Existing Wood Trusses are Failing... 3

[StructuralJoe]

I have any interesting project...

It is a 1-story plaza in Hernando County, FL

It has wood roof trusses spanning ~50' between load-bearing CMU walls

It was built in 1985 and from discussions with the owner the roof originally was constructed of plywood sheathing with Clay Roof Tiles...

Upon inspection the roof has 1 member on each truss that has buckled, typical on every truss

The owner had originally contracted a "handyman" to provide "repairs" although no engineer specified the repairs.

Members where added randomly as shown in attached photos.

I have since been contracted and modeled the roof truss with all applied loads to current code...

It appears that the trusses have no horizontal bracing

My questions:

The only member that fails (on-site) does not fail upon analysis... but other members do, why?

Bad wood grade on every truss?
Load transfer?
Lack of proper permanent bracing?
Current wind loads to excessive?

I have modeled the wood grades to be the minimum that will work for the minimum loads applied and not fail any members... but I don't feel comfortable assuming that the trusses where originally designed to not have any lateral bracing (rat-runs)

Any thoughts on this process...

Any suggestions or recommendations?

 

[msquared48]

Bruce:

I try to keep my mind open to all possible reasons for any failure until the truth comes to the fore. Closing ones mind limits the truth. Sorry for the post. I won't make any more.

Mike McCann
MMC Engineering

 

[BAretired]

Mike,

Keeping your mind open is great! Don't ever close it. Perhaps Canadian lumber was at the root of the problem. Who knows? Certainly not me!

Please don't be "sorry for the post". I admit that I was a little annoyed at you, but I want your continued input because, over the last few months, I have come to respect and value your opinion on a number of issues. Please do not drop out of this discussion.

I would invite you and all members of this forum to review the engineering aspects of this problem. If you do not agree that Vertical #2 was incapable of carrying the load, even with good lumber, please show me your calculations.



BA

 

[msquared48]

I have run no calculations, but base my comments on what could be.

The EOR will need to sort through the local requirements and actual geometry of the truss to determine what was the cause. I was merely giving another alternative reason, possibly an additional one, and a follow up post to my original one with the actual information. I was not discounting your information, just adding to it.

That being said, I feel it is funny in a roof structure for only two trusses of many to fail if it is only a design problem. I still suspect that defective materials played a major part. It could also have been a local overload which happens here with snow driftring. But that depends on the location.

Sorry for the post. I got a little PO'd too.

Mike McCann
MMC Engineering

 

[BAretired]

Thanks, Mike for staying in the discussion. Faulty lumber could be part of the problem, I agree. But there are other possibilities.

You said...

That being said, I feel it is funny in a roof structure for only two trusses of many to fail if it is only a design problem.

It is not only a design problem. Here is what the OP said:

Upon inspection the roof has 1 member on each truss that has buckled, typical on every truss

This suggests to me that one member, namely Vertical #2 had a problem. The problem was that it was too slender to safely sustain the load. Check it out. It is clearly incapable of carrying the load.


BA

 

[msquared48]

OK. Misread it. Probably is a design problem then.

Mike McCann
MMC Engineering

 

[mark1234]

When I see anomoly's like this it has always been to lack of truss bracing.

I hear stories how home owners witnessed loud banging in their attic during heavy winds... when inspected, always a lack of bracing was discovered.

If this building was erected in 1985 the following links may add some insight as to the excessively bowed webs and failed members.

http://en.wikipedia.org/wiki/Storm_of_the_Century_(1993)

http://www.tboblogs.com/index.php/news/story/no-name-storm/

MDJ

http://www.windspeedbyzip.com

http://www.groundsnowbyzip.com

 

[msquared48]

I have to admit that when comparing the computer model with the photo, they do not seem to match.

1. I seem what looks like a ridge beam at the roof, above the truss, and beyond the break point.

2. I notice two five member joints in the bottom chord in the model, and see one in the foreground of the photo. I assume this is the right most joint of the sketch.

3. I see the added vertical, beyond the first five member joint. I see the broken diagonal after the added vertical.

Am I missing something?

Mike McCann
MMC Engineering

 

[BAretired]

Mike,

1. What looks like a ridge beam is the top of Vertical #1 (numbering from left to right on the elevation).

2. The five member joint you see is directly under the ridge (which I believe has a height of 10.9')

3. The added vertical, actually two per truss is a repair immediately adjacent to Vertical #2. The additional vertical shown on the sketch is behind the photographer and does not appear at all.

That is the way I see it.


BA

 

[StructuralJoe]

Group, thank you everyone for all of your input.

Just an FYI, I have posted three times to this and I haven't seen any go thru... not sure what is wrong but...

There is some confusion (my fault entirely) on the support locations, please review attached multi-page PDF...

In addition to clarify, it appears that permanent lateral bracing should have been applied and wasn't.

It also appears that possibly the roof wasn't originally design to support clay tile (i have placed 15psf SDL in addition to self weight).

Is it possible that the SSBCI in or before 1985 allowed for a great LL reduction that currently permitted? I used 20psf unreduced (i may be able to reduce the front slope LL due to slope angle)

In addition it may be possible that the design wind loads in 1984 were much less than currently required...

Also I believe a "stress increase" was permitted back then when multiple load cases were applied (ie DL + LL + WL)

All of this said and done I intend to first reinforce the members that don't meet current code.

Secondly I intend to add quite-a-bit of lateral reinforcing so that none of the member fail under any of the current loading combinations.

