So, what is it? Is it now a ritual for the modern American first-time timber framer to do some type of modification of Jack Sobon’s “Garden Toolshed” as their first project? It seems like it to me. And I’m proud to be engaging in this tradition myself.

So, there are a number of modifications I’m doing, but this question refers to the loft joists. I will be putting in floor joists for the loft. They will be dropped in to two of the three 8x8 crossbeams. It’s important to note that I am changing the original shed dimensions from 12x16 to 14x20, so these joists will have a 10’ span. Instead of having the loft end in a straight line across where the crossbeam ends, I want to cantilever the loft over that beam and give it an edge that will have some curve in it. At the furthest point, I’d be looking for a 2’ cantilever. I am milling the timbers myself, so I will probably use 5x7s or 5x8s for the joists.

I am thinking about allowing part of the joists to run over this middle crossbeam, giving me my cantilever (maybe a 5x3?). This means the loft floor wont be flush with the top of the crossbeam, but that sounds like a fine compromise for me (although if Ive ever seen this in real life, I can at all remember how it looked to me). I was talking about this to a professional timber framer, and he told me a lot of professionals will only “let in” their joists ¾” to 1”. He recommended I do it 1 ¼”. I’m confused. In all the literature I’ve seen for drop in floor joists, the pockets are always much larger than that. In Sobon’s plans, his first floor joists are 4x6s dropped into a 4x4x3 pocket (many of you are probably familiar with this). Also, I found this picture from another post on this forum:

[img]http://www.flickr.com/photos/danielcha/5466223638/[/img]

What is this “let in” measurement this builder was referring to? I’ve been thinking of it as the pocket depth, (which in the above picture is 5” or less on an 8” D. Or is it the tenon depth (3” or more in the above picture)? What do more experienced framers (than me) have to say about this? Is 1 ¼” enough depth for the joist pocket?

So, if I use 5x7 joists, and assuming this 1 ¼” “let in” measurement he’s talking about is the pocket depth, I am thinking of removing 1 ¼” from the whole top of that middle crossbeam, and run 3” up over it? aka – only 1 ¾” of the joist is actually above the top of the crossbeam, but there’s 3” of intact joist allowing me my cantilever. I think that’s structurally sound for both the cantilever (which won’t be supporting much weight) and the crossbeam. If that’s too much to remove for the 8x8 crossbeam, I can size up just this middle crossbeam to an 8x10.

On the other crossbeam, I would make a 1¼x5 x3 pocket.

Any thoughts people? All are certainly appreciated.

I don't see any issue with a 4'(or slightly less) knee wall on this width of a building. the wider the building the less knee wall height.

Rest assured, Jack Sobon has just stuck joist on top of the tie beam with no joinery, I have as well, I think is is a reasonable solution and cuts out a lot of unnecessary joinery and fiddle fuddling.

On the "let in" vs "drop in" I suspect the "let in" is the 3" into the side of the timber and the "drop in" is in the depth of the timber.

I would not use a drop in pocket on the ends of the joist. Instead use a soffit tenon on one end of the joist, they will plug into the end tie beam and flush off to the top of it as well. I find a simple mortice easier to cut than drop in pockets. It also eliminates your question on 1" or 1-1/4" pockets. At the center bent with the cantilevered joist use no joinery just set them on top of the tie. To facilitate this the center tie needs to be the thickness of the the joist lower than the end tie beam. Also make the other end tie beam the same height as the center tie to allow joist to be just set on in the future when you or someone else decides to floor the whole thing over.

Using soffit tenons and mortices will leave valuable wood in the tie beam making is stronger, as well. In the situation you have outlined this is an easy assembly as well, which is something to always consider. And I would make all the ties 8x10's.

Divide the brace leg by .7071 to get the length of the brace. See.... 36"divided by .7071=50.9121" . Adjust the brace longer to fall below the tie beam joinery, using Sobon's directions for the 36" brace only swapping out the 36" for a longer leg and brace length. Longer braces are better anyway. This is for the long wall, in the bents you can stick to the 36" lengths. This means you have to keep things straight when laying out the work, but that is part of the fun in constructing this type of building.

I've certainly used 1-1/4 drop-in pockets for joists and purlins, but always with (hidden) long screws to make sure they stay where they belong. CB.

