Building A Loft Floor

[MountainDon]

So the other day I was looking into a couple of things in the IRC. I noticed something that I had either seen before and forgotten or simply breezed over previously because I was not looking into balloon framing.  

Right there, plain as day in section 6 (walls), is a drawing that illustrates a prescriptive method of connecting the joists for a second floor to the wall studs. They use a 1x4 let into the studs, not a 2x as so many of us owner-builders have done. That leaves the stud with more outward bending strength. The floor joists rest on that 1x4 AND are nailed to the wall stud as well.

The nailing schedule that accompanies that states 3 - 16D nails to secure the ledger strip to the stud. It is not clear to me what nails to use for nailing the joist to the stud, but 10D would make sense, I think?

The figure below is a small section of Figure R602.3(1)



The floor joists have to mate with a wall stud, so that would preclude using 24 inch stud spacing and 16" joist spacing.

Comments?

[PEG688]

I'd use 5/4 stock instead of 3/4.

[archimedes]

Interesting find.

Wasn't there a discussion quite awhile back where someone said that by removing 1.5" from the 2X6 wall stud that you were violating code by removing more than 25% of the wall stud ( 1.5" is 27% of 5.5") ?    If that was true, than maybe this is the solution.

[PEG688]

Yes see  the  : Letting in ledgers : (one of the stickies), thread. Of course on that project the Engineer approved the plans and we had some LVL in play. The plans where drawn by a Professor of Architecture who also was the home owner, along with the Engineer stamp of the design. 

[MountainDon]

Archimedes, yes there is a limit in that same section about notching studs no more than 25% of the actual depth.
A 2x54 (3.5 inch) would be 0.875 (7/8) inches and a 2x6 (5.5 inch) would be 1.375 (1 3/8)

PEG I was wondering. The way they put it, "1 in x 4 in"  and not 1x4, made me wonder if they were talking real measurements?  But that would put it over the 25% on a 2x4 stud. And I have no desire to revisit manhattan42.  

[John Raabe]

Yes, if you are getting issues with a 2x ledger, a 1x is sufficient if you can space the joists and rafters to align and not use the ledger as a short span beam. If 5/4 material is available then that covers both bases. You could also nail a 2x6 backer to the 1x ledger and toenail into the studs where the spans don't align.

Note that when you put in a ledger you are not leaving an open hole, which is the argument for limiting the cut to 25% of the stud (and this was totally arbitrary in its initial determination but is now quoted as scripture). When you have a solid ledger nailed into a well cut stud you have replaced most of the bearing capacity. In bending, the nailed off loft joist - which is above the notch - has stiffened the stud far more than it would be if it had remained uncut and unattached.

My thinking with regard to the 2x6 notch in a 2x6 x 10' wall was that the 1.5" cut left more (4") than the thickness of the 2x4 wall that would otherwise meet prescriptive code. See this from an engineering thread: "2x4 vs 2x6 studs the 2000 IRC Table R602.3.1 shows the prescriptive allowable length of wood stud walls based on the number of floors supported.  In essence it says that a 10 ft tall 2x4 stud wall 16" o.c. can support a floor and a roof for their "model" structure.  2x6 walls 24" o.c. can be prescriptive up to 14 ft without engineering."

[PEG688]

That rule is a $hitty piece of rule of thumb IMO YMMV.  But like a lot of things people  engineers  mostly don't look at the "whole" deal. They look at each piece almost as a stand only item.

And the reason for that is poor building practices over the years , another reason why Simpson Strong Tie Company  has been so successful at getting their products used.  So a Engineer can look at a chart and say IF I make them do "this" / use this clip , hanger , strap etc " It don't matter what the carpenter on-site does , I've reached a certain strength / requirement / point load , what have you , no matter what the builder / carpenter does it's all "covered" by the factors on the chart.     

  This let in ledger rule of thumb or  prescriptive code  doesn't consider any thing but a stud that's notched , doesn't factor in the strength of  the  floor joist resting above that's nail thru the stud , the sheathing that's nailed on to the stud , the siding that's again nailed over the stud etc.  The interior wall finishes dry wall , maybe 1x6 T&G etc.

   

[Don_P]

MD, all dimensions given in the code are nominal... a 1x4 is referring to a 3/4" x 3-1/2" board. Do look at the top of that picture, it applies to a roof with no thrust. If there is thrust the notch is on the tension edge. For myself I'd rather see a 2x notch and ledger on the outside, compression edge. This would give the required minimum 1-1/2" bearing and would then be reinforced by the sheathing. Chances of getting that by an inspector would probably not be good though.

From the other nailing schedule callouts I would think 3-10d (3"x.128) nails would satisfy the joist to stud connection... again if this is not a thrust resisting element. If you are resisting thrust here, and most of you are, this becomes an engineer's turf, much closer to a heeljoint.

As an aside I usually buy 12d's at 3.25"x.131 to get a little more nail in the connection. also notice that when the code calls for a 16d they are very often calling for a hand drive 16 common which is larger than a sinker and alot larger than a gun 16. Even if you are not in code country, if it is calling for 16 commons and you are using a gun, pull the trigger at least half again as many times as they are calling for... when they call out 16 commons there is usually a good reason.

If Manhatten shows up on the job, 2 passes down the length with a power planer might satisfy him... then again  . The 5/4 is a good option. I've been complimented a couple of times for letting in 2x6 ledgers for grab bars and cabinets, you'll run into all kinds.

