Check My Girder Spans Please

[vrf]

This split off from an IRC topic.... 
ED: - MD



Is there a similar chart for buildings of less width? My building is only 10' wide, so I'm going to guess that a double 2x10 will be sufficient girder. The length is 14', with three posts per girder, so around a 7' span. Overkill or not beefy enough?

[Don_P]

I Level's free software should be able to handle that. There is not enough information in the post to say for certain, another floor, roof load, species, grade?
Using my simple beam calc http://theownerbuiltcabin.com/calculators/TT/beamcalc.htm
and plugging in conservative strengths, fb1000, E1.2 Fv135, it was good for ~4000 lbs uniformly distributed load.

AWC's WSDD, wood structural design data, has safe beam load tables for solid sawn beams, 2.5" thick and 3.5" thick where yours is built up to 3" thick, they do go down to 5' spans. At a glance their tables interpolate pretty well with my results from above.

[vrf]

Thank you, Don, for your direction. It seems the built beam I've got in mind will be plenty strong, then. And I actually rounded my dimensions up a bit. The footprint of the main rectangle is closer to 9'x13'. No full second floor, but it does have a loft space with shed dormers. (Inspired by this work: http://www.finehomebuilding.com/item/6838/an-old-time-garden-shed ). Posts will be 6x8, down 36".

Thanks!

[Don_P]

That is a 2 floor building in the article at ~1500 lbs load per floor heading down to the girder. we're at 3000 lbs, of the ~4000 lb capacity, now lets talk about the roof, what is the local snow load, how long are the dormers? The trap door in the loft floor would get old pretty quick, there's got to be a better way there.

[vrf]

Yeah, good call on the trap door. I'll probably just have an open hole and ladder at one end of the building. Snow load is Colorado, but roof will be metal. Dormers are about 7' long, 4.5' deep.

Sounds like I should maybe go with double 2x12s or triple 2x10s? Doesn't seem I need it that beefy, considering the other plans I've seen, but my span is a bit longer than most. I thought the narrower building and the deeper 2x10s would make up for that.

Would you ever construct a girder with triple 4x4s? I have a bunch of those I could use. Or maybe the concern there is that they'd topple off each other.

[Don_P]

I'll bite off the fun part first 

Would you ever construct a girder with triple 4x4s? I have a bunch of those I could use. Or maybe the concern there is that they'd topple off each other.

No you wouldn't ever build a beam of any type by stacking lumber or timber with horizontal joints. Glulam beams made in factories with hydraulic presses and special glues are an exception but you can't practically get there in the field.
The problem is not so much that they will topple, I worry about any unbraced deep girder toppling, in this orientation the stacked 4x4's are little better than 3 individual beams each sagging into the other. Hold a phone book out in front of you, bend it into a curve and watch the pages slip past one another, the same thing happens with stacked timbers or beams made of lumber stacked flatways. The individual pieces slip horizontally with respect to one another rather than being firmly pinned together into a deep member that is trying to rotate and bend around greater depth.

One demonstration of the concept of horizontal shear in a beam is shown in my messy shop. Picture 1 is a stack of thin strips of oak supported on each end. I've drawn vertical lines across the stack at intervals along the "beam".


This pic shows the beam somewhat uniformly loaded with lots of deflection. Notice what has happened to the lines


The center has not slipped, the line is still vertically aligned.


Look at the slippage that has happened here, a great deal of horizontal shear


These are the shear and bending moment diagrams for a uniformly loaded simply supported beam, our beam in the example above. In the shear diagram notice that shear is highest at each end and passes through zero shearing stress at midspan. Notice at the point of lowest shearing stress the maximum bending moment is reached, and right where you would expect it on this beam, at midspan.



One of the strength checks when specifying a beam is to see that the horizontal shearing stresses are at a safe level so that the beam does not try to split lengthwise under load. This is usually more of a concern in a short heavily loaded beam. deflection and bending moment are usually what controls the beam size.

