Built-up Wood Header Calc
Built-up Wood Header Calc
(OP)
Should the Size Factor C_F for a built-up header of (3) 2x10s be 1.1 or 1.2 when using the 2005NDS Supplement Table 4A Adjustment? The notes do not specify if the factor applies to the component 2x10s or the built-up assembly. It seems that using the larger factor of 1.2 might be double dipping if I am already using the repetitive member factor C_r=1.15.
So which is correct
a) F'_b=F_b*C_F*C_r= F_b*1.1*1.15 or
b) F'_b= F_b*1.2*1.15
???
So which is correct
a) F'_b=F_b*C_F*C_r= F_b*1.1*1.15 or
b) F'_b= F_b*1.2*1.15
???






RE: Built-up Wood Header Calc
The repetitive member factor should not be used for built-up members: it is intended to account for redistribution of loads to adjacent members, as in floor loads transferred to adjacent joists through beam behavior in the sheathing. This factor represents a level of redundancy that is not present in a single (built-up) beam.
RE: Built-up Wood Header Calc
The size factor, CF, is used based on the single piece (2" = 1.1 for a 2x10). This factor is used to adjust for the fact that as you get larger and larger members out of a tree, there is a higher probability of flaws, knots, etc. in the wood and thus a lower safe stress.
RE: Built-up Wood Header Calc
DaveAtkins
RE: Built-up Wood Header Calc
As an aside, the repetitive member factor has nothing to do with the fact that there are three 2x10's. It has to do with the spacing of the members. If the members are, for example, joists spaced at 16" o.c. you can use it. If the member is a single (3)2x10 (like a single beam), this factor does not apply. So in your case, if you have (3)2x10's @ 12" or 16" or 24" on center, it applies.
RE: Built-up Wood Header Calc
DaveAtkins
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
It says, "when members are used as joists, truss chords, rafters, studs, planks, decking, or similar members which are in contact or spaced not more than 24" o.c., are not less than 3 in number and are joined by floor, roof, or other load distributing elements adequate to support the design load."
Here's my reasons from the quote above:
1. It doesn't use the term "headers" in the list above.
2. The mention of "in contact" is referring to the decking.
3. The load coming into a header is not via a "distributing element".
and finally - the main reason:
4. A header is designed for the TOTAL load coming into it. The intent of the Cr factor is to suggest that if an overload is applied to one joist or rafter, then the distributing elements can transfer that load (when the overloaded member deflects away from the load) to other members which are not loaded and therefore create a safe system. In the case of a header, there are no adjacent non-loaded members. Each of the 3 2x10's are designed for 1/3 of the total load. So load transfer to an adjacent, lightly loaded member to take up the slack can't happen.
Therefore the Cr factor doesn't apply.
RE: Built-up Wood Header Calc
I am not clear on your interpretation of "in contact". I read it as meaning the individual members can be in contact or can be spaced at 24" o.c. or less.
Based on ANSI/TPI1-2002 Page 29, "Repetitive member factors have long been utilized in wood esign to account for the load-sharing effects of assemblies that strengthen them beyond the strength assumed in designing the single members within them."
Page 30, "These increases apply to chord members where three or more trusses are positioned side by side, are in contact, or are spaced no more than 24 inches on center and are jointed by roof sheathing, flooring, gypsum, or other load distributing elements attached directly to the chords"
I realize this specification is for wood trusses, but the principles should be the same for beams.
RE: Built-up Wood Header Calc
If you have three trusses, and one gets overloaded, the adjacent ones can help. This is because the decking, or bridging, etc. can distribute the load to the other members.
When you have "one" member that is comprised of three pieces (be it 2x10's or multiple trusses in a girder truss) - all three pieces are designed for exactly 1/3 of the total load. If you apply a Cr factor of 1.15 to them, then you essentially are saying that piece 1 gets help from pieces 2 and 3, piece 2 gets help from pieces 1 and 3 and piece 3 gets help from 1 and 2.
But they can't help each other because they are all taking their full share of load - you can't overload one of the 2x10's and not the other two.
So its sort of like depending on three drunks all leaning on each other for support.
That's the way I see it personnally.
However, I note the following publication is apparently a statistical study to prove me otherwise:
http://www.nahbrc.com/bookstore/er0603w.aspx
And check out table B4 and the paragraphs above it here:
htt
So from these two publications, my argument seems weak, I'd agree....but I don't read the NDS that way and for me, I'm sticking to my guns until the NDS is modified to explicitly state that headers are applicable for Cr factors.
RE: Built-up Wood Header Calc
Don Phillips
http://worthingtonengineering.com
RE: Built-up Wood Header Calc
JAE - I don't see how "similar members which are in contact or spaced not more than 24" o.c" is referring to contact with decking. Either the members are in contact {with each other} or they are spaced less than 24" oc. Requiring contact with decking seems just as extrapolated as assuming "beams and headers" should be included.
Also, "are joined by ... other load distributing elements adequate to support the design load" implies that if they are properly nailed together (nails being the load distributing elements), they would meet the requirements of this statement.
Since the allowable loads for lumber are based on the 5th percentile of actually tested specimen values, then further reduced by another factor of safety, I don't think giving yourself 15% back for a 3-ply beam is really pushing the envelope.