Hey,
Thanks for the quick and detailed reply. That’s really interesting you would suggest a soffit tenon. I hadn’t thought about that at all. You say you find it easier to cut the simple mortise easier than the drop in pockets, but the drop in pockets seem so straightforward and easy to me. Plus, if you’re only removing 1¼” of wood for the pocket depth (I think this is what he was referring to), then that’s not much removed, no? But, maybe by keeping the wood you leave with a soffit tenon makes a stronger crossbeam . . . ?

Why would you size all the crossbeams to 8x10s? Necessary? Precautionary?

So, maybe 4’ braces on the long wall, and 3’ braces in the bents? Again, necessary or precautionary? You make the longer brace layout sound simple enough, though. I wonder if its necessary or not because if the 3’8” kneewall is OK, it would be nice to keep it that high as that makes the space more accessible. If I were to use a 4’ brace here instead of a 3’, then to avoid the same problem of the mortises being in the same area, I would have to bring the kneewall down by 5” or 6”. Not the biggest deal (still leaving me with a 3’2” kneewall).

A factor here in terms of bracing is that my insulation will be straw bales with an sand/lime based plaster. These bales have bracing power, especially in a one-story building. If the 3’ brace already works here, I wonder if the addition of the bales might already prove more than enough bracing power?

Im gonna read up on soffit tenons. Thanks again.

I guess an argument for the soffit tenon is that it brings the floor flush with the top of the crossbeam, but it would only do that on one end. The middle would still have a joist (or 3” of it) running over it.

Why only 3" running over the top of the tie? Drop the tie beam the thickness of the joist. I just like longer braces, if the tie is dropped 3'-8" than you would have room to place the 3' brace above the tie.

8x10's just playing it safe, with out running numbers, a few unknowns, loads, use of 2nd floor, etc. Also for repetitiveness, all joints would be the same. You could cut daps into the top surface of the middle tie, as locaters for the canted joist and to bring the tie up a tad if head room becomes an issue.

I think it depends on how the drop in pockets are being cut. Router, could be why 1" pockets are used in the larger shops. A 1/4" more is not that much more. What species are being used? Does it have a high shrinkage rate? How about quality and spiral grain? Boxed heart or quarter sawn? All factors which will impact what happens to that 1-1/4" deep pocket. It has been said, the flooring will hold up the joist!

If using only hand tools, no electricity, a mortice is much easier. If you look at the amount of wood removed the mortice, with a haunch on the shoulder of the joist included, removes less wood. The mortice removes .26 board feet while the drop in removes .30 board feet.

Mass building code requires 1.5" of bearing for wooden joists, and I wouldn't do anything less than that for a pocket depth with green timber. Soffit tenons remove much less wood on the tension side of the beam, and thus are much stronger. See Jack's "Build a Classic Timber Frame House" for a comparison. I think drop-ins were done historically because they are much easier to cut and assemble, especially without a crane and come-a-longs, but they upsized the carrying timber appropriately to account for the reduced capacity. The way to calculate the bending in a notched beam is to, in brief, take the cross-sectional area in the middle of the beam with the wood removed, find the new centroid based on that arrangement of voids, and then figure the new moment of inertia. Not running the numbers, I think an 8x10 would be fine with 1.5"x 3"notches on two sides, 14' span on the tie, 10' on the joists (one side), and run as much of the joists above the tie as practical. Tim's solution of running them over the top is the strongest. Depends on the grade and species of timber, of course. By the way, the joists in Jack's book are dropped into the sill which is fully supported, perhaps.
I remember the Guild conference (RPI?) when in the span of one year we saw the slide show entries move from drop-ins joists and purlins to soffit tenon assemblies pre-assembled on the ground and flown into place. People figured out it was stronger and safer.

Will - is the 1.5" for typical 2x joists?

1.5" x 3" or 5" or 6" for the surface area seems excesssive for a typical floor load. For green pine I have often used 1" for 4x6 or 4x8 secured with either straps or panel screws.

I don't use metal hangers enough to remember how deep they are, 1.5"?

For the most part I am not concerned with how far into the timber the joist is set, due to the load factor. I find it more important, especially when using green timber that it is set in far enough that after drying it is then still bearing on enough wood. To ensure that I prefer to play it safe, at the minimum 1.5" is used and usually 2" to 2.5". It is nice to know that in some areas I perform above the code standard.