[UK4X4]

"Snow load is Colorado"

Thats a very open subject- and can make a huge diference- I'm in CO and have a 90#snow load - 5 miles down the road they have 60#

Contact your local building permit office and check what you have to use.

here at the design stage you have to concider all the enviromental issues- I know designing the building is where the interest is, but the foundation needs time and effort too- not as much fun I know - but it holds the rest up !

Check your soil report to for your area- gives you a basis to work from - my permit office required a soil report and a geotch report.

http://websoilsurvey.nrcs.usda.gov/app/HomePage.htm

The above was not accepted - they wanted a site specific.

Both snow and ground type - plus the building loads its self all take part in deciding the structure and foundation design and size.

Don was extremely helpfull with understanding beams and forces when I was at your stage understanding how all the parts act together made my project much more meaningfull and hopefully better built- I had to use an engineer to sign off on the structure - but I drew and designed the whole thing using knowledge gleamed from this site and its knowledgable inhabitants

[MountainDon]

Just in case the point has not been made clear to someone reading this.... design from the roof down to the earth. It makes complete sense when you think about it.

[vrf]

Thanks, guys, for sharing your wisdom. I'm just going to spend a bit more and build the girder out of 2x12s. Perhaps overkill by a bit, but not a terrifically larger expense on such a small building.

My part of Colorado, near the Springs, is 30# snow load.

Don, this may be an easy question for you. How much is the strength of a beam affected when you notch the posts like this....?

http://0.tqn.com/w/experts/Decks-3464/2009/09/example-H-cut.jpg

(As opposed to face-nailing the 2x12s together.)

[Don_P]

That's not an easy question for me but my mind is also going a different route that makes it moot on this building, I'll explain in a minute. I know it is outside of prescriptive assembly for a girder to be spaced, they call for them to be knitted together. One thing to think about is that with the members acting independently they aren't sharing the load as evenly. I've done something similar in pole barns but also nailing a 2x6 ripped to 2-1/2" wide (2 rips per cutoff)x long enough to fill the gap +  run up alongside the framing member above. Nailed to all 3 members it helps tie the beams and the framing together. You are asking on the net what should really be asked of an engineer, this is outside of prescriptive so answers you get, while they may agree or disagree, ultimately are not worth much. I do agree with what UK4x4 did, plan, then have an engineer look alternative construction over.  I'm not going to say I agree with putting untreated as close to grade as he did in the article but he did put it on a full perimeter of concrete. The questions that should be answered by an engineer, the piers, girders, and possibility of the building toppling over go away.

But let's go up top and look at the roof first, it's pretty neat. If you look at the two ends they are framed with rafters tied to the floor joists forming a closed triangle... a truss. The dormer rafters are not tied across the building and are thus subject to spreading and shoving outward on the walls as that roof takes on a snow load. But, if the ridge of the dormer is a beam the rafters hang from it, as long as it is strong enough to not sag under the load the rafters cannot spread. Now for the fun part. If the truss assemblies at the ends of the dormers are strong enough to do not only their job but also support a ridgebeam across the dormers then that ridgebeam need nod be sized as 14' long, the span is closer to 8' and a well selected 2x10 x14' will work as the ridgebeam. MD's admonition about designing from the top down has a reason that becomes apparent now, I just directed a good bit of roof load to the walls under the 4 rafter feet at the ends of the dormer. As you can imagine the connection of those rafter feet to the floor joists at those locations is going to be pretty highly stressed, nail those heeljoint connections together well. Small structure but a good example.

