A-frame roof to flat roof connection "slippage"


  #1  
Old 03-17-16, 02:09 PM
M
Member
Thread Starter
Join Date: Oct 2005
Location: Florida
Posts: 1,631
Received 7 Upvotes on 6 Posts
A-frame roof to flat roof connection "slippage"

This is a rather complicated one to explain so it's going to be long with many pictures and sketches. Bear with me.

It is a single story residence located in Miami, FL. Concrete block exterior walls with wood framed roof. Monolithic concrete slab foundation. Built in 1955.

There is a sag on the roof. The sagged area is about 4-5' wide. Here is an overall roof layout. Most of the house is on an A-framed roof, but on the SE corner there is a flat roof section. The sagged area is just a few feet west of where the sloped roof and flat roof meet.



Here are two pictures of the roof above. The A-frame portion is shingle roof and the flat portion is flat deck concrete. You can see where the sagged area is at the bottom of the shingle section. I can't tell for sure, and it seems barely appreciable, but the roof ridge line may be off a little as well...but the amount is so little, it could easily be bad shingle work as well.





Next I went below to look at the location inside where it corresponds to the sagged area above, by transferring the measurements. You can tell, the sag is visible. The ceiling is sagging as well.



So this is not some roof deck getting wet and rot and bow, something is off.

Next I crawled inside the attic, to see what's up there. I can't get real close to the sagged area because there is just no head room. But with a flashlight I can tell there doesn't seem to be any roof leak or moisture penetration. The trusses look OK.







So the next step was I moved all the furniture out of the way and took down about a 30" wide strip of ceiling sheetrock where the sag is so I can see how the transitions look between the heels of the trusses and the flat roof framing. By removing the sheet rock in that room I exposed five trusses and I can see four of them have "dipped" at the heel where the transition is. The heel of the trusses and the metal brackets mounted on the flat roof 2X6s came apart a little. It appears out of the five trusses, one seem fine, the other four have dipped lower, from 1/4" to 1" or so. Not only the heels have dipped, the ends of the trusses have slipped off the metal brackets that attach them to the flat roof 2X6s, so that they are now at a "tilt".









Now the entire flat roof section is about 21' wide, the sag is only happening over a 5-6' portion of it. It turns out the room south of the sagged roof room, has a tie beam across that spot.





So it seems I need to jack up the trusses heels back to level, then install a beam across and under it?

I wonder how the trusses will behave as I jack up one end of it's heel for an inch. Would the 1"X6" roof deck pop and warp as a result? Can I use several bottle jacks across at the heels? Thanks for any comments and replies and sorry this got tedious.
 

Last edited by MiamiCuse; 03-17-16 at 04:22 PM.
  #2  
Old 03-17-16, 03:18 PM
Join Date: Mar 2005
Location: USA
Posts: 5,650
Received 1 Upvote on 1 Post
Did you have an engineer and get a permit?

On the face, it seems that bearing a concrete flat roof on a wood truss system is not very compatible, especially, when it comes to deflection. The stresses might be OK, but the mechanics are not so the load could cause superficial deflection/cracking.

Trusses work well for uniform loads, but concentrated loads on one location on a single truss and results are never good unless the truss is reinforced.
 
  #3  
Old 03-17-16, 03:45 PM
XSleeper's Avatar
Group Moderator
Join Date: Dec 2004
Posts: 27,061
Received 1,910 Upvotes on 1,716 Posts
I will have to study this a little more when I can enlarge the pics, but at first glance, it looks like someone removed a load bearing wall at some point and didn't header the trusses sufficiently.
 
  #4  
Old 03-17-16, 04:33 PM
M
Member
Thread Starter
Join Date: Oct 2005
Location: Florida
Posts: 1,631
Received 7 Upvotes on 6 Posts
Well, I am a licensed professional engineer but my specialty is in railroads, highway and drainage. I have done all the structures courses including seismic but I am just not that familiar with residential timber and trusses. Even the structures PhDs in my office are specialized in reinforced concrete and steel structures and at at a loss on small residential timber construction, I bride them with beers and shrimp cocktails but they are useless.

