Deck Lateral Load Research Results!

[Glenn]

There are three code modifications proposed for the 2015 IRC related to the deck lateral load anchor detail and provisions. (2012 IRC: R507.2.3) (proposals: RB260, RB262, RB263)

Recent testing by Washington State University reveals answers regarding the ability of band joists and ledgers to resist lateral loads. I’ve put together a video on the subject.

This 13-minute video covers the past, present and future of deck lateral loads. It describes how the provisions got in the code, how they affect the industry, and the results from the recent studies. (as best a 13-minute video can do, so see the links below)

[video=youtube;nP6QzyxZgAw]

If you’d like to read the entire research report yourself, use the links below. I highly encourage a subscription to Wood Design Focus and reading these articles yourself.

Wood Design Focus: Forest Products Society

Editorial of articles: http://www.forestprod.org/buy_publications/resources/july2013/Wood%20Design%20Focus%2023,%20Number%202%20Bender. pdf

Information about other deck proposals for the 2015 IRC: www.nadra.org/code

If you will be attending the public comment hearings in Atlantic City this October, I would love to hear from you.

Thank you

 

[Mark K]

The original concern was not necessarily the lag bolts in the load path but was that when the rim joist was loaded in a perpendicular direction that it would possibly pull out. Still it appears that the testing was focused on forces applied parallel to the rim joist.

The ability of wood construction to resist forces is highly dependent on quality of the materials and the workmanship. With these tests it is very easy to significantly increase the capacity by the right choice of materials and good control over the workmanship The problem is then when the revised detail gets into the IRC it needs to work over a broader selection of materials and workmanship. Thus unless the testing program tries to address these issues the primary advantage of the testing is in understanding the magnitude of the forces and the mechanisms by which the forces are resolved. The point is that such tests should be viewed with a certain skepticism.

While the testing might help to refine the code provision I believe that there is a real advantage to a direct connection to resist perpendicular forces.

The validity of a lag bolt only detail is much more questionable when you are screwing into blocking between joists as opposed to fastening to a one piece continuous joist.

My sense is that this has become an emotional issue for some deck manufacturers. I have trouble believing the added cost is significant and a good contractor will pass the cost along to his customer. The one valid concern has to do with when a deck is added to an existing building and there is a need to disturb an existing ceiling. In these instances a custom engineered solution will likely minimize or eliminate the need to disturb the existing ceiling. But then many contractors have a thing about engineers since this may mean that they loose a little control.

 

[Glenn]

Thank you, Mark K, for your review of this information and commentary. I would certainly encourage the reading the articles if they were ever available to you. You would find them interesting.

The live load testing with human subjects revealed that the most force could be generated by movement parallel to the ledger. This was due to the nature of the deck's ability to deform. With that knowledge, they tested the connections in the same direction. The resultant force of a parallel force at mid span of the joist is a force perpendicular to the band (rim) joist. The band joist definitely was loaded in withdrawal.

I agree about the quality of the work. The researchers are very descriptive about how the 4 foot wide mock floor assembly was built. They built it according to the IRC with a #2 material. Yes, you never know how something else was built, but we assume it was done to code. We assume this every day with every construction and product. This is why there are safety factors. The discussion and research of lateral loads is not over, but current knowledge does discredit the lateral load provisions in the code today. Those were put in without any research. A single concentrated connection of 1500 lbs for the construction, design and research industries to blindly follow should be over. It's not a sensible or necessary connection. The research revealed many other issues with the connection. Being on one side of the joists, one of the two was put into compression while under load. They also contributed to additional joist damage. Holding two joists of any random deck is not going to stop a failure if the ledger fails (i.e. rot)

I am not aware of any cases of deck failure that included a supported rim joist being pulled from a house. Rotten rim joist from poor or lacking flashing?....YES. Rim joist from a cantilever floor?....YES. I've studied deck failure cases with a sickness, and have found NONE that were of a normal band joist getting pulled. This fear is fabricated, as far as I've been shown. The code definitely needs work regarding decks, including education regarding what we already know. But this anchor is not the solution for the deck failures you read about every day in the news.

Yes. In new construction it's easy to throw money and hardware at the project (but you have to ask the guy holding the money). Sure...installation is not a problem. Paying for something required by the government that does not have a proved return...that's a problem. The minimum standard should remain as such. It should be trustworthy.

For the decks that fulfill the dreams of your neighbors, they aren't new homes and installation can be very costly. Both below the floor and above the floor must be disturbed. Homeowners are as reluctant to their installation as builders are. Quality builders are then passed for the handyman that will do it on the weekend (no permit)...and do it cheap. I am very in touch with the decking industry, and I respect the real affect this provision has. Our decisions in code development and enforcement affect those where the rubber meets the road, and that should matter to us. Authority exists to benefit those being regulated.

Emotional issues are when inattentive parents want guard spacing tightened up to crib standards after an unfortunate or anticipated accident.

Regulatory issues are when an industry is burdened with unnecessary regulation to satisfy convenient enforcement of a provision based on an emotional issue.

An emotional issue is what got the lateral load detail in the IRC in 46 days. A regulatory issue is what should remove it.

 

[DRP]

I agree with a good bit of what you've said Glenn. I do have an observation that I've brought up within this group before. I know what the APA calls for, but back in the real world... Because T&G plywood doesn't cover 48", very often the rim is connected by some toenails to the sill, nails in withdrawal to the joist ends, and a 2 or 3" wide strip of floor sheathing to connect the top of the rim to the joists. The sole plate of the walls hopefully makes a few more connections of joist to rim.

