on top of plate) There currently are collar ties spaced at 2'-0" o.c. located in the upper 3rd height near the ridge board.

What kind of blocking between the rafters and should they be located, i'm assuming they would be located on the top plate between the existing rafters, would you suggest using? I have calculated the thrust PLF along the top plate to be 380LBS for a balanced snow load condition.

 

[ToadJones]

ok-
I haven't read through the entire post as it is older...
I would think that if your double top plate is capable of transferring (laterally) the load with the ties being every 6'-0 you would be ok, no?

If I am picturing this correctly maybe you could build a bulkhead of sorts that could work as a lateral beam transferring the load b/t ties.

M^2 might have some ideas here.

 

[msquared48]

You can design the collar ties for a 6 foot spacing. No problem, just justify your design with calculations. No different than open trusses at the roof.

You may need a dedicated ridge beam member to span between them, and you may have to double up on the rafters to double your bolt values for the collar tie to double shear and to help control lateral spreading of the walls.

Mike McCann
MMC Engineering

 

[Lake06]

msquared48

In your last post you mention I may need to use a dedicated ridge beam member to span between the ties. Can you provide more detail of what you mean? Are you refering to installing blocking at the top of the double plate to transfer the load between the rafter ties? Please see attached sketchs 1-3. Sketch 1 and 3 do not show the pine boards and recessed lights for clarity. Sketch 2 is a cross section of the proposed rafter tie (6'-0" o.c. spacing) which consist of using 2, 1.9E 1.75" by 7.25" microlams, one on each side of the existing 2x12 rafter.