Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Heavy Timber Knife Plate Design

Status
Not open for further replies.

Mstrux92

Structural
Mar 21, 2020
10
Hello, I am designing a custom knife plate for some heavy timber beams. Typically, I would go with a proprietary hanger, but this case is unique. The Main Beam is (2) stacked 12x14 DFL beams, the side beam is a 10x10 DFL. The Side beam is centered in the stacked beam. The connection to the 10x10 must be concealed. I have attached a detail for my proposed connection. For calculations, I checked the connector capacity in the side beam, main beam, row/block tear out in main beam (still need to check side beam), and steel plate welds. I am also planning to check pullout of the connectors in the main beam. My connection must hold 5000 lbs snow load (ASD).

I greatly appreciate any feedback/discussion on my connection.
 
 https://files.engineering.com/getfile.aspx?folder=88aeb1ff-4407-462d-a5f8-a56c022f1ea6&file=Concealed_Beam_Connection.pdf
Replies continue below

Recommended for you

I would try to move screws up as much as possible. Check edge distance of screws in the main beam. I would also use longer and bigger screws probably. Check pullout of the screws because of eccentric connection (moment).
I would try with one column of dowels to reduce eccentricity. Dowel can be 8" long with no problem.
Add vertical full thread screws right by the dowels in secondary beam to prevent splitting (one on each side of knife plate).
 
Mstrux92:
How are the two stacked 12x14 beams attached together, can they move w.r.t. each other at their mid-height horiz. joint? If so, how does this figure into your connection and the 36 – SDS screws. I would be tempted to look at a ‘zee’ shaped end connection, depending upon what both ends of the 10x10 bm. look like and how that bm. would erect. I would have the top horiz. leg on the ‘zee’ dapped into the top 12x14, and the bot. horiz. leg of the ‘zee’ dapped into the end of the 10x10, near its bottom. Then, you would not have to depend so much on screws and dowels in shear and bearing to transfer the 5k reaction, and any eccentricity in the reaction would be smaller and could be taken up in bending in the ‘zee’. You need a row or two of WSNTL3S screws below the upper horiz. leg, and a few up through holes in the bot. leg, to hold it all together. Recess the ‘zee’ into a routed pocket in the side of the 12x14’s. You need to plug a slot in one side and a few holes in the bot. of the 10x10.
 
Can you use a normal mortar and tennon joint?

Dik

Dik
 
Mstrux,

I would check the moisture content of the wood pieces and the moisture content of the environment that the timber is going to be exposed to. There is a concern for shrinkage transversely in both directions that will affect the beam connections separating and failing.
 
Could you go with an "almost concealed" seat connection? Similar to what you have drawn, but turn the shear tab into an upside-down T section so you have something for the 10x10 to bear on at the bottom. If you notch the 10x10 at the bearing plate, the plate can be flush with the underside of the 10x10 and you'd barely see it.
 
Mortise and tennon would be completely hidden and would also be traditional.



Dik
 
dik said:
Mortise and tennon would be completely hidden and would also be traditional.
I'm imagining this would not be easy to build. How much can you really cut primary beams? Probably not much. And still you need some screws or dowels to keep elements together. Could you post a sketch what you mean or more detailed description?
 
molbden: Those are great suggestions. My calc passes, with one row of 8" dowels. I am hesitant to use a knife plate (unsupported bottom flange) although I frequently spec the Simpson knife plate hanger (which has similar capacity values).

dhengr: The beams will be connected together with 1" diameter by 16" lag bolts centered in the beam (8" penetration to each beam) spaced at 9" o.c. I like the idea of the Z-shaped hanger. I am a little worried about how they would erect the beam. I would imagine they would build both stacked beam frames, and then set the 10x10 beams in place. This would be difficult with the dapped ends on the 10x10's.

dik: Personally, I try to avoid mortise and tenon joints unless it's "lightly" loaded or nonstructural. I just don't have a high comfort level with notching beams. However, my original though was to notch the double beam 1.5" at the 10x10.

CANPRO: I like this idea, it is similar to dhengr suggestion. It would be constructable, however, not completely concealed. I will run it by the architect, and see what they think.

bones206: I did consider a double knife plate configuration, and it does make sense for such wide beams. Somewhat uncharted territory for me though.
 
Mortise and Tenon joints have proven capacity many times over in practice. But require a highly skilled carpenter to do properly. I've had them provided by a specialty contractor on a heavy timber frame. Honestly they look absolutely gorgeous, but I wouldn't trust the average contractor to do it properly.
 
