Truss Metal Plate Failure

[idecharlotte]

This past winter Charlotte had a whopper snow storm with almost 6 whole inches of accumulation. I am now looking at my second roof truss failure as a result. Both situations are identical. Churches with scissor trusses spanning 46 feet, both built in the mid-90s. The truss company who did both is out of business. The failure was at the peak metal plate where the vertical chord gave way from the top chords. Failure mode was teeth pull out. Both failures occurred over the winter starting with one truss quickly followed be several others. Walls kicked out a few inches on each side and trusses dropped at least 6" before shoring was installed.
I modeled the truss on RISA and determined the vertical force under design load is 3500 lbs. and under the actual snow load conditions that caused it to fail about 2000 lbs. I did tons of internet research found the ICC report regarding the truss mending plate allowable design capacity. The allowable teeth stress should be about 192psi in SYP#2. I measured the contact area of about 17.5 sq. in which yields about 3360 lb capacity. I figured close enough since I didn't figure any C factors.
As best as I can tell the truss plate connection met the requirements of both the current ASCI/TPI-2007 and the applicable TPI-85 specifications. Upon further reading some engineers have questioned the TPI testing procedures saying they don't account for localized bending stresses and humidity/ temperature stresses that occur in real world conditions. Yet the entire industry follows them. I'm going back to investigate the first truss that failed a little more closely and see if there was some fabrication error but so far as best as I can tell, the actual capacity of the truss plates did not meet the tested/ design values.
Anyone been through this and can help?

http://www.idecharlotte.com

 

[JAE]

Did you get the actual snow configuration at the time of failure?

Sometimes you get unbalanced snow on one side of the roof that creates worse conditions than the design snow.

Also - snow loading is based on general estimates of density. Some wet snow can weigh a lot more than anticipated.
You probably don't have any recourse to find data on the actual snowfall density but perhaps some insight as to snow drifts on
windward/leeward sides of the roof might help.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[idecharlotte]

That snow load really doesn't matter since the design Live load is always higher in Charlotte. So even if I was off on my estimated 7 to 10 PCF density of snow, it is nowhere near the 20psf live load the truss is supposed to handle. Besides, there really wasn't a drift issue where the failure occurred.

http://www.idecharlotte.com

 

[JedClampett]

I know nothing about metal truss connectors. But if the truss plates don't meet the tested allowable capacity, is there an environmental or other external issue that would cause them to not meet this capacity? Were the plate teeth fully engaged? Was the timber in good condition locally, with no rot?
As far as JAE's comment, the roof couldn't take that live load it was designed for. So pinpointing the actual load can be important. Are there other roofs with the same design you could look at? Maybe they are giving early signs of failure that you could observe.

 

[idecharlotte]

Timber was in good condition, plates were fully engaged. No signs of roof leaks or rot.

http://www.idecharlotte.com

 

[woodman88]

As, per the OP, this seems to be a wood not a metal plate failure. I would check the other end of the failed vertical, to see if there is any separation of the metal plate to the vertical wood member.
In any case, as this appears to be a wood failure I would not look at any problems of the plate values. But rather at possible wood or manufacturing problems.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[dcarr82775]

I think the 90's was a little late for it, but was the wood treated with a fire preservative? The old type really hurt wood strength.

 

[idecharlotte]

I reinspected the truss plates again. They definitely failed by teeth pull-out. The wood was in good condition, dry, with no signs of roof leaks. The roof had proper ventilation with ridge and soffit vents. Several other plates on some of the diagonal members have begun to experience plate peeling at the corners, perhaps due to the new stresses as a result of the failure. The wood was Southern Pine #2 with moisture content less than 5%. The only think that I can assume is that seasonal environmental conditions of cold and hot, expansion/ contraction weakened the connection over time, and the snow storm was the nail in the coffin so to speak.

http://www.idecharlotte.com