Scarf Joint question

[Big Leu]

I've been reading and learning a lot on here and have a question I haven't found answered. I'm looking at doing a frame with similar dimensions as the one in this image. Hoping to do a 32x48 or 30x48 4 bay structure with a similar bent and common rafters. 

My question is on the long tie beam. I don't have access to a 30 or 32' beam to make it a single piece. Would my frame be compromised or is it ok to do a two piece tie beam with a scarf joint above the post? I see it done with post and beam using T brackets, but can't seem to find anything with that and timber frame scarf joint. 

Scarf 

 

[Jim_Rogers]

Offset the scarf so that it's over a brace from the center post. Not over the center post. and use a tension scarf. Lot of scarf location info here on this forum. search for it.
Jim Rogers

[Big Leu]

Great, thanks Jim! I've already learned a lot from your posts. Do you have any of the timber frame squares available? I know I read recently they were out of stock but you may be getting more in. 

[Don P]

A floor, a non uniform load, are braces supports or not? Where is the inflection point in this?  You reckon that's a reliable point of low moment? No one has come up with a vetted paper showing how to size a scarfed member. We know it is not as strong as a solid sawn one but that is all. You are being led into your safeties, how deep is the only question in my mind.

You need a timber frame engineer.

[DangleSnipe267]

I posted my take on this subject a couple of days ago in the "scarf joint advice needed" thread

"If a scarf joint is positioned above a post, you have assumed that the plate will act as a simple beam (a beam supported on each end by a post). With a simple beam, the moment is reduced to zero at the supports (posts).

If you have a scarf positioned over a brace, you have assumed that the timber is continuous beam. With a continuous beam, the moment is reduced to zero at approximately the 1/4 point between supports (posts)."

This was more in reference to a top plate and not a tying member but it really doesn't matter. However, because this is a tying member, it is preferable that you utilize a table in the scarf or similar to increase tensile strength. Don P posted some relevant beam stress diagrams from the Steel Manual in the aforementioned thread in this thread as well which you might find helpful.

As a side note, your plate sizing looks small to me.

-Tim

[Don P]

If you have a scarf positioned over a brace, you have assumed that the timber is continuous beam. With a continuous beam, the moment is reduced to zero at approximately the 1/4 point between supports (posts)."

I think there is a lot of assuming going on.
You alluded to something the other day, a "structural" brace. It is obvious that the practitioners of scarfing over a brace think of a brace as a rigid structural element, a support. If so, in the area around the post we have 3 supports. The scarf is not out in span at an inflection point, it is over a support. Nothing wrong with that, there is a better, more direct, support nearby though if that is the direction we are taking.

My question is, what are you gaining by moving the joint off the post?

[Jim_Rogers]

There are some standard rules for scarf locations and sizes.

One that is important, at least to me, is the amount each timber overlaps the other timber it is joining. I used to use 3 times the depth of the timber. But a while ago, I asked Jack Sobon what he uses for a standard. And he said he uses 4 times the depth.

So if your timbers are 12" deep then you need a four foot scarf.

The next standard rule is the location of the scarf. As nice as those drawings are, that thehardway did, that location is not the best. It could be one of the worst.

This picture below is a graph of the stresses on a plate when supported by several posts.

To read this graph you look at the line labeled "Shear". Here you will see that the shear is highest right over a post. This is the reason you don't want to have your scarf over a post. This is where the highest shear point is and you have cut your nice 12" deep beam in half.

Next look at the line labeled "moment" this is where the most load will be pressing down on the plate. It is in the middle between the supports. You don't want your scarf here as you have cut your nice 12" deep beam in half again.

You want to make your scarf someplace where there is a balance between these to force/stress points.




The reason for setting the scarf joint just off the post is that is will receive less stress there.

Look at this drawing:



Between the posts the unsupported timber will sag or deflect. Over the center post the timber will crown.
This causes this "exaggerated" wave shape effect in the timber.
At the point where the sag or deflection changes to a crown is called the point of "inflection".
This point is where the timber is neither sagging nor crowning. It has the least amount of both forces.
This, I am told, is where the joint should be made. If I understand all this correctly.

