OWSJ Failure

[connect2]

We provide structural consulting to several school boards in our area. Recently we recieved a memo wrt to a school in Washingtonville N.Y. were a portion of the roof suddenly collapsed. Fortunately this occured on August 1, 2004 and there were no injuries. The collapse has been attributed to an 'SJ' type joist that had retained water infiltrating thru the roof assembly and subsequent corrosion had weakened the owsj to the point of collapse ... without warning. The memo from the School board in question can be reviewed at

http://www.emsc.nysed.gov/facplan

Question is; Has anybody else had similiar experience with this type of joist?

 

[JAE]

A "U" shaped channel top chord that holds water leaking from the roof - hmm. I suppose that its possible that the roof could leak for that long, rust out the top chord and never, never have a drop of water get down past the joist and be revealed through staining of a plaster ceiling.

Sounds like a Macomber joist.

We have had joists rust in the past but not with that type of top chord where the "evidence" is hidden until too late.

Inspection might include drilling small holes through the top chord to see what the section looks like - lots of visual inspection, too.

 

[CESSNA1]

DEAR CONNECT2: I have had no direct experience with this but here is my 2 cents.

The trusses are obviously corroded. This implies that they were exposed to water and I am sure standing water and other corrosive processes.

There was no evidence of leakage. I suspect during the inspections, if there were any, the inspectors were looking for water marks on the interior ceiling. Could the watermarks, if any, have been painted over?

I suspect condensation. That area of the county is subject to low temperature and high humidity.

Moisture in the interior air could have infiltrated through the ceiling and condensed on the steel trusses. The cold weather would cause condensation/freezing on the steel.

It appears that the top chord is more heavily corroded than the rest of the truss. This could be caused by standing water, possibly with salt in the air, stress corrosion, galvanis corrosion, and other forms of corrosion.

I suggest a metallurgical analysis of the trusses to determine what corrosive processes were at work.

Where and how was the ceiling insulated and was there a moisture barrier, if so, where?

Regards
Dave

 

[jheidt2543]

It would seem to me that it would take a significant time for water standing in the top chord of a joist to corrode it to the point of collapse, whatever the source of the water. You don't mention the age of the building.

However, I was involved in an instance where the metal roof deck corroded on a building less than five years old. The soucre was the wet roof insulation, which was made from an expanded foam material that when it became wet disolved to form an acid solution. It turn out that the manufacturer knew of the problem, but would only correct the problem once it was found (they were not proactive about it!).

The insulation was a very popular one because it had a high R value per inch of thickness and therefore was used extensively as an energy cost saver. You might check that as a source. As I remember it, the name was something like Rx Foam and made in Canada.

 

[connect2]

Just to be clear we are not investigating this failure. We are mearly responding to a directive from the government to inspect our schools for a similiar situation(s). They have issued this directive based on the conclusions of a NY State Structural Engineer that did investigate and report. This can be read on the site referenced. In my case i had never heard of this problem and was a bit surprised ... so have asked the question if others have experienced similiar problems.
To Jheidt; they don't mention the age of this particular portion of the building.
I find their description of the joist as an 'SJ' confusing as well. The 'SJ' descriptor is an old one. Since the founding of the SJI, joists covered by their standard were called 'SJ-Series', on or about 1959 'S-Series' replaced 'SJ-Series', on or about 1961 'J-Series' replaced 'S-Series', etc., etc., But there is no reference to a particular propietory joist type.

 

[JAE]

connect2 - are you then hiring firms to "do" this inspection for you?

 

[connect2]

JAE - no where not hiring anybody. Based on what happened in New York State The Ministry of Education in our Province has issued a letter requiring all school boards in Ontario to review all their schools for similiar joists and or situations as referenced at the NY state Ed. web site in my opening comment. We are Structural Consultants to one of the Boards and so have been asked to investigate. Its quite a large task as there are a lot of buildings with multiple additions extending over a period from 1910 to about 1985. 90% of them contain OWSJ of one description or another. So just trying to find out if anybody has had similiar experience to that referenced on the website.

