Beam width: connected vs dispersed

[Elquery]

I am working off the estimation that a simple beam of width w will be roughly half the strength of a beam of width 2w. (linear relationship, unlike beam depth).

I am wondering if the cohesion of that beam matters tremendously (at practical scales), or if its width can be considered in a dispersed manner, given even loading.

Or said differently: Is there any appreciable difference in strength of 1 beam at width 2w, vs two beams at 1w each, given that the load is evenly distributed between the two beams in scenario two.

My understanding is that the resistance to horizontal shear may be greater in the wider single cohesive beam, but in a practical sense (practical, at least, for something like residential wood frame construction) this would be negligible.

 

[PhanthomJay]

I assume you are talking about solid rectangular wood beams, in which case if the beams are placed side by side, there is no difference in total bending and horizontal shear strength from the single beam of twice their width. This would not be the case however if the 2 beams were stacked one over the other.

 

[Baluncore]

Is there any appreciable difference in strength of 1 beam at width 2w, vs two beams at 1w each,

Two narrow parallel beams can twist and buckle where one beam, of twice the width, will not buckle. Bolting or welding two parallel beams together would give the equivalent buckling characteristics of one wide beam.

Plywood and gluelam are stronger than solid timber because the weaknesses and imperfections are not aligned and a local failure is less likely to propagate.

My understanding is that the resistance to horizontal shear may be greater in the wider single cohesive beam, but in a practical sense (practical, at least, for something like residential wood frame construction) this would be negligible.

Timber frame construction is designed not to flex beyond a specified small limit. Maybe 1 part in 1 thousand.
For that reason failure of the structural material is most unlikely.

 

[Elquery]

Thanks! And you've guessed correctly that I'm thinking in terms of rectangular, wood-frame construction.

"Two narrow parallel beams can twist and buckle where one beam, of twice the width, will not buckle. Bolting or welding two parallel beams together would give the equivalent buckling characteristics of one wide beam."

In a highly idealized and theoretical scenario of bending strength, could one assume these twisting forces are not at play, whereas in the real world with imperfect materials and imperfect loading, they indeed come into play?

I am interested in the operation of 'sistering' floor joists (doubling them up) and whether this gives us much advantage over simply decreasing the spacing between joists. (Not considering increasing beam depth as an option in this hypothetical). The relationship between the stiffness of the superstructure (i.e. subfloor/flooring) and the sub-structure (joists) would seem to be the primary determining factor.

If a superstructure is relatively weak, then closer joist spacing would reduce the deflection between joists(waviness); however the deflection parallel to the joists would in turn be slightly increased since point loads would load a singe joist further before transferring load to the surrounding joists (blocking may improve this situation though).

On the other hand, if a superstructure is stronger, the need to support it as often may be reduced... yet at the same time it will be better at transferring point loads to surrounding joists, in which case a dispersed model (more joists evenly spaced) may still see a reduction in deflection parallel to joists (on par?) with doubled joists.

 

[hutchphd]

If a superstructure is relatively weak, then closer joist spacing would reduce the deflection between joists(waviness); however the deflection parallel to the joists would in turn be slightly increased since point loads would load a singe joist further before transferring load to the surrounding joists (blocking may improve this situation though).

On the other hand, if a superstructure is stronger, the need to support it as often may be reduced...

By this (correct) reasoning you should make the superstructure strong enough to support the entire floor and eliminate the beams! I think the practical solution here is to determine the maximum beam spacing allowed by your chosen "superstructure" and then specify the beams accordingly.
Any "sistered" beam will be as good as a single larger one so long as the constituents are absolutely rigidly conjoined (up to the design load requirements...no relative motion anywhere). Does this help?

 

[anorlunda]

I am interested in the operation of 'sistering' floor joists (doubling them up) and whether this gives us much advantage over simply decreasing the spacing between joists.

Sorry. As an abstract question about beams, this thread is OK. But when you cross the line into structural engineering, we can't comment.

For structures, you need building odes, building permits, inspections, and perhaps a sign-off from a professional engineer. Our opinions and your opinions don't count.

Thread closed.