The flat deck concrete roof side has not deflected at all. The A-frame side has. I tend to agree with XSleeper that a load bearing wall was removed at one point. I did a permit search to look up previous improvements but problem is the city keeps records up to ten years...I can't believe this, they didn't scan or microfilm old records, they just tossed it. Being that this was built in 1955, I have no way to do any forensics on alterations.
 
  #5  
Old 03-17-16, 05:12 PM
B
Member
Join Date: Oct 2008
Location: New England
Posts: 9,460
Received 47 Upvotes on 43 Posts
Any similar homes in your area where you might impose to see how that roof was (maybe) supported?

Bud
 
  #6  
Old 03-18-16, 07:09 AM
Handyone's Avatar
Member
Join Date: Aug 2000
Location: U.S.
Posts: 4,807
Upvotes: 0
Received 0 Upvotes on 0 Posts
I'm no engineer, so some terminology might be off.

I going to call the board where the trusses attach a Beam (the worse beam ever).

A beam has to do two things: support the weight that's above it or attached to it (the sides) and the beam needs installed so that it cannot rotate. Your beam is doing neither and it looks like it was doomed to fail.

The idea I have is to completely remove the old "beam" and install new.
New beam would have to have a bearing surface at both ends and fit into pockets or have blocking to prevent rotation.
Once the new beam is in place the trusses can be attached to the beam with hangers. The new hangers won't fail like the old ones did.

That's obviously some pretty serious work.
 
  #7  
Old 03-19-16, 06:39 AM
M
Member
Thread Starter
Join Date: Oct 2005
Location: Florida
Posts: 1,631
Received 7 Upvotes on 6 Posts
Handyone, thanks. I will take that into consideration. I agree a beam is needed somewhere. Whether as a replacement beam to attach the trusses, or under the existing trusses to take the load off that beam, I am not sure.

I wanted to find out more about the original floor plan, was a wall removed? Was the flat roof section added later on as a "bad DIY"?

There is no permit record older than 10 years for ALL properties in this city. Some sort of a paper reduction act or whatever. I have had this happened in another city but they microfilm everything so there is an old archive you can get back to. Did some digging. Original architect is out of business and dead. I did find one neighbor who told me the flat roof section was added later on, in 1970s...by then it was part of the county and not part of the city yet, looked up the county records and the only thing I can find is the tax appraiser's office did noted the added square footage was added 1974 and additional tax imposed...that does not mean it was permitted either but in 1970s most of the hurricane code has not yet been put in place, I don't see any hurricane straps to tie the roof framing to the exterior tie beams. We are concerned about uplifts too.

In any event, whatever is here has been here for over 45 years. Of course, there may have been a wall that's removed later without permit later on.

So yesterday I decided to put up a temp wall right under the heel of the trusses. Putting double bottom 2Xs to not crack the tiles. Put double 2X vertical members right under each of the three sagging truss heels. Attached double 2Xs across the heels, and put three bottle jacks right underneath.









I slightly jack each up for about 1/4"-1/2" just to see. So far so good. In that I was able to remove the bow along that temp wall.

I am going to leave it like that for a few days while I scratch my head some more. I probably will go back say 10'-15' back on the trusses, find a spot where the web and bottom chord converge...and put up another temp support and jack that and see.
 
  #8  
Old 03-19-16, 11:49 AM
Handyone's Avatar
Member
Join Date: Aug 2000
Location: U.S.
Posts: 4,807
Upvotes: 0
Received 0 Upvotes on 0 Posts
I'm curious. I mentioned rotation of a beam.
Is the board that the trusses are attached to out of plumb? I assumed it was when I mentioned rotation, that would explain a lot such as why the trusses are separating.