I hope this helps to explain why joints shouldn't be made over posts or directly between the posts.

Jim Rogers

Don: 
This is my standard reply to why a scarf should not be over a post. And I am open to understanding any reasons why this information does not apply.
Jim Rogers
hidden images:


 

 

[DangleSnipe267]

Don,

That's honestly a great question and its something I have given some thought to.

In my (limited) timber framing experience, I size a beam as a simply supported beam entirely ignoring the brace. I do this because it is the most conservative method of doing so. By adding the brace, one is in fact adding a support member (with reduced capacity) to the structure and as a result, reducing the span of the beam and increasing the beam bending capacity. The majority of the load will always be transferred to the most stiff element though (which in most cases is the support post).

If I were to size a continuous beam, I am of the opinion that the scarf would actually be positioned in "space" (without support) at the 1/4 point of the beam span such that the moment was zero (the inflection point). I would size the connection to match the shear capacity of the existing member and go about my day. This is how steel elements are designed all the time, whether it be a bridge girder or a waler. However, given the nature of how a traditional timber scarf is constructed, one really cannot match the original shear capacity of the beam. Lets not continue down that road.

I think that regardless of this, we can agree that it is certainly preferable to have something underneath the scarf to provide some level of support.  It just feels a little bit better and perhaps stronger to do so. Herein lies a problem. You have introduced an additional support to your equal span continuous beam which you have assumed does not exist. In actuality, you probably have a indeterminate unequal span continuous member. This altogether changes the beam stress diagram (though likely for the better) but ultimately there probably is a moment induced into the scarfed member (which is probably fine). Without rambling any further, I ultimately agree with your question, what are you gaining by moving the scarf off the post?. If you are going to provide support for the scarf, why not do so over the stiffest element?

Now a similar argument can be made that the addition of a brace creates something more complex than a simple beam between supports. I wouldn't disagree with that. but if I were to theoretically cut my beam in half at each post and over each brace I would just be creating a series of simple beams with zero moment at each support. My beam was already conservatively sized based on a much longer span with greater induced moment so I am a little more comfortable with that.

In the end it really doesn't matter where you put your scarf joint (to a degree). Historical structures have proven that frames built with a scarf located over a post or over a brace have both long withstood the test of time regardless of the math. I just prefer mine over a post. 



 

[DangleSnipe267]

To satisfy myself a little bit a ran a VERY quick model analyzing a common bent arrangement:

I analyzed a 28-foot-wide bent with a center post and 3-foot-long braces. I utilized 14-foot-long 8X12 eastern white pine and loaded it up to 1000 lb/ft to exaggerate the stress diagrams. The results were more interesting than I anticipated. For the record, I assumed every connection to be pinned (which I feel is applicable to a timber frame with wooden pegs). 

This model assumes that all the braces are acting as support.

I sketched the bent arrangement in red (poorly) for clarity



 

It is a little surprising that the braces exhibit both the highest shear and moment. This is likely because the model is assuming the braces are supporting the majority of the span. 

I think I will play around a little more with this

-Tim

[Don P]

Tim posted twice while I was writing a reply to Jim, so this is all referencing the diagrams in Jim's post not Tim's, but this all looks interesting to think about.

A continuous beam is a one piece stick of wood. A scarfed timber in the traditional timberframing sense has about 1/4 of original beam capacity in the area of the scarf. It will not stress or strain the same as a solid member.

The elevation in the top diagram shows braces. The moment and shear diagrams neglect them. There is no joint shown in those diagrams, that is not a scarfed beam diagram in any way shape or form.

The number of supports and locations of the scarf in relation to supports is not as the diagrams represent. You and Sobon have called a brace a beam support and here are clearly saying it does not exist in those beam diagrams.

Mental exercise; Remove those braces, that is what Jack drew in the beam diagrams there. Still comfortable?

What is the behavior of that beam going to look like. Do you think it is going to be nice smooth fair curves indicating smooth flow of stress or is there going to be something going on around that scarf.

