Rafter to plate question, 1 1/2 story house.

[xcgreene]

First time post, LONG time reader. This frame will be my 4th design and construction, but my first 1 1/2 story. So my questions come from this weakness.

Jack Sobons book has been a bible for me over the years, but Ted Bensons "building the timber frame house" has the best reference for a 1 1/2 story design I am looking for. With 3 children, I'm looking for the best way to optimize space, and use minimal timber.


As you can see in the diagram below, using the struts and upper tie beams will allow for maximized space while providing adequate  support where needed. (this particular design was discussed briefly in Bensons book)



My questions reside in the circled area above, where the rafters (which will carry purlins) will meet the plate, or top of posts. Should I use a continuous plate across the tops of the posts or use a girt type tie in plate between each post?  See my diagrams below to illustrate each idea.


 
And below is a more detailed view of the joinery of each option.


 
The continuous plate on top of the posts would be stronger in one dimension, but I fear it could roll outboard a bit with rafter forces, and I also don't like the solution of the peg drilled down as the only retention.

Installing the plates down, socketed to the sides of the posts, would negate the fear of plate rolling, would offer a better way to pin the rafters down, and would be easier to handle as well. But I fear 3 tenons at the top of these posts may be too much to ask, structurally speaking.


With your experience, what are your thoughts on these two options?  I'm leaning towards the rafter to post top, but I'd love to hear from you all.

[everythingwood]

I am certainly no expert but the first rafter connection option (with the birdsmouth) is a no-no.  It reduces the effective depth of your rafter and is prone to split at the birdsmouth.  Better to use an option where the rafter bears fully on the plate.

How far apart are the bents spaced?  As long as the plate is sized to bear the weight there is no rule that the principal rafters have to rest on the plate at the post.  You could alter the spacing to avoid crowding the joints on top of the posts.

An oooh boy, on the option with the continuous top plate you show a scarf joint!  those baby's seem to be a little controversial!  I kid a little... despite being used in thousands of structures through time they apparently aren't in the engineering standards so they seem to lead to a lot of discussion but, from what I have seen, no resolutions.

I hope you keep us posted on your progress.   

[everythingwood]

Also, those struts are actually working in tension no?  If so, should you consider using a lapped dovetail joint?  I have read they are better in this situation since you aren't counting on the peg alone to resist tension?

[bm.stanford]

I too am interested in knowing which way you choose to build and why.  I'm designing my first frame and have been stumped about this same issue.  I've read multiple posts where there are concerns the top plate could roll but on the other hand I've looked at countless pictures of old frames and they appear to be built just as your first diagram.  Is it the weight of the insulation or heavier roofing materials that have changes to make one look at this differently?

[xcgreene]

The bents are only 12ft and 8ft apart.  So that is a comforting attribute.  though the bents are 20ft wide.  But that length is perfectly manageable too.  This building will be 20ft x 32ft.

Regarding the birdsmouth, it is not overly clear in the drawings, but there is a tenon there that would assist it structurally if it were placed on top of the post.  But if the rafter is placed on top of the plate, I can certainly see what you are implying about reduced effective depth.  But I compare it to a half lap floor joist in a way.  The belly of the rafter (or joist) is still providing rigidity. but a floor joist can be tapered to reduce checking, I can see where the rafter does not have the possibility of this benefit.

If I place the rafters on the plate, I'm not too worried about crowding of joints, in this case, I'm more worried about HOW to tie in this rafter to the plate in a 1 1/2 story like this.  I really don't like the methods suggested in the books I've read.

Yes those struts will be under tension. I agree a dove tail is perfect for tension loads.  But it would take away even more meat from the top of the post where I am concerned about already. A mortise and tenon, as long as it meets proper end distances, edge distances, and spacing, should be fine in this regard.  The TFEC standards provide all the calculations to use for Mortise and Tenon connections loaded in tension.  But anyone reading this, I encourage you to use those calculations before moving forward with any designs you read like this online.

