Wood I Beam of dimensional lumber

[sixgun]

My first post here. I've lurked around, used the search function - but I can't seem to find what I'm looking for. So I'm hoping someone can help me.

I am designing a shed for my backyard. It will be used to house a storage tank for water heated by solar collectors on the roof of said shed. I have the rafters and walls designed to bear the loads of the collectors, snow (60#) and dead weight of the roof.

The shed wil be 12' X 16' and will set on an 8' x 16' basement foundation. 4' of the 12' width of the shed will cantilever out from the foundation.

6" x 10" pressure treated floor joist on 24" centers appear, by my calculations, to be acceptable for the floor joist. I am interested, however, in building wood I-beams of treated wood to use for the floor joist. I can not find, however, a formula (or formulas) that will give me the data I need to decide on what lumber to use for the chord and flange sections of the I-beams. (My google-fu is usually pretty strong, but on this question I have failed...)

For example: Can I just calculate the S and I for each member and add them? Seems I'm missing some of the advantage of the I-shape then. I'm really only adding depth to the member with no accounting of any change various widths of flange would create.

Can anyone help me out here?

 

[TVP45]

If you want to make a structural shape, sources such as Machinery's Handbook are valuable. Be aware that you cannot get away with just nailing the flange to the web.
Also, if you are subject to a building code for this (sounds like you are), you really need to check with the code authorities. There are 2 issues: pressure treated lumber may be significantly weaker; and you may need a registered engineer to review and approve any beams.

 

[sixgun]

Thanks for the reply, I'll check out the book.

The shed is below the square footage requiring a permit, but I want to get it correct...

It is an experiment for me in many ways. I've gotten the technical data on the pressure treated wood from Boise Cascade, the producer. They've been very helpful about recommending adhesives etc. I plan to router a groove in the flanges, glue the web to the flanges and then compress them while the glue cures. Timber screws every X inches, but X is dependant on the lumber used.

I could just use the timbers. The problem with them is that they generally are checked and split in many places. The dimensional lumber is visually graded Select Structural and generally straight. I'll have all winter for them to drop moisture content - then I can manufacture the I Beams.

I have a large hoist available to me at a hanger here. So after I build the beams I'm going to test one or two to see if the calculations I make about Fb and Fv are close...

Like I said, an experiment for me. I was medically discharged from the Army after 20+ years. To shot up/beat up/broke down to keep flying gunships, so now I'm occupying myself with learning something...

Again, thanks! Anybody else with suggestions for where I can get the formulas please chime in.

 

[FredGarvin]

You can calculate the resultant moment of inertia for the cross section, but they are not simply added. If you are really interested in finding the info, look for an engineering statics website that talks about moments of inertia and the parallel axis theorem. The math is easy with these shapes. It really just takes some accounting and keeping your numbers straight.

The other big issue that seems that you have a good handle on is the actual joining of the members. Any variation from a solid shape will be detrimental (says captain obvious).

BTW...Single rotors are for kids.

 

[sixgun]

...and any aircraft that takes off in a mid-air with itself is not for me.

Thanks, I'll try and find a site that covers your recommendation.

 

[TVP45]

Sorry if I sounded too cautious. You get a lot of eager novices on here.
There is a book that shows this all very well:
Parker and Ambrose Simplified Engineering for Architects and Builders. You should be able to find a used copy pretty reasonably. It's a standard and has been for many years. Feel free to e-mail me directly if you need any help with how to set up the formulas.
Tom

 

[sixgun]

Tom, Thank you for your kind reply. We / I used to get a lot of "eager novices" in the attack helo business - and caution was always a wise move...

I will try and find the book you've recommended - actually that will be my wife's job, she's better with e-bay and amazon than I - and once I've given it a good read I'll see what I can do. If I get stuck somewhere I'll email you direct. Thank you for the offer.

 

[sixgun]

Found the Ambrose text and it is very helpful. Seems I have two options:

Calculate the values for the cross section of the full size of the I-Beam, then subtract the values for the space created by the removal of the areas between the flanges, or

Use another formula given in the book involving transferring the moments of inertia for complex shapes, i.e. adding the following to the basic equation: I = Io + Az squared

Anyone have any experience that suggest one method is more reliable / accurate?

Additionally, is there any experience with a factor to apply to the end result that would account for the I-beam being "made-up", that is, not a solid milled piece. ~ Possibly multiply by 0.8 (or some other number) to account for a glued and screwed connection between flange and webs? Would that factor vary based on the width of the web - flange connection: I.E., 2x10 flange is 1.5" surface mated to web material vs. 4x10 with 3.5" surface?

Thanks for any reply.

& remember - Gravity is a cruel mistress...

 

[TVP45]

Glad you found it. It's pretty good.

I always use the parallel axis method Io + Az^2

In steel your question is simple - you just weld and don't worry. For your case, the answer really depends on the glue, that is, how strong (and stiff) does it get? I've never made engineered wood, so I don't have a "rule of thumb", but I'll take a look and see what I can find. In the meantime, you might do an experiment. Make up a 6' long beam with your joints and glue and screws. Calculate the deflection as though it were really one piece. Then load it and measure the deflection to see if you made it.

 

[sixgun]

I'm already making up a 10" beam for testing purposes and I plan to test it with the same 8" bearing surface on each end the beams will have on the solar shed I'm building.

I also plan to build a beam using only screws, no glue, and no routered chanel in the webs. I want test the first beam for deflection at various loadings and cycles of loading. Then I want to load it to failure. I will then load the simple no glue, no routered chanel beam to failure and see what the relative performance is.

Ah, when the snow begins to fall here I will have plenty this winter to keep me busy...

 

[jim mcnamara]

Wish I had seen this earlier...

I-beams 101 from the FPL -which is the good source for elementary wood engineering stuff - IMO.

http://www.fpl.fs.fed.us/documnts/pdf1990/leich90b.pdf


The lack of stiffness of fully cured glue is referred to as creep - heat creep or moisture creep usually. Most decent wood application glues have greater resistance to disruption than the wood itself, so that is not a problem. You need to worry about waterproof, too - consider polyurethane glues.

Epoxy is also wateproof. Generally epoxy glues for an external shed application have a problem with heat creep due to insolation. Bostik 7522 is a high temp epoxy structural adhesive - supposed to have low heat creep, but I've not tried.

Another factor is the quality of the joint. If there are gaps, well, you guessed that part.
Unless you are a competent woodworker consider buying those I-beams at a lumber yard.
Gluing fresh-from-the-mill S-DRY (US grade) and keeping long beam stock straight is also fun. :ugh:

 

[awsaua]

I'm already making up a 10" beam for testing purposes and I plan to test it with the same 8" bearing surface on each end the beams will have on the solar shed I'm building.
Then I want to load it to failure. I will then load the simple no glue, no routered chanel beam to failure and see what the relative performance is.

When you are finished testing can you share the results of your experiments.
thanks jan