 

[JAE]

We've never had a joist failure for a rusted condition like that - pretty rare. We have reviewed joists for damage from fires, overload from snow, damage from impacts, etc.

I would think that the investigation would have to be based on an actual visual review of current conditions. By going through all the old structural plans, you may get an idea of where these joists were specifically called for, but in reality, the various joist types used in past years were substituted on an "equal" basis so you probably don't know if you have the channel chord type unless you specifically see it in each building.

It sounds like the channel-joist-holding-water concept is what is the focus here. I just don't believe that a traditional double angle joist would have this problem as long as the leaks were found and fixed in an appropriate time.

 

[JedClampett]

This rusty joist solution seems too neatly wrapped up for me. I might believe one joist rusted through without any water damage visible below. But to take down a whole section like that doesn't ring true to me.
I'd be interested in the interval between the failure and the last big rainstorm. I'd also like to know the condition of the roof drains.
It might be just a corrosion problem. It also might be a soggy roof that failed at its weakest spot, a rusted joist. Or possibly, a ponding failure due to clogged roof drains.

 

[Lutfi]

I must say that I have never encountered such a situation in my 21 years of working as SE.

It sounds to me that someone is rushing to blame the channel for holding water that lead to sever corrosion. While this may hold true (fully or partially) I have few points:

1. Obviously corrosion is the culprit from the few pictures that I have seen posted on the web site. Then one must as what caused the corrosion. I like to raise the following possibilities:

A. Water leaks from the roof. It seems to me that it will take a long time for this to reach the condition the joists are in. Of course, I do not know when the school was built.

B. Dissimilar metals. Yes, we all know that create galvanic action. I would check into the type of steel and its composition, look into the roof deck and its metal composition as well and see if there are elements that would create a galvanic action. I recently learned that when sitting a Stainless steel frame on HDG steel frame, this would create a galvanic battery if moisture is present. Keep in mind moisture comes in many forms such as liquid, steam, and vapor.

C. As an engineer who designed many roof systems utilizing steel bar joists, I do not ever recall that I had to do special coating on them just because the roof may leak! I have asked for special coating on joists because they were exposed to corrosive environment.

D. Why was the roof membrane leaking? How long was it leaking before it was repaired? Why no one inspected the structure underneath for water damage?

E. My experience with plaster that it will show stains from water leaks. Were those leaks visible? Did someone investigate where the moisture was coming from? Were those leaks covered up with paint as it was suggested above?

F. Historically speaking, what was the space below that roof used for? The use of the pace could have contributed to the moisture migration to the space above the ceiling. I have done finite element software analysis with moisture transport using various construction materials. Vapor will transport though concrete, block, paint, dry wall, plaster, and stucco. To reduce the vapor/moisture transport, you must use a vapor barrier having a very low perm rating (less than 1 and close to zero). This phenomenon is impacted by the climate conditions, temperature and humidity, through out the year. The basic premise is vapor/moisture can enter spaces as long as the drying rate is higher you will be on. Drying can be accomplished by heating and or air conditioning. I would be very interested in seeing the composition of the roofing system.

G. Thermally speaking, insulation is rendered ineffective once it is wet and holds moisture. If it gets wet, can be a source of corrosion; slow at work.

I would be very interested in the final report on this failure.

 

[connect2]

Agree with all. Never in my 25 years have I run across this. I don't know the exact situation so its hard to comment about the conclusions given in the NY State report. However away we go. Thanks all for confirming my initial reaction. Getting to old for visual inspections ... always seem to have the wrong glasses on hahaha.

 

[connect2]

Update,
Out of 28 schools inspected three with corrosion, 2 insignificant but none the less corrosion, one with significant corrosion. Two related to condensation (insignificant), one related to water penetration through roof membranes, through poor deck welds, into the joist, in areas where large roof top units and associated curbs and supports located (significant corrosion), noting also the top chords of the OWSJ in this this instance are cold rolled shapes. humph!

 

[JAE]

Thanks for the update.