PDA

View Full Version : Splices in girder/built up beam not over vertical support posts



jar546
November 4th, 2010, 07:56
Built up girders/beams with nominal lumber. Let's use a (4) 2x12 for example.

If the splices are in between vertical support posts, can you allow it?

Although it has been common building practice not to do this and common sense, the IRC does not truly cover this issue on built up girders/beams.

502.6 is pretty close to covering this but does not appear to have the correct language. There is language elsewhere (I don't have time to look it up right now) that states ".........sufficient to handle the load imposed" or something like that which is also a stretch.

What say you.....?

Jobsaver
November 4th, 2010, 08:12
We have always required the splices to bear.

fatboy
November 4th, 2010, 10:21
Over bearing........(is that a pun?)

We've felt that without direct bearing, then you are having to consider the shear of the fasteners, which then we consider engineering......... JMHO

jeffc
November 4th, 2010, 10:44
I'm not sure I follow, are the joist abutting each other over the bearing point and not lapped 3"?

Mac
November 4th, 2010, 11:13
From AF&PA Wood Frame Construction Manual: "End joints of the nailed (girder) lumber shall occur over the supporting column or pier"

fatboy
November 4th, 2010, 12:03
jeffc, no, if I understand the OP, the butting junction is not occuring over a column........

Paul Sweet
November 4th, 2010, 12:28
The girder can actually be stiffer if the splices aren't over the bearing points. In a continuous girder, the maximum negative moment is over the supports, the maximum positive moment is near midspan, and there is a point of inflection, where moment is zero, near the quarter point of the span.

You can splice all of the plies at this point, as long as you use the proper type of hanger to support the side of the girder that doesn't bear on the support. (http://www.strongtie.com/products/connectors/HCA.asp#gallery) This method is more common in steel or laminated timber construction than in light wood framing.

In a built-up girder the splices should be staggered so that only one ply is spliced at each quarter point, and each ply bears on at least 2 points. The only problem with doing it this way is that it's a bear to build up a 40 or 50 foot long girder spanning over several piers. It's usually quicker & easier to just throw in an extra ply and put the splices at the supports.

GHRoberts
November 4th, 2010, 12:30
If you look at the equations for a beam that is supported at the center and uniformly loaded, the span and max deflection are identical to a beam simply supported at both ends.

I have no problems with the ends falling near the midpoint of a span.

jar546
November 4th, 2010, 13:29
From AF&PA Wood Frame Construction Manual: "End joints of the nailed (girder) lumber shall occur over the supporting column or pier"

What section please?

jar546
November 4th, 2010, 13:36
Getting 100% resistance on this one and they are not buying anything I have to say. Little help here please.

FredK
November 4th, 2010, 13:58
jar I went here:

http://www.awc.org/HelpOutreach/faq/FAQfiles/BeamConnections.html
and then here

WCD#1, Details for Conventional Wood Frame Construction
and under illustration 15 shows how a built up beam should be done. Anything else I'd ask for engineering.
This is the book and wait for it to download and click on illustration.
http://www.awc.org/pdf/WCD1-300.pdf

Mark K
November 4th, 2010, 14:05
This is not something that the code specifically allows. Thus engineering calculations are required if they do not want to try to justify what exists.

fatboy
November 4th, 2010, 14:22
"This is not something that the code specifically allows. Thus engineering calculations are required if they do not want to try to justify what exists."

BINGO

Mac
November 4th, 2010, 14:32
Engineer it or.... maybe additional posts can be added to support the splices?
That might be easier than replacing part of a completed floor system.

Remember - you are not responsible for this condition. You haven't done anything wrong.

jar546
November 4th, 2010, 16:55
UPDATE:

Just got an engineered fix with a stamp from a PE on it. Called the contractor back and told him thank you and by the way, if what you did was OK why didn't the PE sign off on it as is?

We spoke for 30 minutes and have an understanding now.

All is well that ends well.

FredK
November 4th, 2010, 18:26
....... if what you did was OK why didn't the PE sign off on it as is?

Was the a silence/long pause before the comment "Damn engineers don't know nothing," kicked in?

fatboy
November 4th, 2010, 18:38
"
....... if what you did was OK why didn't the PE sign off on it as is?"

BINGO!

Good call Jeff, sticking by your guns...........

Uncle Bob
November 4th, 2010, 18:44
Don't have my book with me right know; but, all end joints must be supported. A splice that does not occur over a support must be corrected by an Engineer. Not occuring over support makes the splice Engineering; and, all that I have seen; the Engineer required a support below the center of the splice.

Uncle Bob

jar546
November 4th, 2010, 19:08
I think it was my avatar that got me the engineers report spec :)

fatboy
November 4th, 2010, 19:16
And it is AWESOME!

peach
November 4th, 2010, 19:33
I'm offended by jars avatar, by the way... partially because even in my prime, I didn't look like that, and now that I'm past prime, I'm jealous..

Besides that... splices can be offset (and really should be). How stupid is it to build a beam with all splices occuring at the same point? You can build a really long, strong beam with pieces parts of different lengths.

GHRoberts
November 4th, 2010, 20:57
peach ---

The typical girder is built up of pieces each with three points of support - the ends and the center. About 1/2 of the pieces are spliced at any support.

RJJ
November 5th, 2010, 07:39
Jeff: I just like to know how you got a picture of my wife? And converted it to an avatar!

Jobsaver
November 5th, 2010, 07:56
On the avatar. Two thumbs up! Still, I've already been try'in to convince my suspicious wife that this is just a building code forum . . . and the avatar ain't help'in any!