And a note, I doubt that the room was original to the house. A load bearing exterior wall was probably torn down as I think was suggested.
I also don't think an architect would design the house with a flat roof. It is an easy way to add an addition though.
 
  #9  
Old 03-20-16, 05:55 AM
M
Member
Thread Starter
Join Date: Oct 2005
Location: Florida
Posts: 1,631
Received 7 Upvotes on 6 Posts
I'm curious. I mentioned rotation of a beam.
Is the board that the trusses are attached to out of plumb? I assumed it was when I mentioned rotation, that would explain a lot such as why the trusses are separating.
Yes the triple 2x6 was out of plumb. There is bending and shear stresses at that connection, but trusses don't behave quite the same as regular beams, but I do believe the bracket and heel coming out of alignment, the beam going out of plumb are caused by the same thing.

[uote]And a note, I doubt that the room was original to the house. A load bearing exterior wall was probably torn down as I think was suggested.[/quote]

As I mentioned above, I found a notation at the county tax assessor's office records that the added footage was introduced in 1974. So it's a 45 year old addition.

However I don't believe an exterior bearing wall was removed. In my sketch you can see there is a concrete tie beam on the southern half of the flat roof. That tie beam did not run all the way across. Only the southern half. Then on the northern half, there is no tie beam right underneath the roof connection line, but instead, a concrete tie beam 10' further east of it, right under the middle of the flat roof as illustrated.



I think more likely an interior wall somewhere under the trusses was removed.

Also it is possible the flat roof was original, but it was over an outside porch, and then it was closed in later.
 
  #10  
Old 10-01-16, 06:41 PM
M
Member
Thread Starter
Join Date: Oct 2005
Location: Florida
Posts: 1,631
Received 7 Upvotes on 6 Posts
OK it's been six months so an update (plus a few questions).

First, after I put up the temporary wall across the bottom of the trusses at their heels, and slowly jacked them up a bit here and there, wait for a week, jacked some more, after a month or so I got them to be nice and level all the way across, then I jacked them up by another 1/4' because I wanted to put a beam across that space eventually.

I left the set up in that position for a good six months while I was busy dealing with other things.

Meanwhile I found some more information.

(1) I ran into a neighbor who was visited by another neighbor who lived there in the 80s and 90s. He told me he knew the prior-prior owner and have visited the house. So I asked him to stop over and tell me what he remembered. Well, he told me yes the house has been modified, but not where I was working on. The modification was they closed in a porch...the roof and slab were already there, there wasn't any new structural additions. So whatever was added or altered was there since the 1980s.

(2) I tend to agree with that neighbor because I recently had to open up some sheetrock in that area to run some electrical...well these are 1" thick sheet rock. Hard like cement. Sawzall just bounced off it can't cut into it. Had to use an angle grinder with a diamond blade to cut. The sheetrock has three layers, a regular layer of gypsum about 1/2", then 1/2" of this cement layer which looks almost like stucco, then an eggshell layer over it, then like 4 layers of paint over that. The wood furring behind the sheetrock was attached to the concrete wall with hard cut nails. I don't think any remodel work done in the 90s or 2000s would involve this type of sheetrock and hard cut nail fasteners.

(3) The most revealing, is the truss heel affected the most, the one with the biggest slippage, I noticed drywood termite "dust" dropping off it. When I grabbed it by my hand, it feels like swiss cheese. That wood at the heel has gone bad. The other three adjacent heels seems OK.

My plan is to relieve these 2X6 members from the duty of supporting the heels of these five trusses, by putting up new beams under the heels. I will be using a set of double 2X8 as my beam.

NOT ONLY at the heels, but also two more locations. I have identified and exposed two more panel points of these trusses, and will also put up same beams underneath. So a total of six 2X8.

The single termite damaged truss heel with rotted wood, I will first patch the rotted area with structural strength wood epoxy filler. Once that's rock solid I will put up two triangle pieces of plywood gusset plates on each side and glue/nail the heck out of it.

