Floor Joist Notched for Ledger 3

[SteelPE]

I have a project where I have to calculate the capacity of an existing wood framed floor (should be simple enough). I made a visit to the site yesterday and took a few photos and took some measurements. I noticed that all of the floor joists were notched in order to rest on a flush framed wood beam (see photo).

The notches on these joists is over the allowable notch located in the building code (which I believe to be 10% of the overall depth of the wood member). NDS does give an allowable shear equation for notches of fv = [3V/2bdn][d/dn]. Using this equation I am calculating an allowable end reaction of 154# which gives me next to nothing in capacity for the floor (17.5 psf).

I know this method of construction was used in the 1970's (my parents house has this exact detail). Question is, do I ignore the detail based upon it's the historical use of the detail or do I report the capacity of the floor as 17.5 psf? I am leaning towards the 17.5 psf as it's really not my problem to deal with at this time.

 

[phamENG]

A notch for the 2x ledger is permitted by code so long as the notch meets the bearing notch requirements from R502.8 (D/4). So a notch for a 2x2 ledger (cue the shivers up my spine) would be "acceptable" for any 2x8 or larger joist.

The 2x2 is explicitly permitted in the IRC:

Joists framing into the side of a wood girder shall be supported by approved framing anchors or on ledger strips not less than nominal 2 inches by 2 inches (51 mm by 51 mm). (Emphasis mine)

While it may be "permitted" it's still a terrible idea and has no place in an engineered design.

 

[lexpatrie]

phamENG - i already said a 2x2 ledger was permitted in the current code...

yes, a notch at bearing is permitted in the IRC, it's not in the same section as the R502.6.2 2x2 ledger is my point. The notch on side seems to visually exceed this limit (say 50% of the depth)

It's existing construction. Nobody said it was engineered. It's "implicit" I suppose that somebody is asking the engineer to vouch for the live load of the floor. I don't see any research on the capacity of the 2x2 ledger strip connection and it's not inside the prescriptive provisions anyway. If somebody desperately wanted to stamp this detail, they would need to get over the NDS hurdles we describe here. The alternative is in-situ testing to demonstrate it's safe.

I think we are getting distracted with the IRC, here. The OP mentioned a "workshop". That doesn't sound like the residential code. (Prescriptive framing in the IBC may have similar provisions).

 

[Eng16080]

lexpatrie - I suppose you're right that the code doesn't imply the joist is notched when a ledger strip is used. I was picturing how I've normally seen the detail constructed, that being with a notch.

What I should have said is:
When a 2x2 ledger strip is used, it's not uncommon for the supported joist to be notched. It's also very likely that the ledger strip is attached below the neutral axis of the supporting beam. Either of those conditions alone is bad in my opinion making this a terrible detail.

 

[phamENG]

lexpatrie - sorry, I guess I misunderstood:

That idea that the code "implies" a notched joist isn't founded in the code itself, I don't have the IRC commentary, and there's no figure or language in the code itself permitting it.

You're allowed to bear joists on a 2x2 ledger, and you're allowed to notch joists at bearing. Since there's no prohibition on notching joists at a 2x2 ledger, it follows that it's a permissible condition. The code isn't designed to lay out every possible permutation of construction configurations.

Though, if you want one, you can look at the IBC, conventional wood framed construction section (the prescriptive design rules for wood in small, non-residential structures) where these do occur side-by-side:

2308.4.2.3Framing details.

Joists shall be supported laterally at the ends and at each support by solid blocking except where the ends of the joists are nailed to a header, band or rim joist or to an adjoining stud or by other means. Solid blocking shall be not less than 2 inches (51 mm) in thickness and the full depth of the joist. Joist framing from opposite sides of a beam, girder or partition shall be lapped not less than 3 inches (76 mm) or the opposing joists shall be tied together in an approved manner. Joists framing into the side of a wood girder shall be supported by framing anchors or on ledger strips not less than 2 inches by 2 inches (51 mm by 51 mm).
2308.4.2.4Notches and holes.

Notches on the ends of joists shall not exceed one-fourth the joist depth. Notches in the top or bottom of joists shall not exceed one-sixth the depth and shall not be located in the middle third of the span. Holes bored in joists shall not be within 2 inches (51 mm) of the top or bottom of the joist and the diameter of any such hole shall not exceed one-third the depth of the joist.

I usually go to the IRC first since that's the only place I fall back on prescriptive stuff (residential contractors and engineered designs rarely mix well in my area, so I give them "the usual" whenever it's reasonably justifiable), but a lot of this stuff is in the IBC.