I laughed when you mentioned the scarf joint too.  I agree, lots of contention out their for sure. A timber framer criticizing the structural viability and use of a widely utilized scarf, its almost an oxymoron. If someone would like to mill me four 32ft long 8x8 or 8x10 beams, carve them up, and install them for me, to save me from the certain death of scarf joints, please send me a PM.  Sarcasm aside, I suppose we all get lost in the black hole of calculations sometimes. because its important.  Back to basics and pragmatic, I have to remind myself this all the time. 

BM.stanford, as the saying goes, if in 1969 we had the rules we have today, we would never have gone to the moon. Eventually, everything will have caused at least one incident, so everything will eventually become illegal, by current rule.  I exaggerate of course. But allowing people to have the option of taking their own risks is more important than I think we currently realize. Yes, perhaps that means someone has to learn a lesson on their own that others have learned a thousand times over, and that's really hard to watch. But somewhere buried in that idea, and just outside of my preceptive reach, is our freedom.

[Ljohnsaw]

One way to eliminate the outward thrust of the rafters is to install a ridge beam.  Have you considered that?

 I have a tremendous snow load so just a ridge beam was not enough without having beam-sized rafters.

[xcgreene]

ljohnsaw, A ridge beam would absolutely solve the outward thrust problem, and is typically an excellent solution, but requires a lot of extra timber.  I would need another post in the middle of each bent, from the basement on up.

[everythingwood]

My shop has 24" knee walls in the attic for extra storage space and there has been no evidence of the plate "rolling" in 20 years.  The roof has a 7:12 pitch.  Granted I did toenail the plate to each post with two 8" pole barn nails because it was my first and I was leery of joints that were pegged only.  I also have a scarf joint and, surprise!, it shows no sign of failure.  



 

If you worry about plate roll you could add some straps that hook over the top of the plate and run down the inside of the post for a foot or so.  Make them wrought iron and they would even look nice.  I have seen similar things on a barn that was much larger than your frame and may have actually had an issue with the plate rolling.

For my next build I am using principal rafters on 6-foot centers.  I was planning on a step-lap rafter seat as shown in Beemers book, but now I'm thinking of using a housed birdsmouth joint and adding a tail.  The housing should eliminate the threat of the rafter splitting at the birdsmouth angle.

ETA:  I'm sure you know this but you don't necessarily need posts all the way to the basement if you use a ridge beam.  You just need to size your anchor beams to handle the uniform load from the upstairs living space, and the point load from the roof ridge beam.

[mattgancz]

Dear xcgreene:

The outward thrust of the rafters is going to be an issue in either case with your design.  The struts that you added will be in tension, so essentially you will be relying on the shear strength of the pegs to hold everything together --- bad idea.  

I would use a top plate if I wanted to use common rafters, but in that design I would add either a ridge beam, or midspan plates.  In either case, I would add additional posts to the design.

Your drawing depicts principal rafters, so I would move in the direction of using the rafter foot to post, but I would add an additional post.  Without a post, the rafter spread is definitely going to be an issue.  Along with that, getting the second floor deflection within acceptable values is going to mean putting a large tie beam in for the 20' span --- it would look overpowering in a small space in my opinion.  If you add an additional post you could bump the building to 24' or 28' wide and gain some cheap square footage on both floors.  

You seemed concerned about putting three tenons into the top of one post.  Since you have wall plates (the lower ones), and top plates (the upper ones), it is the wall plates that you need to be concerned about in terms of loading.  They will carry the second floor loads, and include the braces.  The top plates really only hold sheathing and stiffen the frame --- they are not load bearing per se.  Putting rafter feet and top plates into a post is a common high post design; you have lots of standing frames to back you up.

[xcgreene]

Dear xcgreene:

The outward thrust of the rafters is going to be an issue in either case with your design.  The struts that you added will be in tension, so essentially you will be relying on the shear strength of the pegs to hold everything together --- bad idea.  

I would use a top plate if I wanted to use common rafters, but in that design I would add either a ridge beam, or midspan plates.  In either case, I would add additional posts to the design.