But hey, rather than waste time on engineering, bump it to (3) 2x12s or go to LVL.
RE: Built-up Wood Header Calc
My though experment goes as follows. If I test three 2x10s fastened together as a header i get a specific breaking strength. If I had tested those three 2x10s sperately and added together their breaking strengths I would get the same value. (the whole is equal to the sum of the parts). Now, if I had placed those three same 2x10s in a flooring system and tested each one seperatly. I would find that each 2x10 would carry a large load, due to the fact that the decking (load distributing member)applied some of the load to the adjacent members. (the whole is greater than the sum of the parts)
I feel the distinction comes in the design process. I design the header members to each carry a third of the load. I design the floor members for their tributary load.
While the argument could be made that the repetitive member factor accounts for the propablility that some of the adjacent members will have a greather allowable strength than expected. (There is also the proprability that they are weaker, when i build up headers i always but the the worst one in the middle of me header). This consideration is accounted for with the size factor, CF. Three 2x10s are stronger than one '5x10'. As support of this I provide the observation that there are size factors for Fb, Ft, and Fc, not Fb alone as is applicalbe for the repetiative member factor.
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
Aren't three drunks leaning on each other considered structurally stable?
DaveAtkins
RE: Built-up Wood Header Calc
Physicists use space time thought experiments, we use stability of drunk guys.
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
Scotch has waaay more second order effects.
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
"It (referring to the 15% increase) is based on the increase in load-carrying capacity and stiffness obtained when multiple framing members are fastened together...."
"It reflects two interactions: load sharing or redistribution of load among framing members and partial composite action of the framing member..."
"The repetitive member increase also applies to an assembly of three or more essentiall parallel members of equal size and of the same orientation which are in direct contact with each other"
"The required condition is that the three or more members act together to resist the applied moment"
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
Of course it makes sense, I'm not sure the same can be said about your example.
If it was purely due to load sharing with other members because it is deemed (somehow) that they are not fully loaded, then the principle should be valid for all construction materials.
The repetitive member factor is used in timber design because it is statistically very unlikely that all the members in a connected group of 3 or more will have the characteristic design strength. Due to the high variablity of timber properties it is likely that most or all of the members will have a strength significantly higher than the design strength.
This is not a comment specifically on NDS (about which I know nothing), but rather, on the basis for British and Australian Standards.
Also refer <http://www.timber.org.au/NTEP/menu.asp?id=187>.
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
CJSchwartz - I agree that the references suggest that a stress increase for typical headers is appropriate. I feel this would be fine to incorprate into the IRC header tables where there are prescriptive construction requirments. I'm leary of its application with regard to the NDS, where I'm able to design a large range of built-up header and beam configurations.
GT1878 - I'm not sure the commentary really clarifies anything.
To recap I feel the discussion is centered on two concepts:
1) Is the intent of repetative member factor to account for the likelyhood that some of the adjacent members will be stronger than the expected design value.
2) Or, is its purpose to account for the redistirbution of load between framing members and partial composite action.
I feel the majority of the evidence leans toward the second.
The NDS seems to have specifcally not included built up headers or beams in its list of members and I think we can agree that they are not similar to joists, truss chords, rafters, studs, planks, or decking. In addition the commentary is clear that that you can not apply a repetitive member factor to an assembly composed of adjacent 2x10, 2x8, 2x6 and 2x4, or to an assembly of only two 2x10s. Both of these assemblis have the possiblity of adjacent members being stronger than expected.
Now there is the possibilty that the initial intent of this factor has mutated into a more narrowed definition of just load sharing. Specifically, the reptitive member factor was first added to the 1968 NDS edition based on a recomendation in ASTM D245. This would suggest the factor has some expected allowable stress component. Unfortantly I don't have a ASTM handy to further investigate.
My final comment on this matter is that because of the complex nature of allowable wood stresses I can not be sure that the obvious benefit of using multiple 2x members over a large solid sawn member is not accounted for with other adjustment factors or within the base allowable stress value. If I was designing a header and came out overstressed by 15% I would just note the top plates, rim joist, etc and move on, but then engineering judgement is a relative thing, I've hung small block chevy engine from a single grage truss with some 10 penny nails with great succes.
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
I personally believe apsix has the correct logic on why.
JJMT I believe your questions are answered in the commentary that I referenced and JAE posted. The answer is both.
RE: Built-up Wood Header Calc
I still think I won't use it for the reasons I posted above. Despite the commentary, which seems pretty clear, I just haven't used Cr for headers in the past and with the quality of wood we get these days, the fact that headers are usually fully loaded at some point in their lives, and the fact that I'm an old stick-in-the-mud, I'm using 1.0 on headers and 1.15 on joists/rafters, etc.
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
Don Phillips
http://worthingtonengineering.com
RE: Built-up Wood Header Calc
RE: Built-up Wood Header Calc
I'm with the post that said to make it 3 2x12's and quit fighting it if its that close.
RE: Built-up Wood Header Calc
But this thread is typical of what is found here at Eng-Tips.
A "nuanced" question arises from our work, we discuss it from the different perspectives of our education and experience, we research a bit, we discuss the ramifications of that research, and then we conclude a bit.
This thread then exists for future searches by others who run into the same issue. We all benefit. Thanks for the great discussion all, and thank you kar108 for the initial post.