On numerous occasions I am asked what I do about green timber and how it is dealt with if it is used. This is one of many solutions to working with green timber.

The OP mentioned he will be sawing his own timber and most likely it will be used in a green state. If you were having dried doug fir shipped in from the west coast, 1" or 1.25 may be just fine. Not in green timber. I feel very reassured in saying this, just the other day, while doing some remodeling on a frame I cut 5 years ago I have had a first hand look at what dried timber looks like, it can move a bunch.

In the MA code 1.5" is minimum bearing for any wood member resting on wood or metal, be it a beam or a 2x joist. 3" bearing required if wood is resting on masonry. Joist hangers are 1.5" deep. Although, as Mike says, the loading can be handled with less area, I think it's to accommodate other factors such as shrinkage, poor workmanship, etc.

A long time ago, and using a book on timber framing as a reference we have probably all seen, I did a 1" deep joist seat for a green red oak frame. Visiting the house a few years ago I found one of the joists completely out of the housing on one end and being held up by the nails from the subfloor above. Luckily it was a short trimmer but not so lucky to be right in front of the bathroom door. Decorative oak stirrups glued and screwed on the carrying beam covered it up. Pine would shrink much less.

Agreed Will.

I learned valuable lessons on my first frame for my father's screen porch that I cut in the summer of 2000. White oak for exterior use, and fresh as could be.

Nothing as dramatic or structural as you mention - but lots of little things where you can see into housings, or where a timber is now hanging on a peg in its joint...

In white pine I would never go less than 1" for something like this, but also make the decision based on loading, use, etc.

Thanks so much Tim, Mike, and Will for the very helpful responses.

I'm keen on Will's recommendation of using 1.5"x3" notches on two sides in an 8x10 (not 8x8) tiebeam, and then running as much of the joist as practical above the tie on the middle bent. I had been thinking about doing a pocket depth of 1.25" or 1.5", but extending it all the way thru the 8x10 tie as opposed to just 3"(see #2 in picture below). Will's recommendation (#1 in picture below) would remove less wood on the top side of the tie, and I would run 3" over the middle-bent tie. Im really liking that. It's clear a soffit tenon on one side would be stronger, but doing it this way keeps the location of both ties at the same height, and allows me the adzed reduction on both ends. Do I have to worry about the strength of the joist where the "notch" is made that transitions from the joist tenon to the 3" tail running over the tie?

[img]http://www.flickr.com/photos/danielcha/5483014336/sizes/l/[/img]

Thanks again guys.

It really depends on the intended load to be placed on the joists. Storing regular attic stuff is one thing, placing car engines up there is another....
You need to not only design for your plans but be aware of what future plans or storage will be put up there.

Cutting across the full width of the tie beam will mean that the cross section is reduced greatly and may need to be verified by running the numbers again to see if it will be ok for your load. Normally, I believe, they (engineers and code officials) don't like to see cuts completely across the top of the beams.

If you cut a 1.5" notch all the way across the beam then instead of an 8x10 you now have an 8x8-1/2 as far as the engineering goes. If this is in midspan of the beam it could be a problem. Basically the most correct way to model a beam's capacity is to recalculate the section modulus of the beam's cross section at the biggest notches. It's just a time consuming math problem, but it can have a significant impact on the strenght of the beam. I would recommend keeping those notches to 1.5 x 1.5, there really is no need to take the full 3" out, especially across the top. By doing so you have reduced have of the beams depth by 1-1/2".

Also, while I haven't run the numbers, intuitively 3" of carrying capacity for a 7" deep timber seems undersized. The "scoops" are dramatic and for some maybe aesthetically displeasing. While the 4-1/2" seems more practical. One way to look at this is to figure 3x5 gives you 15 square inches, just calculating for vertical shear in pine gives you 3x65 (going by memory here) for only 195 pounds of carrying capacity on one end, that's only 40 pounds per lineal foot, so with 4' spacing that gives you 10 psf, which is barely enough for dead load alone!...oops, that's right, I just looked it up, they bumped up shear for pine to 135 in the current NDS... so now you can add 10psf live load!