Jobsaver
November 5th, 2010, 08:00
Oh yeah, and Peach, don't be jealous. You can still but you some the same way she bought hers!

Uncle Bob
November 5th, 2010, 18:44
GH,

"The typical girder is built up of pieces each with three points of support - the ends and the center. About 1/2 of the pieces are spliced at any support."

George, it's a girder; not a girdle.

Yep, that avatar has to go; it's messing with people's minds. :)


Uncle Bob

peach
November 6th, 2010, 06:51
Just like top plates where splices occur in the center of the cavity.. it's a small version of what Jeff is describing.. (Jeff, you're having way toooooo much fun with the photo I sent you of myself.. :)

Architect1281
November 6th, 2010, 12:40
A little knowledge is a DANGEROUS THING

Paul Sweet said >> The girder can actually be stiffer if the splices aren't over the bearing points. In a continuous girder, the maximum negative moment is over the supports, the maximum positive moment is near midspan, and there is a point of inflection, where moment is zero, near the quarter point of the span.

What Paul Sweet said is true for a multi (3 Minumum) span girder
What Paul may not know is that WOOD dependes on parrallel greain cellular structure to transfer its loads to the ends or bearing points.
(the glue between the continuous straws transfering stress from strand to strand)
cut through those straws and there is ZERO stress transfer
A tree piece built up girder with a single splice not over the bearing point is no better than a two piece girder.

You may occasionaly see in steel girder multi-span construction what is reffered to as Cantilever design where the splices and the center span is actually between supports
With a major distinction. thare are bolted splice plate connectors gennerally at the points of inflection of the deflection curve where the moment is ZERO.

In wood construction in the TImber manual you will see steel saddle plates at similar points to transfer items such as shear and bending

So Paul after you learn about Moment transfer keep on reading.

JAr let me guess that the engineered fix if not remove and replace involved some sort of bolted steel plate

jar546
November 6th, 2010, 21:05
The engineered fix involved adding another layer on each side of the girder with the new layers having their ends land on the support posts then a specific bolt pattern to attach it all.

DRP
November 7th, 2010, 09:00
Timberframers have successfully used scarf joints at ~ the inflection point to create long beams for centuries. Here is more reading on that topic;
http://www.tfguild.org/joinery/part6.pdf

That entire book is available here and is excellent;
www.tfguild.org/joinery/joinery.html

I'm not expressing personal view, just info, I've argued both sides of this sitting atop a drywall bucket. To my eye it looks like beams overhanging posts rather than a continuous beam. I believe 1/4 point splicing is allowed under Canadian code.

Uncle Bob
November 7th, 2010, 10:32
Well,

If the end joints of the pieces are not over supports; then, It's not conforming to the code; and therefore requires an Engineer and an Engineer's design.

Uncle Bob

Architect1281
November 7th, 2010, 10:51
So the fix was make it be as many continuous pieces as it is supposed to be and discount the spliced piece to zero
a logical approach. and adequate bearing was supplied. and the engineer had the answer the framer (cough wheeze) should have had.

more of a continuous load path as opposed to the pinball table load path

GHRoberts
November 7th, 2010, 11:26
So the fix was make it be as many continuous pieces as it is supposed to be and discount the spliced piece to zero
a logical approach.

It is only logical if you have engineering that shows it to be logical. Your opinion; my opinion don't matter.

---

The engineer came up with a logical (engineered) answer. Other engineers would have come up with other logical (engineered) answers.

Architect1281
November 9th, 2010, 17:45
Si Sienor Eppes (Numbe3rs) you got me there
fortunately for me I have the stamps to stamp the answer as well as the CBO to ask the question
besides logic is a subset of mathmatics (also known as symbolic logic)

peach
November 9th, 2010, 17:53
The best way to have a girder is for it to span continuously (which is why we sometimes see steel beams instead - to keep the depth of the thing reasonable). A mid bearing point is maybe best placed where there are no splices..

GHRoberts
November 10th, 2010, 13:56
The engineered fix involved adding another layer on each side of the girder with the new layers having their ends land on the support posts then a specific bolt pattern to attach it all.

All this talk about bolts and shear transfer got me thinking (as opposed to just typing.)

If the joists rest on top of the girder, there is no need for shear transfer. The various members will just deflect and the deflection will transfer the "shear."

On the other hand if the joists sit on hangers that are attached by "short" nails. There is a real issue. In a 4 layer built up girder to transfer 1/2 the load to the second layer you need about 1/2 of the hanger nails to extend through the second piece of wood in the built up girder. That seems like a lot. (I think that is 3 nails between each joist.)

jar546
November 10th, 2010, 14:20
Since they are not adding depth, simply width, it will be harder to correct the deflection issue too

ajweaver
November 11th, 2010, 16:44
I was taught all the shear is transferred through the required fastners of a built up beam/girder.

A three ply we can verify both faces were nailed.
On a 4 ply or 5 ply we require bolting.

ajweaver
November 22nd, 2010, 14:04
I just reread my post-

Just to clarify...

I WAS NOT taught the splices in the beam are ok-
The fibers in the wood are required to be continuous for deflection issues.
In specific-the tension side(bottom).

If you have "pieces of wood" scabbed on to the sides of continuous beams, they act more as "stiffeners".
They do not count as additional plys.


I WAS taught

The loads imposed by hangers are transferred through the fasteners connecting the beams themselves.

I'm sure that was a given..but just to make sure.


I was taught all the shear is transferred through the required fastners of a built up beam/girder.

A three ply we can verify both faces were nailed.
On a 4 ply or 5 ply we require bolting.