Also there has been some question as to the orientation and the members that have failed. 1 member in each truss has completely failed... another member in that truss (opposite)fails much higher... I wonder if the longer legth member bowed but didn't break whereas the shorter member (when load was subsequently dumped into it) cracked much easier... neither of which is here nor there... i intend to bring these trusses up to meet current code regardless...

Thanks especially to Mike and Bruce, I greatly appreciate both of your inputs! And also everyone else involved.

Thanks, Joe

 

[BAretired]

Joe,

Thanks for the response. Mike and I can go and have a beer now.

The last page on your PDF shows that the truss is higher than I had thought. The left support is at Vertical #2 where I had assumed it to be at Vertical #1. This makes the stress in Vertical #2 considerably more serious than I had calculated.

I assume that the ratios on your last page are F(calc)/F(allowable). If that is right, you have some pretty impressive ratios.

Thanks again, Joe.


BA

 

[Jeff2009]

StructuralJoe,

When I looked your second picture, I can tell without any doubt that the vertical green member will buckle without bracing. The blue one is a tension member at vertical loads therefore it can't buckle. I don't think the failure has anything to do with the lateral wind loads.

 

[SlideRuleEra]

I agree with BAretired that the most likely cause of the problem is compression members that are too slender. Of course bracing will resolve that, and is certainly the way the truss system should have been originally designed/installed. However... just because bracing "should" have been used does not necessary mean that adding bracing is only way to address the situation as it exists.

IMHO, rather than retrofitted bracing, requiring accurate placement to work properly, a less sophisticated approach is probably more cost effective:

Use glue and deck screws to "sister" a new, near full length member (2x4, probably) to any currently unbraced compression member that is too slender (greater that 6' 3" unbraced length). Doing this would increase resistance to buckling by merely doubling the width of the compression member. Members that are buckled could have a pair added, one on each side.

Advantages are:
1. Truss members needing reinforcement can be determined in advance and the lumber cut and drilled (for screws) before it is taken into the attic. (Assuming that lumber approx. 12' long can be taken into the attic).

2. The work is repetitive, chance for installation error is reduced. Inspection and acceptance of the work are also simple.

3. In the photo the attic is already "crowded". Once installed, there is still reasonable (future) access to all parts of the attic. - better than with a bracing system installed.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[msquared48]

StructuralJoe;

Not time for a beer yet...

Sorry to belabor the point here, but I still have a couple of concerns with the computer model:

1. Are all the joints really pinned? Normally the top and bottom chord are continuous over some joints, but not at the splice points.

2. You have shown roller supports at both supports in edach direction. I do not agree with this. The vertical direction should be set so no vertical translation can occur. The horizontal direction should be set so no translation can occur at one joint, but can at the other.

See what happens to the deflections and member moments/shears after these adjustments are made.

Mike McCann
MMC Engineering

 

[StructuralJoe]

Mike, correct on the support, that was one of many iterations to determine what is happening... great catch thou

To further clarify:

I have modeled supports as pin-pin and pin-roller
I have modeled all points pin-pin
I have modeled all points fix-fix
I have modeled points at continuous Chords to be fixed and splice locations to be pinned - along with all web members pinned (I did record splice points on site)
I have modeled with lateral bracing
I have modeled without lateral bracing
I have modeled with some members laterally braced and other members "built-up"
I have modeled with 75% of LL and WL
I have modeled with 5psf TC SDL + SW instead of 15psf (the non-clay tile option) and 75% LL & WL

To note SlideRuleEra option that is what I intend, but under DL+LL+WL I will still need some additional "rat-runs" to strength members that will undergo compression...

I intend to use all #2 southern pine 2x4 so size and length shouldn't be too cumbersome... and for this space rat-runs shouldn't be to difficult IMHO...

Thanks again to everyone posts... I may upload my intended repairs to hear additional input...

Does anyone know of a particular "GLUE" to specify for attic spaces here is Florida? I plan to develop the forces using screws but glue would be a good way for me to CYA...

 

[BAretired]

I wondered about the supports too. At each support there is a horizontal and vertical roller. That would be the equivalent of a hinge at each end. Based on that, there should be no stress in the bottom chord at all, so I concluded that I must be misinterpreting the boundary conditions.


BA

 

[StructuralJoe]

I applied a line load of 5psf x 2' = 10plf to the bottom chord to acount for ceiling, MEP etc...

I have seen this on current truss shop drawings...

Technically it is a has resistance at both end due to the coff. of friction on the truss bearing of the CMU and also the tie-down in shear... that's why i did it both ways...

Thanks, Joe

 

[SlideRuleEra]

I have specified "Liquid Nails" products (or equal) for attic truss repairs (in South Carolina) - there are several such as this one:

http://www.liquidnails.com/products/product.jsp?productId=36#msdsTech

Can be applied in temperatures up to 100 deg F. with a maximum service temperature of 140 deg F. That should be high enough for a ventilated attic.

You will have a large area of contact - properly applied the glue will be better than screws. Will also help to compensate if the existing members are of inferior wood.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[StructuralJoe]

The comment about supports brought up another thought entirely...

I ran pinned-pinned and pinned-roller... and compared the results...

Then after the discussion I just tried roller-pinned... the stresses change entirely for the entire model... other members fail that never before did...

very odd... any thoughts... in the dozens of truss shop drawings I have review I have never seen a roller pin always a pin-roller...

 

[BAretired]

Using a pin at the left end of a truss and a roller at the right should give precisely the same member forces throughout as a roller at left end and a pin at right.


BA

 

[kslee1000]

One thing I don't under stand - how could the program let you have Roller-Roller support without flagging (global instability). I wouldn't trust that output.