Agree with jayrod12, I would be very hesitant to design a mortise and tenon joint unless I knew the Contractor could execute.

Since you will be removing material in the beams to form the mortise, beams may need to be oversized some to accomodate this. The tenons will also need to be checked to be able to transfer the shear load since it too will have a reduced area.

 
With the connection being at mid span of the stacked beam, I would rather not allow notching for a mortise joint. I am detailing a mortise and tenon for the condition where a post is located.

I expressed my hesitation with a knife plate to the architect, and they said an "almost" concealed connection per dhengr and CANPRO would be acceptable. I'm going to go that route. I think if the stacked beam were a solid member, I would feel more comfortable with a custom knife plate, modifications made per molibden's comments.

Thanks everyone for the input, that was very helpful!
 
Mstrux92:
Slow down a little and reread my first post (last few sentences) for its full meaning. Ask the wood fab. guy what the width of the chain on his chain mortiser is and make your ‘zee’ just a little thinner, if that’ll work strength wise. He can cut thicker slots too. The ‘zee’ is set back into a routed out pocket in the 12x14’s, with its own horiz. slot for the top ‘zee’ leg, and the Simpson WSNTL3S screws can be countersunk, so the ‘zee’ is flush with the side of the main beam. The 10x10 beam is end slotted for the horiz. ‘zee’ leg from one side only, and the 10x10 is slid into place from one side, and that horiz. slot is filled with a DF ripped piece of wood. The bot. screws get plugged too. That end slot is .75” clear, up in the end of the beam, so check the beam for a 1.25” (or whatever) end notch.

Edit: You should probably put a small stabilizing angle connection atop the 10x10 beam and screwed into the top 12x14 main beam.
 
OP:
dik: Personally, I try to avoid mortise and tenon joints unless it's "lightly" loaded or nonstructural. I just don't have a high comfort level with notching beams. However, my original though was to notch the double beam 1.5" at the 10x10.

I've done a lot of timberframe buildings... actually, my favourite type of structure. M&T joints can be used for real loads... it does take a quality woodworker...




Dik
 
dhengr: I believe I understand now. The ripped section on the far end will screw to the bottom of the plate, concealing the connection. I like this detail. Hot rolled Z sections are difficult to find, instead I have detailed two angles with 2 stiffener plate connecting them. The stiffener plates will knife into the beam ends.

dik: Thanks for the input. I could limit the tenon notch per NDS. For the Mortise, I could analyze the main beam with a reduces cross section. However, NDS does say "No notches permitted in middle third of span". This is my only hold back.
 
Is there not provision for rational design? Almost any old timber framed barn will have numerous notches in the middle span?

You can also put the pin in from the bottom proud so it is exposed.

Dik
 
Mstrux92:
There’s still something lost in the translation, and I can’t scan stuff at the moment. Hot rolled Z’s are the wrong proportions, and angles are a crappy choice too for the ugly toe to toe weld condition and large radius. Make your Z out of bar stock and plates, and thick enough so you don’t need any stiffeners. The two legs are 3-4” wide bars, about 7-8” long (thus about 1” left on each side of the 10x10), and the web is about 7-8” high. The web gets bevels for groove welds on opposite edge corners, and on the other side from the groove welds is a fillet weld. The web butts to the bot. or top faces of the top or bot. legs. The Z web gets drilled and countersunk up near the top leg for the WSNTL3S screws so the beam reaction can’t tip/roll it out of its upper slot. The bot. leg gets 3-4 holes for screws from below, and up through the Z leg, to keep the beam on the Z. I think I asked and you never told us, so I assume the other end bm. support detail is the same. The 10x10 end (end view) gets horiz. a slot, 3.25 or 4.25” deep (into the length of the bm., matching the Z leg width +) with its bot. 1” up from the bot. of the bm., plus the leg thickness, plus a little clear. That horiz. slot is cut through one beam face, across the bm. width, to within 1” of the other bm. face, and stops. The bm. then slips onto the Z’s from one side, and is pushed horiz. into place. The 3-4 bot. screws are driven up through the holes in the bot. leg to pin the bm.in place. You might install a few more screw up into the bm. bot., just beyond the cut slot to protect against notch corner splitting. Then, the slot on the open side of the bm. is filled with a DF ripped piece 3.25 or 4.25” long (slot depth), .75” wide and the height of the slot. I’m tryin to paint you a word picture, draw that/those sketchs as you read. I’ve not run any numbers, the exact sizes and dimensions should be based on your calcs. and detailing.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor