Wood-and-Steel Flitch Beam Analysis

[codySTR]

I'm designing a single, bolted flitch beam comprised of a 3/8"x7" A36 steel plate sandwiched between two 2x8 Southern Yellow Pine members (1.5" x 7.25" actual each). Each element is oriented with the long dimension vertical. The overall section will be subjected to strong-axis bending.

This problem seemed extremely straightforward to me when I started out. I thought: I'll assume that the overall section is fully braced against lateral-torsional buckling and I'll used the transformed section properties to determine the maximum bending capacity (i.e. transform the steel into wood and use NDS2015 Fb values; also transform the wood into steel using Fy=36ksi). You find that the section is either controlled by reaching the maximum allowable stress in the wood or the steel (generally wood controls). No problem there.

The design aspect that is stumping me is how to determine the required number of bolts and the required spacing. I found a white-paper/guide published by STRUCTURE Magazine (see here if you're interested). The author's approach is to determine the amount of load seen by the plate based on relative stiffness of the wood and steel, determine the bolt capacity using NDS yield limit equations, and divide the former by the latter to get a maximum bolt spacing. At first thought this seemed reasonable and easy to me. The more I've dug into this, though, this approach seems to neglect some key design considerations. First and foremost, I'm stumped as to whether this overall member must truly satisfy the requirements to make it a composite section. In other words, do you have to consider shear flow (assuming wood is loaded first and the bolts transfer load into the steel plate)? Or are you accounting for shear flow already using the STRUCTURE magazine method? Also, the author doesn't seem to directly account for the unbraced length of the member; he just assumes the section can reach the maximum stress in the controlling member. Is that appropriate? Or do you need to somehow limit your unbraced length to some amount (e.g. distance between bolts? Further than that?)?

I know this type of construction has been around for a long enough time that people have thought of these things, but I can't seem to find a guide. Any pointers?

Thanks!

 

[BAretired]

It is not a composite section. If load is applied to the wood members, the bolts must be capable of resisting the load attributed to the steel plate. This is based on the stiffness of the members.

If the load is applied to wood and steel members, i.e. the top of all members are flush, then the bolts can be nominal as they are not required to transfer any load between members.

BA

 

[codySTR]

BAretired,

Thanks for your reply.

Not a composite section -- got it. So determine the percentage of load that is attributed to the steel plate based on relative stiffness and determine the bolt spacing based on NDS bolted connection capacites? No specific requirements for unbraced lengths?

 

[BAretired]

You assumed the overall section is adequately braced against lateral torsional buckling. The individual members should be braced in accordance with code; this would determine the maximum spacing of bolts.

BA

 

[msquared48]

Remember that the lateral stability issues apply to both the steel and wood members, so both met be braced.

If the wood is already laterally braced, using thru-bolts should brace the steel even if the bolts are not needed for shear as BA mentioned.

Mike McCann, PE, SE (WA)

 

[HouseBoy]

I agree with the others so far however.....regarding this aspect: "If the load is applied to wood and steel members, i.e. the top of all members are flush, then the bolts can be nominal as they are not required to transfer any load between members..." I find that the steel and the wood members are not usually the same height and (more importantly) IF the bearing is on wood (usually the case) then it seems to me that the wood members are bearing the load at that point so.... seems like all of the vertical load must be transferred into the wood (by the bolts). It seems to me that it is hard to do that without several bolts near the bearing ends and that can be a tough situation. I'd be especially leery of 2x8's.

With that said, I've seen plenty of old flitch plate beams with a pair of bolts at each end and staggered bolts over the length. Still, I don't think I've ever designed an flitch beam that didn't require 4 bolts near the bearing ends.

 

[BAretired]

Yes I agree with HouseBoy. It would be difficult to align all members precisely, so better to design the bolts to transfer the load carried by the steel plate to the wood members.

BA

 

[SlideRuleEra]

I'm designing a single, bolted flitch beam comprised of a 3/8"x7" A36 steel plate sandwiched between two 2x8 Southern Yellow Pine members...

I know this type of construction has been around for a long enough time that people have thought of these things, but I can't seem to find a guide. Any pointers?

I believe the reason for the lack of a guide is based on the answer to a key question... How long is a wood beam going to be?
If it is "too long" and heavily loaded the wood will be overloaded in shear at the supports.

Consider the 2x8s mentioned above. Say, maximum reasonable length (even with a flitch plate) has a span/depth ratio of 20. Then maximum length = 12'.

Using the nominal 16" bolt spacing from the Structure Magazine article empirical method, 10 bolts are needed.

Now say that calculations indicate 24" bolt spacing is acceptable. Seven (7) bolts are needed... and they must be accurately spaced (even if there are interferences).

Constructing any flitch beam is going to be relatively expensive. IMHO, in this example the benefit of saving three (3) bolts is not cost effective.
Suggest using the empirical 16" spacing and maintain reasonable flexibility in bolt placement.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[dik]

Can you glue and screw a BAR 1/4 x 1-1/2 to the bottom? and will that provide enough strength?

Dik

 

[tony1851]

Glue and metal fixings do not have the same characteristics and cannot be assumed to act together.

 

[dik]

tony:

based on? There are some excellent adhesives out there... I remember, as a kid, epoxying a small stone to a metal hydro pole 'step' in the fall, and it was there in the spring, through a winter where -40 is not uncommon. I've used adhesives for 'strap/bar' reinforcing numerous times without an issue...

Dik

 

[msquared48]

No different than putting reinforcing steel in concrete...

Mike McCann, PE, SE (WA)

 

[OldBldgGuy]

Since the 70s we've been splicing ends onto timbers (among other things) in heritage buildings where they had rotted in masonry walls, but the balance of the timber was good and disturbing all that original fabric to replace the whole thing was unacceptable. Picture rebar dowels between new & old, for the simplified version. It worked then, works today, and 40 or 50 years of service shows that the repairs are sound. Steel & adhesives can work with wood.
Saying that "Glue and metal fixings do not have the same characteristics and cannot be assumed to act together." is kinda like saying steel & concrete don't have the same characteristics so they can't work together. Not too many engineers would argue that reinforced concrete won't work.