With regards to the eng-tips thread.  I've also done "the math" and with allowable safety factors a 2x4 wall 8' tall is indeed slightly overstressed. We recognize that the assembly as a unit works and it is allowed. So this is a case where the engineering would be more conservative than the code. One gentleman brought up the interaction of bending and axial loads and another mentioned that bending loads need to be looked at. True.
There has been mention of the ledger filling and replacing the notch, it doesn't. The materials are assembled at best at 19% moisture, they will go down to 12% or below in service, there will be a gap, load goes to stiffness, the remaining wood, the notch is gone. You're also on the tension side, kneewalls on high posted capes have been leaning out due to bending since the style was introduced, there is a compressive and a bending load. Timberframers have used the same thinking with drop in joists into notches in beams, testing of full sized material showed that the logic of the joist filling the notch doesn't hold up. In fact it caused a rethinking of that whole area. It is worse than even the pessimistic professor predicted, you can't even claim the material in the beam under the notch. The video was neat though, nothing like an exploding glulam bouncing off the technichians booth as he dived for cover   .

Where did we come up with the 25% rule... a stud is a #3 piece of lumber with #1 edges. A #1 edge is allowed a 25% width defect, same as the notch, these both give the same strength range. This comes from breaking samples. I've been at the machine when notched 2x4's were broken. Make the notch across from another 25% defect or an area of maximum slope of grain and that safety is blown. This is why I'm cautioning about horizontal thrust in the notch area. We are applying a rule outside of what it was intended to be about and using sketchy logic to make ourselves feel we're covered. It might be fine, it might not. I find I have to start a critical thought process by being honest about what I do and don't know. Our minds try to close and cover rather than open and learn.

Simpson has done a good job of making it easy to use their products, they've done an excellent job of testing and promoting their line. We, the public, have done a mighty poor job of doing the same for alternative methods, we're lazy. We have, and we pay for, cutting edge engineering labs, professors and grad students. we should be lobbying them en masse to work for us. I've been having a discussion with a building inspector about this. He's as sick of Simpson as I am and realizes there are many areas where I could nail in a pressure block and develop the needed strength but it is sure easier to look up the numbers in a simpson chart rather than trusting my math, and by the time I hire an engineer, the hangers look damn cheap. The solution is prescriptive tables from testing done at the college. So far we've been ignored. Why? Cause if one guy walks in singing a bar of Alices Restaurant they're gonna think he's crazy.

[John Raabe]

Good point PEG. Looking at everything in isolation and over-designing at each stage leads to expensive and unthinking robotic building.

Glenn and I are working up information on the Ken Kern CBRI light duty plunger pile concrete floor. You would not want to try this in a code area, but it has a long history of spanning 3' every direction using very thin concrete shells and small plunger piers. This floor has been tested in India and in Glenn's underground home, that is IN REALITY NOT IN CODES.

I don't think that this floor is necessarily a great idea for most people, but it is fascinating what can be done when you're guided not be codes but by experimentation and testing to see how little will actually work. Then you can choose, like Glenn did, to double that to be conservative.  

(Admittedly it is easier to test a residential floor than a wall or a roof in a hurricane.)

[archimedes]

If you use a 1X or 5/4 instead of 2X for the ledger,  how do you get past  the 1.5" bearing requirement for the floor joists?  You solve the notching problem,  but don't you just create a different code problem?

[John Raabe]

Interesting!

[MountainDon]

fourth line from the word except...

[John Raabe]

There you go. Thanks Don 

[archimedes]

 

[glenn kangiser]

Good point PEG. Looking at everything in isolation and over-designing at each stage leads to expensive and unthinking robotic building.

Glenn and I are working up information on the Ken Kern CBRI light duty plunger pile concrete floor. You would not want to try this in a code area, but it has a long history of spanning 3' every direction using very thin concrete shells and small plunger piers. This floor has been tested in India and in Glenn's underground home, that is IN REALITY NOT IN CODES.

I don't think that this floor is necessarily a great idea for most people, but it is fascinating what can be done when you're guided not be codes but by experimentation and testing to see how little will actually work. Then you can choose, like Glenn did, to double that to be conservative. 

(Admittedly it is easier to test a residential floor than a wall or a roof in a hurricane.)

Codes are many times written to sell materials and increase the tax base.  Sometimes it has nothing to do with durability or necessary strength of materials.  The CBRI floor is a great example of using materials in an environmentally responsible fashion without sacrificing safety or durability. 

Harry51 is going to use one of my design modifications of that floor with a concrete admix to make it waterproof for an outdoor deck in an area that really catches rough weather.  Hot enough in summer that polyurethane will not last a year and rain and snow blows straight sideways here on the mountain in the winter.  He will be covering exterior grade plywood.  Harry has seen my CBRI floors and has little fear of having any problem with it.

[Don_P]

Sometimes it has nothing to do with durability or necessary strength of materials.

I don't disagree, but you had better know when.

The codes related to strength of materials have this clause;
"When a building of otherwise conventional construction contains structural elements that exceed the limits of this code those elements shall be designed in accordance with accepted engineering practice. The extent of such design need only demonstrate compliance of nonconventional elements with other applicable provisions and shall be compatible with the performance of the conventional framed system."

The intent is simple, show that you have designed a safe building.