Your answer [from a previous column concerning the use of collar ties to fix a sagging roof] was that collar ties would not do anything vertically to correct the deflection and you advised the man about a Plan A, use of truss uprights off the tension cord, or Plan B, installation of sister rafters. The thought that occurs to me is that there was no mention about how to achieve the correction to the original rafters.
Are you expecting the use of posts and jacks to push out the deflection prior to attaching companion boards?
If so, where is the best starting point or sequence for that: center, ends, series of jacks, one at a time? It seems to me this is a tricky staging proposition to achieve the right set-up prior to executing.
Really appreciate your thoughts here.
Before answering your excellent questions, a little clarification is in order. My original column on collar ties ran in November, 2004 and was fairly short and sweet. I upgraded and expanded it considerably for publication in my book, “Cracks, Sags and Dimwits,” Lulu press, www.lulu.com, a year later. If you haven’t read the book version, you’re only getting about half of the story. And, that original column — unlike my book — did not include the case study of the old church building, which had the same affliction you’re now wrangling with and what we did to correct it.
The gist is that collar ties will reduce the stress in a rafter slightly, but in doing so shift forces around to other elements and connections that can’t take it — and so are not a recommended fix. The recommended solution is to add vertical or diagonal support to the rafters by installing posts (aka “kickers”) and connecting them to bearing walls below.
You are correct in your observation that I did not say how to get back the sag. More on that in a minute.
First a little background on bent wood. When a load — any load — is placed on a beam, joist or rafter, said member will deflect (sag). If the load is small, the deflection may not be noticeable. If the load is large, the deflection increases proportionately.
If a large load is applied for a short term, say from temporary construction materials or from snow, and everything else about the member is okay, the deflection will be elastic. That is, like a rubber band, the wood will spring back to its original shape when the load is removed.
I italicized “and everything else about the member is okay” because most badly sagging members fail this basic tenet. Consider: there is a span for which any beam’s own self-weight will cause it to fail. Say, for example, we have a 2x4 that is 50-feet long. Have a guy on each end pick it up and it will break in the middle under its own weight.
The point is that any beam, joist or rafter can be overspanned even before snow or construction workers or any other live load is applied.
Most of us in the construction industry aren’t so dimwitted as to try to span a 2x4 50 feet. However, we may try to span one 15 feet and then subject it to a Rocky Mountain winter.
In the old days, BC, (Before ConstructionCalc software), builders sized beams, joists and rafters using the “back of the thumb” method, which went something like this:
FRAMER FRANK (from his teetering position atop the walls of a house under construction): “Hey, Joe, I’m framing up this roof system. What do we have a lot of in our boneyard?”
LABORER JOE: “Looks like we got a big batch of 2x4s left over from the Yellapiddle job.”
FRAMER FRANK: “Think they’ll span this 30-foot roof system?”
LABORER JOE (sighting down the length of his arm at the back of his skyward-pointing thumb): “Yeah… yeah, I think so. That’d be 15-foot per side. We’ve spanned 32-foot before with 2x4s and never had no prollem.”
FRAMER FRANK: “Okay, pack 'em up here. You know, Joe, with your knowledge of shears, moments and deflections, you should have been an engineer.”
LABORER JOE: “An engineer! Frank, you’re killing me. That’s the most idiotic thing you’ve ever said. I been building this-a-way for 50 years and ain’t never needed no engineer yet.”
Overstressing can be caused by overspanning, as in the above example, or it can be from overloading. Or it can be from both. Some common causes of overloading follow:
- From a freakishly heavy storm.
- From a normal storm, but the snow is extraordinarily wet. Water content of snow can vary wildly.
- From a normal snowstorm followed by a cold rain, which the snow soaks up like a sponge.
- Snow dumping from an upper roof onto a lower one.
- From a strong wind that piles it up against a parapet, chimney or leeward side of a roof.
- From ice damming at the low edge of an eave, which can back up piles of snow on the roof, and in extreme cases, can curl under the eave, hanging there, adding tremendous weight.
- Rain on a flat roof with a plugged downspout. (This caused the collapse of the roof on my favorite grocery in Sacramento, Calif. when I was in college there. Interesting story, that. The roof creaked and groaned so loudly under the extreme load that everybody inside was scared out of their wits and fled the building before it caved in. That’s one nice thing about wood — it generally warns you before rupturing.)
- A party in which too many people crowd too small a space. Chintzy decks in older apartment buildings are notorious for this.
- Builders stacking too much material in too small an area on a roof or floor, especially if the decking isn’t installed yet to distribute the load and to keep the rafters or joists from flopping over.
- Any cantilevered (overhanging) joist, rafter or beam with an overhang of three-feet or more. This is generally not a case of rupture (breakage) but rather a permanent deflection (dip) at the far end of the cantilever.
Wood creeps, meaning that when it is overloaded or overspanned for a long time (years, usually) it will deflect permanently. You can’t get this “plastic” deformation back. Well, let me qualify that. You could get it back using heat, moisture and constant pressure, like wood benders do. However, in most rafter and joist situations that is not an option — which is why I didn’t mention it in my original collar tie column.
Now, to answer the specific question: no, I’m not expecting the use of jacks or posts to recapture any permanent (plastic) deformation. You might undeflect the members slightly with jacks or come-alongs, but the instant you removed them, sproing, right back they’d pop. So you’re left to make your repairs as best you can dealing with permanently sagged wood. If you can’t live with that, there isn’t much you can do but tear the whole shootinmatch out and start over. Or, hire a professional wood bender.
Infomercial Postscript. Every builder, remodeler, and designer should own a beam, rafter, joist sizing computer program for the very reasons alluded to above. As a testament, after five years of beating my contractor-remodeler brother over the head to give ConstructionCalc ProBeam a try, he finally took the plunge. “Timultuous,” he gushed a few days later, “Timultuous, you really should market and sell this program! It gives me 20 alternatives — in any material I want! And, when I show the printout to my clients, they are soooo impressed.”
I didn’t bother to remind him that I have been marketing and selling it for over five years now. Brother.
Tim Garrison of ConstructionCalc.com, is a professional engineer, author, and software producer for the building industry. Check out his new book, "Cracks, Sags, and Dimwits — Lessons To Build On," available at www.lulu.com, Amazon and Barnes and Noble.
Send e-mail to email@example.com. Tim reads every one.
This column cannot be reprinted without permission from the author.
The views expressed in this article represent the personal views, statements and opinions of the author and do not necessarily represent the views, statements, opinions or policies of the National Association of Home Builders. NAHB does not necessarily endorse any of the views expressed by the author and NAHB is not responsible for any direct or indirect consequences arising out of the views expressed in this article.