Gambrel truss joint mathmatical solution

[sm69]

I am in the process of designing a shed with a Gambrel roof with little to no webbing utilizing wood plates to attach the joints to maximize my storage space. I have learned that when the joints are not coincident the soulution is not unique. My question is how are these types of trusses analyzed? Are there other ways to get a solution with confidence if indeterminant ? Below is the apporximate shape (diagram is not displayed well)

><
> <
> <
/ \
/ \
/ \
/__________________\

I see these types of truss designs at a local home depot with the joints connected by steel plates with little to no triangulation and would like to use this design, however I would like to learn from it and not just wing the design.

 

[SlideRuleEra]

Gambrel roofs have been discussed a few times, there is almost nothing on the web about them.
...however, in thread507-127734 Eng-Tips member "theclipper" has an interesting observation about the intersections all being on the arc of a circle. I don't know if it is true, but is makes sense.

Also I still have the document "Designing Barns To Withstand Wind Loads" on my website (link below). It will not answer your question but does have some (1930's) info on gambrel roofs.

http://www.SlideRuleEra.net

 

[sm69]

Thanks SlideRuleEra.

I was looking through "Designing Barns To Withstand Wind Loads" and noticed the complexity of the joints. It appears that when designing the joints, pre FEA, each triangle is analyzed separately and then assembled as a whole with many assumptions.

In my design, using only a plate to assemble the joint mimics a triangle with some amount of bending resistance required. This type of design appears to be a truss and a beam combination without knowing what percentage of each it is acting like. The amount of bending resistance needed is what concerns me. Might you have any info on moment transfer of wood joints using nails or nails and glue? I actually joined a couple of 2x6's together to form an 8' long beam with 3/4" wood plates on both sides. I made two samples varying the number of nails but keeping the spacing the same. From memory the nails were 8D common (.162" dia.) and just penetrated all the members of the joint (nailed on one side only). Two rows were created with edge spacing of 1" (vertical spacing of 3.5") and started 2" from the cut. The horizontal nail spacing was 5" O.C. Sample one had 6 nails per board (12 total) and sample two had 4 nails per board (8 total). Sample one held at least 400 lbs acting at approximately the center with no apparent yielding. Sample two on the other hand plastically deflected. Clearly the number of nails increased the moment transfer, however what I still don't know does the horizontal spacing have a major influence on moment transfer? The nails closest to the supports have to be taking a large percentage of the load somewhat like threads do in a bolted joint. I am sure that these nails yield and redistribute load to the other nails. A research document on this subject would surely be helpful in obtaining some sort of nailing schedule. Might you be able to give me some direction?

 

[SlideRuleEra]

sm69 - I do have a few leads that may be helpful:

"Mechanical Connections In Wood Structures"

http://www.fpl.fs.fed.us/documnts/pdf1997/solti97c.pdf

"Wood Mechanical Fasteners"

http://www.fpl.fs.fed.us/documnts/pdf2001/ramme01c.pdf

"Longtime Performance of Trussed Rafters With Different Connection Systems"

http://www.fpl.fs.fed.us/documnts/fplrp/fplrp444.pdf

all from the Forest Products Lab.

http://www.SlideRuleEra.net

 

[sm69]

Thanks SlideRuleEra. Good info