The issue, as I see it, with this particular subject, is that the NDS doesn't do anything beyond saying it's bad. They don't go into detail attempting to show why it's bad. So when the code writers see that in the NDS, but get push back from industry groups and then states get lobbied by industry groups (I'm looking at you, NAHB) to let them keep doin' it the way their grand-daddies taught them, the NDS hasn't really explained why. Combine that with the fact that the Truss Plate Institute does give values for design (because every truss has tension perpendicular to the grain at an truss plate that doesn't run right up to the edge of the member), and it's pretty easy to see why getting rid of some of these details hasn't happened yet.

 

[lexpatrie]

I concede your point that there are engineering concerns with the connection. In the engineering sense you'd have to lean into the NDS commentary that it's "lightly loaded" or provide mechanical reinforcement somehow. I believe I've seen them notched as well, but I'm not positive. It wouldn't surprise me to see them notched, but as the prescriptive code, there are limits to be respected.

 

[gte447f]

Not to be harsh, but some of you folks need to get out from behind an FEA model on your computer screens and into the real world a bit more and experience real live construction of real live buildings. Notched and un-notched wood floor joists bearing on 2x2 and 2x4 ledgers fastened with nails at the bottom edge of wood floor girders has been standard construction in millions of homes for decades and decades throughout the United States. (Sidenote: the 2x2 instead of the 2x4 is actually standard and preferred historically, specifically, at least for one reason, because it allows for an acceptably small notch in the joist.) There is nothing inherently wrong with this type of construction. It has and is performing adequately in millions of real life structures in service. For those worried about attaching the ledger below the neutral axis of the girder, keep this in mind: it is called "conventional light-frame wood construction" for a reason.

Now, if the detail is constructed deficiently, such as excessively notching the joist or using too few fasteners in the ledger, etc., then of course the load capacity of the detail could be limited. In this case, the OP calculated the reduced capacity of the joists due to the excessive notch and concluded that the allowable floor capacity is 17.5 psf. Well if you deduct about 10 psf for self-weight/dead load, then that leaves you with about 7.5 psf of live load capacity. That isn't much at all and implies that the structure is severely limited in it's usefulness and probably unsafe under a lot of imaginable service conditions. Well, job well done Mister OP. Inform your client and be done with it. No need to go down a rabbit hole crusading about 2x2 ledger strips, and neutral axes, and such.

 

[phamENG]

gte - I'd agree with you if it weren't for the countless houses I've been called to evaluate with a failed, notched ledger condition. And many more that, while the whole connection hadn't failed, it had deformed enough that the ledger clearly wasn't what was supporting the joists anymore...

No FEA/computer screens about it for me. I realize my sample population suffers severely from selection bias (I don't get called for nor do I volunteer to go observe perfectly good connections), but I've seen enough to know that I don't trust anyone with a saw to make this connection in a truly reliable way.

 

[gte447f]

phamENG, I have read enough of your posts on this forum to know that you are out there accumulating a wealth of practical engineering knowledge from first hand experience assessing and solving real world problems in structures. Kudos to you, and please keep sharing your experiences here on the forum.

I too have seen problems with this specific detail, but I have seen many more instances of it performing serviceably for the life of a structure. One of the most frequent problems is overcutting a notch (poor workmanship) or using a 2x4 ledger instead of a 2x2 ledger, which necessitates an excessive notch in the joist, unless the girder is deeper than the joist. I also often see too few nails in the ledger, and sometimes nails causing splitting in a 2x2 ledger. The list goes on. However, as a counter point, I bet I am as likely as not to see a metal joist hanger installed incorrectly in some fashion or other, which boggles the mind, because it should be so damn simple a child could do it right.

 

[lexpatrie]

Where in this thread do you see a computer screen or a finite element analysis? Seriously. Your tone, furthermore.

Further "if the detail is constructed deficiently" I don't think there is any IF here. Nobody has said "this is fine" and neither do you. Don't present it as some kind of hypothetical so you can more effectively attack it.

 

[phamENG]

gte - thank you, I appreciate the compliment.