I presented my 1-2-3 plan to a structural engineer friend of mine and he said it should work but I am overdoing it a bit.

Next I will post my first attempt at putting up a beam across the bottom of the truss heels with the temporary support wall in place.
 
  #11  
Old 10-01-16, 06:44 PM
M
Member
Thread Starter
Join Date: Oct 2005
Location: Florida
Posts: 1,631
Received 7 Upvotes on 6 Posts
OK so I decided I will first install the new beam under the truss heels, go back 5' or so and put another one under the next panel point, then back another distance to another panel point. So a total of 3 sets of double 2X10s.

First, I cut down some sheetrock on the two walls that are about 9.5' apart where the new beams will span. Solid poured concrete wall the entire wall top to bottom, sweet. I attached the 2X10 on there like an ledger board, using 1/4" Tapcon screws. Yes, I know Tapcons are not going to cut it for shear loads, so I will put in some 1/2" lags later. But in any case, this board will not be carrying the bulk of the load. The reason I put up the boards across the wall, is to provide an anchor point for the joist hangers for the three sets of beams.





The other side is part solid concrete up top, concrete blocks below, and a section of it is 2X4 wood studded wall. I attached like the other side the 2X10 across the top, using Tapcons on the solid concrete, and headerLok screws to the wood studs and top plate.





So now I have this temp wall across the space, and the two walls have the "ledger boards" in place.



I then raised the double 2X8 beam up against the old jacked up joist hangers. The beam has a real tight fit between the two sets of 2X10 ledger boards. Then I used a set of Simpson Strongtie double shear hangers HGUS28-2 under the 2X8s. I attached the hangers to the beam, not the header.

I then installed a 2X4 under the header all the way to the slab (had to cut away tiles to allow the lumber to sit on the concrete slab).

Then I put a 4X4 post under the double 2X10 header on each side of the wall.

So this is what I have now.

The beam under the original joist hangers supporting the truss heels. You can still see the temp wall to the right that is now a tad loose. I haven't removed that yet. It's there for insurance LOL.



The two 4X4 post with a 2X4 behind it, so it's like a 4X6 post sitting on the concrete slab.



Top of the post where the double 2X8 sits. The 2X10 header in this case is not really taking the load. On top of the post is the double shear hanger with the double 2X8.







Note that nothing has been permanently fastened yet.

I have not nailed in the joist hanger to the header, or to the beam. I have not tied the trusses to the beam with any straps, I have not attached the 4X4 posts to the walls. Everything is now just "sitting" there from the weight.

I now have a few questions.

(1) The Simpson Strong Tie instructions said to use 1.5" 10D nails to attach the hanger to the header. OK the 1.5" nails work because the 2X10 header is 1.5" thick. Behind that is solid concrete. But there are 2 sets of diagonal holes I am supposed to pound in to attach the beam to the header lumber. What size nails should I use? Surely longer than 1.5" since the skewed holes are 1" back from the header and driven in at an angle?

(2) It also said to use 1.5" nails to attach the hanger to the joist. In my case it is a double 2X so I have 3" of width. Should I drive in nails LONGER than 1.5"?

(3) The HGUS28-2 is specifically designed for using a set of two 2X8s. Yet when I wrapped it around the 2Xs, there is a gap of about 3/16" on the side. Why is that? Any idea? Isn't it supposed to be a snug fit? Am I supposed to shim that gap? Or put a piece of 3/16" thick filler between the two 2X8s? If you look at the pics below you can see the gap I am referring to.



Once I have this all sorted out and permanently tied in, I will move back 5' to the next panel point and put up another double 2X, then a third set. This is why I put a longer header across the wall on both sides.

Thoughts?
 
 

Thread Tools
Search this Thread
 
Ask a Question
Question Title:
Description:
Your question will be posted in: