Wood Beam Girder

[MKN10]

I have a (7) 2x12 joist girder supporting ground level floor that is overstressed (1400 psi actual - 1060 psi allowable). The joist girder spans 4 bays 13'-2" each long supported on brick piers. The ground level will be converted to a restaurant and the basement will be used for storage. Therefore, the owner opposes the use of additional posts in the basement to reduce the joist girder spans. The other option that could be used is using 1/8 inch steel plates on each side of the beam girder and bolt them together through the (7)2x12 members. Any other ideas?

 

[JAE]

I'd consider replacing the girder with microlams (LVL) lumber. This would require temporary shoring of the floor joists but would allow essentially the same depth (or less) girder.

I assume the floor joists go OVER the supporting girder. If they are face mounted to the side of the girder then that might require another solution as the LVL's won't be as wide as the (7) 2x12 assembly.

 

[SlideRuleEra]

If you go with the steel plates, 1/8" thick plates that are about 11" high may require a large number of bolts keep them vertically aligned under load - this is because they are so thin. Suggest that your consider thicker plates, say 3/16" or 1/4" - this is NOT for bending strength, but for rigidity of the plates.

A way that may allow use of 1/8" thick plates would be to add a 2x12 on each side of the existing 7 members and sandwich the plates on each side between the existing members and the added 2x12s. This sandwich should keep the plates vertical without an excess number of bolts.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[jdgengineer]

What about just stitch nailing an LVL to each side of the existing joists? The LVLs are a little stiffer than sawn (say 1900ksi vs 1600ksi) so you will want to distribute load to the joists based on their stiffness. But 2 LVLs may be enough to get your stresses down. I'd suggest LVL instead of additional sawn to avoid shrinkage issues. Then again, I've never done a built-up joist consisting of 9 members. Sounds excessive so not sure of there are other things to think about.

I'm assuming the 7 joists have staggered splices though out the length? At 52' seems unluckily they would be continuous. Are you assuming any continuity over supports or are you just treating it as "pin-pin"

I'm not a huge fan of plates on the outside for the buckling reasons mentioned above. We do use flitch beams somewhat often, but those have wood on the outside to stabilize. What about channels bolted to the outside of the built up joists? We do this often for residential structures when adding basements below existing floor girders.

 

[Triangled]

1400, hmmm... I don't have my books with me down by the ocean this weekend, but.... Mom Nature used to make some pretty good lumber back in the day, and rumor has it, still does in the Old Woods. 1400 doesn't strike me as excessive absent evidence such as grade stamps, relatively recent construction date, as-built drawings... The In grade testing program for west coast species in 1991 and then for southern pine recently which resulted in significantly reduced values, was as I understand it, at least partly a reflection of and recognition of the growing dominance of farmed lumber.
I'd be reluctant to drill bolt holes through really good lumber, unless of course, one must.
But then again, what is the origin on the 1400 number. Is there a potential for an excessive load in this restaurant from a default exit corridor or a dance floor or..?
Just some thoughts, happy 4th!

 

[XR250]

1400 is little high, but as Triangled said, wood was pretty darn good back then. I usually try to keep old lumber under 1200 psi. Look for ways to reduce the stress in the calcs. For instance, you likely only need about 2" of bearing at each end, so do not use c/c spans. Also, take reductions in the live load if possible based on the square footage supported. I typically lag screw channels instead of plates as they are more efficient per lb, however, as others have mentioned, simply sistering a 2x12 or LVL on each face would likely take of your issue.

 

[MKN10]

Thank you for all that responded. The structure was built back in the 1950s early 1960s. The joist girder is supported on 12x12 pier (over) and splicing is staggered. My concern is if adding plates or C-channels on either side and through bolting will weaken existing joists. Nailing new 2x12 on both sides of existing (7) 2x12 will stiffen the edge members.

 

[Triangled]

Wondering if there is a typical board length and a patterned layout such that you have continuity or cantilevers as opposed to simple spans?

 

[dhengr]

MKN10:
To make any new sistered members effective in sharing the loads and stresses, you have to jack the existing members up; in affect de-stressing (unloading) them. Then apply the new side members, and then let the girder back down. This allows the new fasteners to do some bending and crushing in bearing, so that they do start to transfer some of the loads to the new side members. Otherwise, the existing members start out with their stress, whatever that is, and that stress (and deflection) must get even larger before the new fasteners and side members are brought into play. 7-2x12's is already a lot of plys and you still didn’t say that the floor jsts. are bearing on top of the girder, but I assume they are. From that era, those 2x12's are probably 1.625" wide, so your new side members will not be bearing on a 12" pier. What are you going to do about that? Also, look at the existing splices and consider that at any given location you probably only have 5 or fewer 2x12 making up the actual member, at that splice location. Stretch a string (piano wire) on that girder and study its deflection, maybe load one span and compare its deflection to a calced. deflection. This, and a lack of much permanent set in the girder should give you come confidence in the quality of the lumber used. How has the girder done under the type of loading it has seen over the years? Look for any grade stamps. In that era 1400-1500psi was a pretty common extreme fiber stress for both Doug. Fir and Sou. Pine. Look at some of the new structural screws for fixing the new side members. They are quite effective in shear and in pulling members together, both important in this application. They or bolts or nails shouldn’t appreciably weaken the girder since you shouldn’t be applying them that close to the extreme fiber, of course, they should be staggered.

 

[TDIengineer]

because your existing member is sistered already, maybe the below doesnt help you but worth a look.

http://www.quakewrap.com/pdfs/Structure Magazine.pdf

 

[TDIengineer]

a steel channel on each face works well with through bolts. Getting a long member into the space is often a problem. a bolted moment splice is possible with the channel. a tricky detail, but possible . A welded splice isnt possible with the wood base material.
destressing the existing member is key and engaging the new channels. the steel at each face works well to force a nice clamping action to engage the VQ/I of your composite shape. i am assuming shear is a non issue for the span. the channel has a much better Ix than flat plate, and with the steel Member EI versus your wood beams, channels will suck the load.