Your drawing depicts principal rafters, so I would move in the direction of using the rafter foot to post, but I would add an additional post.  Without a post, the rafter spread is definitely going to be an issue.  Along with that, getting the second floor deflection within acceptable values is going to mean putting a large tie beam in for the 20' span --- it would look overpowering in a small space in my opinion.  If you add an additional post you could bump the building to 24' or 28' wide and gain some cheap square footage on both floors.  

You seemed concerned about putting three tenons into the top of one post.  Since you have wall plates (the lower ones), and top plates (the upper ones), it is the wall plates that you need to be concerned about in terms of loading.  They will carry the second floor loads, and include the braces.  The top plates really only hold sheathing and stiffen the frame --- they are not load bearing per se.  Putting rafter feet and top plates into a post is a common high post design; you have lots of standing frames to back you up.

Yes the struts will be in tension, but its not the peg that would be of concern. black locust has a bending yield strength of 18,000psi. The edge and end distances seem to be far more critical in this situation, as they are the load bearing weaker wood. 
 A mortise and tenon connection may be loaded in tension provided the following:
-faces of the tenon and mortise are in close contact 
-the load is perpendicular to the peg
-edge and end distances in relation to the peg are adequate, 
-the thickness of the mortise side walls are the same or greater than the tenon
-the size of the peg is 3/4" to 1 1/4"
Now this certainly isn't to say its the very best thing to do.  But it is saying that its acceptable, and not a bad idea.
The 20ft span will have interior stud walls under them, minimizing the issue of 2nd story floor bounce.  The outer most ties will have hefty sheathing nailed to them as well.  They will be 8x10 spruce. I have a 20ft 8x10 tie running free through my kitchen, I personally like the larger timbers, but I can certainly see where some may not like it.
Perhaps I'm trying to get something for nothing with this design, but I've seen it represented and described in Bensons book, and I really want a crack at it.
 

[mattgancz]

Indeed, black locust has a bending strength of 18,000 PSI --- whether it is a toothpick, or a 1-inch peg.  I wouldn't recommend toothpicks in your design.

In regards to tension joinery, you said, "Faces of the tenon and mortise are in close contact."  Shouldn't this always be the case?  Here is a link to the TFEC Standard of Practice.

You also mentioned loading perpendicular to the peg.  I don't typically (i.e. ever) load parallel to my pegs.  It is interesting to imaging though.  Here is a link to the TFEC edge spacing of pegs.

I didn't catch that you were infilling with a stud wall; I was imagining SIP of Larsen Truss; so that is helpful information, but it doesn't affect the outward thrust of the rafters.  If you are bent (pun intended) on using the design as you have it; go for it!  It is your house.  If you do not need a PE stamp for your project, you could probably pay a PE to take a look and give you his advice.  It might be a cheap $120 to pay them for an hour of their time.  

[xcgreene]

Indeed, black locust has a bending strength of 18,000 PSI --- whether it is a toothpick, or a 1-inch peg.  I wouldn't recommend toothpicks in your design.

In regards to tension joinery, you said, "Faces of the tenon and mortise are in close contact."  Shouldn't this always be the case?  Here is a link to the TFEC Standard of Practice.

You also mentioned loading perpendicular to the peg.  I don't typically (i.e. ever) load parallel to my pegs.  It is interesting to imaging though.  Here is a link to the TFEC edge spacing of pegs.

I didn't catch that you were infilling with a stud wall; I was imagining SIP of Larsen Truss; so that is helpful information, but it doesn't affect the outward thrust of the rafters.  If you are bent (pun intended) on using the design as you have it; go for it!  It is your house.  If you do not need a PE stamp for your project, you could probably pay a PE to take a look and give you his advice.  It might be a cheap $120 to pay them for an hour of their time.  

Mattgancz, I may seek out a PE for more advise, always worth the conversation and information.  When I was mentioning in my last post about mortise and tenon connections that may be loaded in tension, the criteria I provided was paraphrased directly from the "TFEC 1-2019 Standard for Design of Timber Frame Structures and Commentary". I agree with your comments about the tenon and mortise faces, and parallel peg loading, but have a read on this standard, and see what you think.