Heavy timber with turn buckle beam??

[AlmostPE]

I went to this old warehouse that is now a night club. At the roof it has a dropped heavy timber every 16' spanning about 48 ft. The timber has a turn buckle (2" solid) about 26" from the top of the beam. Here is a snap shot. of the blocking holding this turn buckle underneath the beam. Has anyone dealt with this before? How do you analyze the deflection of such thing? They want to put some heavy lighting for the dance floor on this thing. Any help would be appreciated.

Not almost anymore!

 

[whir]

I don't know how they are planning to attach the lighting, but here are the initial concerns that I would have just looking at this. If the lighting is to be attached to the side of the beam, a heavy enough eccentric load could introduce torsion. Any small torsional deflection in the top member will be magnified at the bottom of the blocking where the tension members are located. If that rotates, it will place one of the two members more directly under the beam and consequently cause the load path to be distributed unevenly between the tension members (the one rotated to the center taking more load). Be very careful when you are looking at the capacity of those.

I would rather bolt in a frame made of tubes between the existing members that I could make stiff enough to reduce torsional deflection in the existing structure. You can sell them on it with something about it giving more versatility for lighting locations.

Congrats on the P.E.

 

[AlmostPE]

whir, excellent point. I think they are going to put steel plates that will protect the beam, and then they will loop around a couple of steel braided wires (the plates will protect the beam from the thin wires). It will have motor on it so they can raise or lower the lights. The motor has 2000 lb rating and will operate rather slowly and will not be moving constantly. So each of this beam will have 2 of these connections. I guess I ll just model it with RISA and assume that the tension member is not prestressed. What do you think?

Not almost anymore!

 

[OldPaperMaker]

AboutPE,
This looks like a King-Post truss, turned upside down. The truss rods, in tension, provide the support for the center reaction for the beam. But they also create an axial compression load in the beam. Hopefully, the ends of the turnbuckles are at the center of the cross section of the beam or you will have to deal with the P times eccentricity. There should be a "turnbuckle" so the truss rods can be tensioned or the beam won't have any support under the center post. Of course, the beam also has the lateral load from the roof above so you have to deal with interaction equations. From the stress in the truss rods you can calculate the deflection of the center support post.This will take a few interations to complete.
Depending on the roof loads and the low slope of the truss rods you could have one hell of a load in the truss rods.

 

[msquared48]

What you have there is a spring support with the rods in tension and the wood beam and column in compression. As you mentioned, you should be able to model this in RISA quite easily. I would assume that the rods are on both sides of the beam.

Mike McCann
McCann Engineering

 

[msquared48]

Further - the idea is not new and really no different than strengthening an existing RC beam with exterior post-tensioning. Just different materials.

Mike McCann
McCann Engineering

 

[AlmostPE]

I dont know how they connect the rod at the support because it goes inside the finish wall. I am hoping my model will be good as is. I am just afraid that they put some stress (turn the buckle) on the rod to make it act like a prestressed concret beam or someting. I will let you know my findings once they sign the contract .

Not almost anymore!

 

[AlmostPE]

OK, here is the update. I modeled it and it looks like it does not meet today's code. It deflects 3.6" (48 ft long) in the middle before I even added the load from the lighting. The stress is all fine. What do you suggest? 48 ft is pretty darn long for a timber beam. The client isnt going to be very happy when they hear about this.



Not almost anymore!

 

[msquared48]

Question:

What force would there have to be in the rod for the live load and total load deflections to meet your needs - L/240, L/360? Can the connections handle it?

Mike McCann
McCann Engineering

 

[AlmostPE]

Mike, you and I are thinking the same thing. I am not sure how the rod connects to the beam, but I assume it must have big end plates to transfer the axial of the rod to the beam (become compression). Now if I figure out the force, how do I tell the contractor to tighten it? How do I relate how to turn the turn buckle to achieve the total and live load deflection? How do I know how much of prestress they already put when this was built? This is very stressful. I am sure the turn buckle is painted over and will not be able to turn EASILY!! :-(.


Not almost anymore!

 

[LittleWheels]

A couple of thoughts:
- Apply a specific force to the middle of the tie rod. The resulting mid-span deflection should indicate the amount of preload.
- Tension and mass in a long flexible member is related to fundamental frequency.

 

[msquared48]

Good idea Dave...

Mike McCann
McCann Engineering

 

[AlmostPE]

Ok this is what I am goin to do. Pretty much after you load the extra load from the lighting, it will deflect an extra .425 inches. So I am going to tell them to measure this low point. Then they need to adjust the turn buckle and turn it 1/4 turn at a time (each side back and forward) until you change the deflection to back where it was. What do you think? How hard is it to turn this buckle? Has anyone done this?

Not almost anymore!

 

[PMR06]

You are going to tighten the turnbuckle which will induce a force you cannot calculate thru connections you are unaware of... IMO, not a good idea.

Have you measured the current deflection, under the load you can see and estimate?