I agree that most of it comes from inadequacies in execution. But correcting joist hangers is often easier than correcting something like what we see in the OP. Not always, but so far I feel comfortable saying usually. And, anymore, I only see inadequate executions. Whether it's this sort of notch, cutting stair stringers, or anything remotely similar...the guy running the saw invariably runs the max depth of the saw up to the mark and leaves a huge gash in 'remaining' material. As a hobby woodworker/carpenter I'm embarrassed for the pros.

lexpatrie - the if came about when we started discussing the code provisions and their intrinsic merit. That's what usually happens on good threads....a question comes up, we hopefully help the OP find an answer, and we go off discussing some detail(s) in minutiae to suss out some deeper understanding of their origins, intentions, and theoretical validity. It's not all about the OP. We're here to have fun and learn from each other.

I find it best to ignore any perception of tone. It's just text based communication. Somebody could be happy as a clam but I think they sound angry. Or they could be angry but are trying and failing to hide it. Who knows and who cares? If I go looking for a fight, I'll find one. If I go looking for an engaging discussion with some really bright engineeers, I'll find that, too.

 

[gte447f]

lexpatrie, It's just humor. Done poorly, apparently. My tone was inferred, not implied. This is just a discussion. There is no attacking going on.

Hypothetical? I commended the OP for rationally analyzing the as-built conditions of this specific connection in-situ.

 

[KootK]

I find it best to ignore any perception of tone. It's just text based communication. Somebody could be happy as a clam but I think they sound angry. Or they could be angry but are trying and failing to hide it. Who knows and who cares? If I go looking for a fight, I'll find one. If I go looking for an engaging discussion with some really bright engineeers, I'll find that, too.

I'd get that tattooed on my chest and show up here shirtless if this was a real pub rather than the internet.

I feel that humor has an important role an this forum, like most. You know, to keep us from all just sounding like a bunch of dull-ass robots all of the time.

All good humor has to "ride the line" however. Everything that is truly funny has to skirt the boundaries of what is not. And that line can be hard to see through the fog of cyberspace. As such, I feel that anyone bold enough to attempt humor online deserves a little sympathy when it doesn't land as intended.

 

[phamENG]

I'd get that tattooed on my chest

What's with you and JAE and your chest tattoos?

(For context: thread784-512101; scroll down to 12 Oct)

 

[Eng16080]

This isn't a real pub?

 

[bones206]

Given it's Friday and I'm at peak procrastination, I decided to try to reverse engineer the capacity of the prescriptive IBC ledger strip. If we set aside the splitting and look at the roll-over/pullout capacity, I figure the following gives a good range of capacity:

Given: (3) 0.162 X 3.5" nails per joist, W = 26"/in of embedment, Z = 120"/nail (assume all adjustment factors = 1.0, G = 0.42)

Capacity based on withdrawal (lower bound):

Wtotal = (3)(26 #/in)(2" embed) = 156#​

Pmax = (0.5)(156#) = 78#/joist​

Capacity based on shear (upper bound?):

Pmax = (3)(120#/nail) = 360#/joist​

Maybe an appropriate limit would be somewhere in the middle, say 100#? That would align with the NDS commentary definition of a "light load".

 

[gte447f]

bones206, nice contribution. Thanks for the effort.

I assume you are taking the eccentricity to P as the full 1.5" dimension of a 2x2 ledger based on the conservative assumption that all of the reaction is delivered to the corner of the ledger due to deflection/rotation of the joist? I would be comfortable assuming uniform distribution of the reaction on the ledger, so eccentricity would be 0.75 inch, and therefore capacity based on pull out would be 156 lbs instead of 78#.

In my neck of the woods, I would also assume G=0.55 for southern pine instead of 0.42 for SPF. So W=50#/" instead of 26#/". Therefore, capacity based on pull out would be 300 lbs instead of 156# or 78#.

Still, point taken, that the connection doesn't have a lot of theoretical capacity.

 

[phamENG]

Another thing to add: a lot of the issues I see with these in older houses where there are not obvious workmanship errors are in crawl spaces. Moisture levels in crawl spaces around here are high, and elevated moisture wrecks nail pullout pretty quick.

 

[Eng16080]

bones206, I think you can use NDS Eq. (12.4-2) to combine the lateral and withdrawal capacities. Doing that, I get 62 lb per nail or 187 lb total. Still not great.

 

[bones206]

I visiting my brother-in-law this summer and his deck had one of these connections. It looked exactly like my sketch, maybe worse. Definitely has a progressive failure aspect to it, where the eccentricity slowly grows as the ledger rotates.

 

[WinelandV]

This isn't a real pub?

Nah, the real pub is over in the "Member Moderated Groups" - Pat's Pub.

Koot should be able to get you in, if you ask nicely - KootK being an eng-tips super-user and all.

Please note that is a "v" (as in Violin) not a "y".