I've asked for engineering regarding sizing beams that have had this joint inflicted upon them... crickets.

In this situation we have roof point loads, nonuniform floor loads, a dubiously cobbled together timber joint in span (or maybe just notched some unknown amount out in span), and a uniformly loaded continuous beam diagram served up. I've got an order for a hamburger and a recipe for duck sauce. What could go wrong.


Alternatively, join the 2 beams over the direct path to ground, as the building code and most engineers would have it done. Don't use a scarf if it cannot resist the shear... you don't need one. You need to transfer some amount of tension while properly supporting a pair of simple beams.

We aren't to this magnitude I think but more to consider.
From the red book;

The classical man made splices cannot develop half the capacity of Nature's basic material. The only way to carry more than half of a timbers gross capacity through a splice is to transfer the force out into side plates, past the break in the timber, and back into the other tension member.

I suspect a simple steel strap or a spline between members would suffice here.

If you put a bolster over the post and braces it provides room to develop good connections, effectively reduces span, you could use the fixed end beam diagrams. This is easy to document.

[ben bb]

@Don P Thanks for sharing this.  Is there a traditional way of of connecting the plate to the bolsters in that pic?  Or do you just screw down from the top?

[Don P]

I just went looking for the awc.org 's WCD5 "Heavy Timber Construction" ... it looks like the free links to their publications are gone  :-\. I hope not.

Typically I have lagged down. Check your uplift, all elements in that load path have to add up, beam to bolster, bolster to column, column to foundation. I'm getting ready to thru bolt side plates on each side of basement posts that will extend up and sandwich the girder running over the posts, thru bolting the side plates through it as well. This will help with lateral support for the beam as well. That is more what i was linking to in the awc pub, more mill type construction.

These were on the porch above. That had a built up and trim wrapped 4 ply beam above the bolsters so I used Simpson angles from inside plies of the beam to the bolster as I assembled the beam. I had a factory in the basement and thought "these should be a commodity item at the building supply".



 

On these I ran the post tenon thru the bolster and up to the top to the beam. The beam sections fork around the tenon hopefully helping keep everything in line. These were prebored, counterbored, and piloted then 3/8 x 10" lags from beam to bolster. The weak link is the bolster to post pegs, i had 2 offset and caled it 2,000 lbs uplift and the braces are gravy.





I'm amazed she has put up with me, that may be the night I ordered the chain mortiser :D.



 

I remember that stack behind her, it is ailanthus that became manshed panelling, and looks good.. the new ash  .

I've got another pic somewhere from in an old gristmill that I don't see in my gallery. The central post in the mill was collecting 3 stories of load, cleaners ,a sifter that had a counterweight heavy enough to show the shaking that the building was taking, grain bins filling the second floor, many grain elevators and chutes, etc, serious load. The bolster under that floor girder was a good 12' long gracefully tapered from post to ends by several inches, chamfered edges. The post was a capitalled octagon about 12x12.

Whoops, sorry, I was rocked back remembering and smiling. I could see retiring to a life of leisure at a water powered mill  :D. I'm sure it would be fun for the first hundred bushels.

[Don P]

Also take a look at Asian and japanese inspired craftsman style work, you'll often see iron straps wrapping the beams and bolsters.


I found WCD5 here;
PDF Viewer (awc.org)

Read note 2 under "construction details" on pg 2, (this is written into the commercial code... uhh, somewhere around section 2302) and then see the sketches on pg 9.

[ben bb]

@Don P Thanks so much for the thorough response.  Very helpful as always.

[Don P]

I was looking on the old computer for the gristmill pic. I didn't see it but came across this from the Greene's Gamble house. I hope the copyright expires when I forget where it came from. Beams coming in mainly from a bridge on your left. IIRC that is a post in a wall corner that runs to the right. This is mostly decorative but it is most certainly structural, mainly showing the heavily Asian influenced Arts and Crafts style.



 

[tule peak timber]

Took the tour of Gamble, Doug Fir a GoGo ! The docent running the tour boasted that ANY of our local woods were worthless for anything .....Still laughing about his comment. The subset metal is a design feature I use frequently.