Expert: Bruce E. Johnson - 11/20/2008

Question
QUESTION: I am designing a small stick-built home, and want to ensure it is structurally sound.  I would appreciate evaluation of my analyzation, and I have a roof-framing question.

The building is 20' x 56', with a concrete block perimeter and four piers centered down the length.  Stick built gable roof, with 2x6 rafters 16"oc spanning 10' + 2' overhang.  No structural ridge beam, only a 2x ridge board with ceiling joists holding the roof together - two 10' 2x6 joists 16"oc lapped midway atop a bearing wall.

This bearing wall runs the length of the building and is supported down through the foundation across the piers.  There are 3 beams along this wall spanning gaps of roughly 10' and 12' (these may still be adjusted).  I have sized these beams assuming only the ceiling will present a load of 20psf live and 10psf dead.  30psf spanning half the ceiling joists (10ft) is 300 lb/lft.  Assuming a 2x12 supports 161 lb/lft, I calculate a built-up 2-2x12 beam will sufficiently span both the 12' spans and the 10' span.  I plan to hang the ceiling joists flush onto these beams.

I ascertained the 2x12 supports 161 lb/lft over a 12' span from WSDD safe wood beam load tables, using an Fb of 1100.
An Fb of 1100 is obtained via the 2005 NDS Supplement using:
- Reference design value Fb 875 for SPF
- CF size factor for 2x10 of 1.1
- Cr repetitive member factor of 1.15
So Fb'=Fb*CF*Cr=875*1.1*1.15=1106

After sizing these beams, I want to check the point loads at each end of the beams.  However, I wonder if there are additional point loads bearing on the ceiling joists above these beams.  My roof construction references, written by Larry Haun, stipulate the use of sway braces at each end of the ridge board, and every 25ft along its length.  These braces connect the ridge board to the ceiling, and I wonder if this poses additional load on the ceiling?

Googling this issue, I've found such sway braces to be rarely referenced.  Gable space braces for truss systems are popular, but I see little mention of additional bracing along the length of such constructions.  I did stumble upon a reference article for New Zealand building code describing the necessity of what was termed 'roof space bracing.'  It stipulated simular to Mr. Haun (every 7.5m rather than 25ft), but alternately specified spacing per roof-plane area.

Are these sway braces necessary?  Does IRC mention them?  And do they contribute additional load through the ceiling?

Also, the two 12' ceiling beams in this design meet above a single column.  As there is no lateral support for this column along the building's 20' dimension, aside from the ceiling and floor joists, it seems a weak link.  I am constructing in Seismic Design Category A with 100mph wind speed, and have the exterior walls sheathed in sufficient OSB.  Should I be concerned about this?  Should there be any interior walls in this dimension that are OSB sheathed for shear resistance?

Thank you for your time,

Mike

ANSWER: Hi Mike, First of all, the sway bracing may become necessary when using skip sheathing or some other non diaphragm sheathing material.  If you are using 5/8" plywood sheathing nailing it minimum of 6" on center, this will provide the lateral support you need.  Gable end bracing however may be required by local codes so if you need to pull a building permit then you may need to include this in your design.   

Your house is probably long enough for a shear wall somewhere near the middle that is sheathed with plywood.

If you have two posts supporting a beam that has a sheathed floor system attached to it then in most cases the diaphragm of the floor sheathing ( 3/4" tongue and groove plywood) will prevent lateral shear from taking place.

As far as figuring point loads on beams etc.  you need to follow the loads from the top down.  Rafters create a point load on the bearing walls they sit on.  Those walls transfer that load down to whatever that wall bears on.  If it is a beam then the point load would then transfer to ends of the beam and be concentrated at the base of the posts that support that beam.  The base of the columns would then bear on the foundation...etc.  

I hope this information helps feel free to write again regarding this or other matters, sincerely bruce e johnson..bejohnsonconsulting.com

---------- FOLLOW-UP ----------

QUESTION: Hi Bruce, I just looked over my beam calculations and noticed a mistake, so readers can take note..

I used a CF size factor of 1.1 for a 2x10, but I was sizing a built-up beam of 2x12s.  The 2005 NDS Design Values Supplement lists a 2x12 CF of 1.0.  This reduces the Fb:
Fb'=Fb*CF*Cr=875*1.0*1.15=1006 psi
Cross referencing the AFPA's Wood Structural Design Data, a 2x12 spanning 12/0 with Fb=1000 is 146 lb/lft
So 2-2x12 would support 292 lb/lft, just shy of the 300 lb/lft necessary for 30psf loading.

Although the calculations assure me of the design, I may just follow the prescriptive tables within IRC, Tables R502.5(1) and R502.5(2).  I am not a PE, and I hope to have my plans and framing approved by the building inspector without the cost of an engineer's stamp.

In this case, I assume R502.5(2) for interior bearing walls would apply, with a 20' building width, single floor support.  The closest I see is 4-2x12s to span 11-9.  A beam twice the size spanning less distance!  Should I be using R502.5(1), girder supporting roof & ceiling?  3-2x12s spanning 12-2 sounds more reasonable.

Answer
Hi again Mike, over the years I have used the rule of thumb method for sizing residential beams.  For every foot of span I use one inch of beam height in a 4" or double 2x.  So for a ten foot span a minimum of 4x10 should be adequate.  I size my joists similarly one inch of height per foot of joist span.  I have found over the years that this simple formula falls in line with most beam charts etc.  However a lot of engineers add another 10-20% into their structural designs for a safety factor.  So in that respect the 4x10 for a 10 foot span becomes a 4x12 or in some cases a 6x10..I hope this information helps feel free to write again regarding this or other matters, sincerely bruce e johnson..bejohnsonconsulting.com