The dormer roof is ~8x10' or 80 sf x 30 psf snow+10 psf dead load or 3200 lbs. A ridgebeam would carry half of that load or 1600 lbs, each sidewall would carry 800 lbs. Each end of the ridgebeam would bear half of the beam's 1600 lbs , or 800 lbs borne by each dormer end truss, divide that in half and each of the four rafter feet carries 400 lbs of dormer ridge load. Plus it's share of the steep roof load for that rafter pair ~100 lbs and the weight of the dormer cheek wall ~150 lbs. We have a point load of about 650 lbs under each of those 4 truss rafter feet. Look down, we're at midspan in your girder down below... remember he drew it on a full perimeter foundation. Lets see if we can bridge that load back over the posts with the second floor rim, hey it can be a girder too 

The second floor rim could support the floor and the roof load in spans over and between the piers. The tributary area supported would be 1/4 the floor per rim section, or ~5'x7' x (30+10 psf) or ~1400 lbs uniformly distributed plus the 650 lb point load from above, plus the uniform roof load at the ends. A quick rule of thumb, a point load at midspan is the equivalent of twice that load uniformly distributed. We already have a uniform floor load of 1400 lbs and we'll add 1300 lbs uniform load to account for the truss point load. In this case there may be one more 200 lb rafter load, I'll be conservative and spread it out along the rim, and we have 400 lbs of the sidewall load from the dormer rafter bottoms ~ 3300 lbs on a 7' span. If the second floor has a well selected single 2x12 or doubled 2x8 rims with the joists hangered on them and doubled studs over the piers we have just taken the roof and second floor loads and directed them to the piers below. Now the main floor girder need only be sized to support the main floor. Double 2x8 or single 2x10 works.

If you run the 6x6 posts full height to the upper floor and notch single rims of main and upper floors in, the posts are braced by being buried within and cantilevering from the walls.

Edit;
that was a long description above, this is by no means a complete frame drawing but it shows what I was describing. The end trusses support the ridgebeam over the dormer so no ties are needed. They need design. The roof load and 2nd floor load are carried by the inset beam under the second floor. The main floor is carried on its rim beam. I'm not a real fan of pier and beam, this is another way. Lateral bracing of the posts is accomplished by post framing and cantilevering them from the wall framing above.

[vrf]

Hmm... very interesting. I do like the triangular trusses supporting the dormer ridge beam, although I had planned on some light balloon framing to get a bit more head room up in the loft, since the depth is pretty shallow. I went to the contractor's site that built Edrington's garden shed, and you can see that there's maybe 18" of stud height that go upstairs. I'd guess 10' studs, although that doesn't quite seem right.

http://www.marvglover.com/Marv_Glover/Projects/Pages/Garden_Shed.html

I think your idea still works, of course. It just means the rafter/sidewall fasteners are where the strength burden is. Of course the joists nailed into the studs helps a lot.

I feel like I'm hijacking a thread meant for a narrower discussion, but here's another idea: Why not build one tall wall that includes the dormer walls, balloon framing all the way up? That way, the "hinge" on the dormer wall to main mall connection is removed.

[Don_P]

There isn't really a hinge at the dormer floor to wall, braced walls are connected to floor and roof diaphragms. I can't think of a reason not to do it the way you propose if it works better.

I was trying to avoid the kneewalls, certainly incapable of the dormer ridgebeam to rafter loading I'm proposing. By cantilevering the floor joists over the walls and sitting the rafters on the joists, with ply gussets, the triangle, or truss, formed is rigid. That heeljoint is where the stresses accumulate to in a truss. Prescriptively what I've drawn would be an engineered cantilever truss. With kneewalls the shape is 5 sided and not a truss. Kneewalls with untied rafters is not a good design, the unrestrained rafters are pushing outward on the kneewalls. However, keep mentally extending the cantilever I drew, you'll pass code fairly quickly but then in the engineer's realm it can extend out to the point where vertical web members are needed over the support below up to the rafter above, forming framing for a kneewall. At those very steep pitches I would keep exploring the cantilever route withing the codebook limits as you get headroom fast by cantilevering relatively smaller amounts. As small and tall as the building is I'd drop the ceilings to 7'6" if possible.

I have no objection to the mods moving this, we have wandered pretty far afield... again  .