old bowstring truss reinforcement

[diggingdeeper]

my client has a warehouse which he wishes to sell with 70' span bowstring trusses, several of which are broken. the top chord is 4-2x6s flat, apparenetly nail laminated together as no bolts are visible. the bottom chord is a single 6x6 spliced in the middle. the webs vary from 3x3 to 3x4. the panel point connections are a U plate configuration with a single bolt through each web and the wings of the "U" and two bolts through the base of the "U" and through the chord, top or bottom. the heel appears to be a thrust plate receiving the top chord and welded to two side plates bolted through the bottom chord.

i would really appreciate any insigths into these trusses, particularly in the analysis of this type of panel point system. thanks!

 

[BAretired]

Sounds like a real can of worms. Under uniform load, the web members carry virtually no load because the top chord closely follows the bending moment curve.

The worst case for web members will be unbalanced snow load. In the Canadian code, this can be a pretty severe condition because snow blows clear from one side and is dumped down on the opposite side. You would want to study the web/chord connections to determine whether they are adequate to handle unbalanced load.

The worst case for the chords, I would expect, is uniform snow load. The top chord, I assume is relying on the 4-2x6 laminations butting together where they meet. The nailing pattern should be adequate to prevent laminations from buckling individually. Otherwise, they do not participate in load carrying.

The bottom chord is spliced only at the middle? That seems odd because 35' long 6x6 are not readily available. In any case, the details of the bottom chord splice and the end connections would be important features to check.

The grade of lumber, particularly in the top chord is almost impossible to check properly. Hopefully, the top chord stress is not so high as to require a high grade of lumber.


BA

 

[msquared48]

BARetired: Depending on the age of the trusses, the 35 foot length of the 6X6's is possible. Difficult, but possible even today with a special order. But that is $$$$$$$$$$$$$.

I would ask:

1. Are any of the members rough cut?

2. Were the trusses manufactured on-site?

3. How and where are the trusses "broken"?

4. How long have they been "broken"?

5. How many trusses are there and how many are broken. Where arfe they in the roof structure - in the middle of the diapgragm, the ends, etc?

Mike McCann
MMC Engineering

 

[diggingdeeper]

Mike:

I would ask:

1. Are any of the members rough cut?
actual chord size is 5 5/16 x 5 5/16 and midspan splice is all there is. beautiful tight grained lumber. i'll see if i can upload a photo.
2. Were the trusses manufactured on-site?
this is unknown at present. but spacing is 20' 5 1/2" which makes for a perfect 'off-the-shelf' uncut rafter (in hangers), which leads me in the direction of thinking it was designed by a contractor and so built, which further leads me to think it may be an onsite fabrication.
3. How and where are the trusses "broken"?
2 cases.
case A: vertical perp to axis break in bottom chord at second panel point away from heel. one has mech unit over 1st TChord PPoint from heel and one does not. there is no sign of water near these trusses.
case B: split parallel to axis from heel to (situation one) through (situation two) the 1st PPoint at BChord from heel. both have same mech unit similar to case A above over 1st TChord PPoint from heel. one has evident signs of water on underside of sheathing around top chord.
4. How long have they been "broken"?
this is unknown at present.
5. How many trusses are there and how many are broken.
see above
Where arfe they in the roof structure - in the middle of the diapgragm, the ends, etc?
picture rectangular structure, grids A,B,C,D,E across the top each 70'-6" and grids 1,2,3,4,5,6,7,8,9,10,11,12,13,14 down the side each 20' 5'1/2", trusses 70'6" long spaced 20' 5 1/2".
the breaks generally occur close to grid B generally towards the center of the building (grid 5,6,7 and 9)

 

[BAretired]

A photograph or two wouldn't hurt.

BA

 

[diggingdeeper]

truss pic

 

[diggingdeeper]

truss pic 2 of 4

 

[diggingdeeper]

truss pic 3 of 4

 

[diggingdeeper]

truss pic 4 of 4

 

[diggingdeeper]

one more pic showing the typical panel point configuration which was a main part of my original query. thank you all for your interest and support.

 

[msquared48]

The first two look like tension failures of the bottom chord - the first at a knot??? and the second at a vertical hole for a tension rod -

The third and fourth look like a bending/tension failures of the top and bottom chords.

Did you notice any rot in the failure areas?

The fifth picture shows the joint with added holes beyond the connection holes for the sprinkler system.

I wonder if these holes were considered in the original design. Any record of what the live load was that these trusses were originally designed for? Have there been any serious snow overload conditions in recent years?

Doesn't the owner have insurance on the building? If so, why haven't the repairs been done yet, assuming the problem has been there a while?

Mike McCann
MMC Engineering

 

[hokie66]

As to the web connections, the single bolt through a web member would not have a lot of capacity unless there is a split ring or shear plate in the web. But failure of the bolted connections does not seem to be what happened.

Mike has correctly identified the failures as tension failure of the bottom chord, and bending failures of both the top and bottom chord. This tells me that gross overload has occurred, and the trusses were probably not designed at all, just built.

 

[BAretired]

It is pretty clear that the main problem with these trusses (so far) is in the bottom chord. The bottom chord of every truss must be considered incapable of developing the tension required to carry the roof load.

One possibility is to add a steel tension member each side of each and every bottom chord. It would need to be welded or bolted to the end plates.

In addition to that, it would be wise to review the connections of the web members to ensure the structure is competent to carry the snow load.

There appears to be no other viable alternative.

BA

 

[msquared48]

From the looks of these trusses, I honestly do not know what is holding the roof up except the limited arch action of the top chord of the truss. Hopefully this building is not occupied at present. Even unoccupied though, it is still a structural hazard.

If it is not now, this structure should at least be red-tagged until the fixes are made or temporary shoring is installed.

Mike McCann
MMC Engineering

 

[Sawbux]

It appears to me from the photos that there is a valley at the end of the trusses that were broken at the heel joint. Has this ponded water or snow, causing the over load in the trusses? The uneven distribution of load at the end will cause shear/ tension across grain at the heel joint. Wood is very weak in that property!

The upper chord is laminated horizontally, which gives little bending resistance for the secondary stresses between the panel points. This may cause the upper chord to act as supported beam rather than an arch, which puts load into the web members and bending in the bottom chord.

The two photos of the breaks near mid-span shows signs of "brash" failure, this is indicated by the almost straight across grain break. This may occur when wood is exposed to heat for a long period of time and becomes brittle. Stick a knife point into the wood and pull out a splinter. If the splinter comes out long- wood is OK. If it comes out in short pieces- not good.

One photo shows what appears to be paint in the break - must have been there a long time. Another thing that can cause a break like this is known as "compression wood" and may be caused by wind damage to the standing tree or damage during the felling. The wood cells are crushed in compression in an almost straight line and are very weak in tension. This property is very difficult to detect visually when the timbers were graded.

Bowstring trusses are always problematic. I have repaired some with steel straps or cables. These seem to have enough defects both in design and history that they may be beyond repair.