Are you requesting reports from insurance companies on deck claims? I can only relate that I don't believe I've ever seen an accurate news report regarding a deck failure and was wondering if you all are checking with insurors rather than just monitoring news feeds.

 

[RJJ]

I will review this info and respond! I will be in AC this year not sure of what dates yet.

 

[Glenn]

Are you requesting reports from insurance companies on deck claims? I can only relate that I don't believe I've ever seen an accurate news report regarding a deck failure and was wondering if you all are checking with insurors rather than just monitoring news feeds.

I have not made any requests to insurance companies. That is a good idea. Thank you. My research of collapses starts with news stories, and goes from there. The internet is powerful and folks these days love to share their stuff. Here is a good example.

This collapse was described to me as a band joist failure by a zealous advocate for the lateral anchor. I was told the ledger was bolted to the band joist and pulled it from the house. Upon research of the address and region, I came across this personal photo gallery. Hmmmm... This is a classic example of deck failure from utterly poor construction being used and construed by those seeking to keep this proprietary anchor provisions in the IRC.

Have a look at the photos. Decks like this were built by the tens of thousands over the last two decades. We will see many deck collapses in the coming decade, and we need to address the problem, not blindly point to the anchor detail as the answer.

This failure captures the most prevalent causes for deck collapse.

1) Only nails in the ledger (photo proof)

2) No flashing and the band joist rotted into garden compost.

3) Connection to an unsupported band joist

4) Connection to brick veneer is another common problem, but not in these photos.

LINK:

Aston - Weir Road Deck Collapse album | Robb Ware Photography

.

 

[Glenn]

The Professional Deck Builder has gained permission to reprint the Forest Products Society "paid" article. This is from well-recognized and respected researchers.

Now everyone can read the same thing I paid to read last year when I made my video (posted above). Please do read it. Don't take my word for it. It is very very eye opening.

http://www.deckmagazine.com/Images/MeasuringLateralLoads_tcm122-2103359.pdf

 

[Glenn]

I hijacked someone else's thread, so I'm posting my old video here. This video is more comprehensive than the one in the original post. Watch this one instead. No, I did not get this out of the code. It is still there and it's still what it is...

 

[Sifu]

Way back when, while a contractor and inspector in NC we had some prescriptive code for the lateral resistance on decks. Like the ICC, it was a performance requirement, and left the method up to the designer/contractor and inspection approval. It had prescriptive "options", which I really found helpful. They were good to fall back on and in most cases easy to comply with. When they were contrary to what was desired, they had the option to get an engineer or come up with an alternate method that the inspector would accept. However, just like has been found in the ICC, the "option" is often interpreted to be required. Personally, I was open minded about it then, just as I am now.

That doesn't mean that plenty of noise wasn't made about the requirements, so I had many conversations and interactions with the NC code council on the subject and what I found was pretty interesting, and somewhat contradictory to the current mind-set.

The first point was that the NC code was more concerned with the lateral forces acting on the deck at the outer rim, parallel to the ledger. The concern was largely with the live loads and wind loads at that location causing a swaying motion. In my discussions, the current deck tie option was just being introduced by the ICC and NC was considering those reasons and options. I think that the the attention paid to the outer band location was in fact transferring the lateral force at the outer edge to a tension force on the ledger, putting the ledger in withdrawal like you mention, and thus requiring more attention to the loads at that point. According the NC engineers, the prescriptive bolts/lags were more than sufficient to resist this, so though the study cited is a good one, but the results aren't surprising.

The second point is that we are all focused on the bolts at the ledger to band connection, which I don't think many argue with. But I do question the band connection to the house diaphragm, and maybe that is a reason for the tension tie introduction. It's easy enough to verify the sheathing is nailed to the band in a new build or a more modern house. But in a replacement situation, how can we be sure the band is adequate to withstand the withdrawal force? So is it a concern that we are relying on a band joist that is nailed with a couple of nails into the end grain of the joists or no joist at all? Or with a few nails through the sheathing that never gets inspected because there is a wall plate on top of it? Are we assuming there is a band joist at all? I can't seem to access the study you cite so I will ask this in case you know. Did the testing of the lateral force at mid span include simultaneous live load? And was the ledger/band connection to a house diaphragm considered, or just assumed to be a non-issue?

I am NOT an advocate for the tension tie, though if I am right, I do see some cause for concern if the band that the ledger is bolted to can be ripped off along with the deck. There are certainly other ways to shore up that connection than the tension tie.

Maybe I am off base here, after all I am not an engineer but I would like to see more attention paid to lateral resistance where it seems like it would be more likely, at the outer edge. The DCA deck hand-out includes a little of this.

Anyway, great video, I enjoy and appreciate your expertise on decks.

 

[Glenn]

Hey Sifu. I wonder if you watched the first video of this thread from years ago, or the one I just posted #8 (directly above your post). That video has more information and addresses some of your questions.

 

[Sifu]

#8, I will go back and check the first one.

 

[steveray]

I do see some cause for concern if the band that the ledger is bolted to can be ripped off along with the deck.

I also believe this is a concern with I joist floor systems where the rim is held on with very few 8d end grain nails...Which is another reason this came into code (I believe)

 

[Sifu]

Just watched the first. It does address the question at the end of the video. So my question about the load path to the house diaphragm was examined and found to be ok. I just looked at the 18 IRC, and the issue is addressed by adding another prescriptive "option", which is an improvement. I assume that is in some part thanks to you. So thanks.