through bolting in thin wall HSS sections 1

[tstructural]

See attached for tube section

I'm working on a design that requires a lifting lug plate attached to an HSS 12x2x1/8 tube. The lifting lug plate is an angle that is secured to the side of the tube with (2) 5/8" diameter thru bolts. The first bolt is located just 2" above the bottom of the tube and then another bolt is 3" above that. The tube is the base of a large AHU. While adding the lifting plate to the mechanical unit base they yielded the tube wall. I'm still trying to find out what the specified torque value was. They said either 75 in-lbs or 75 ft-lbs - I think the 75 ft-lbs makes more sense but if that is true then the tension in the bolt just with torque and no outside applied forces is 7.2 kips. I believe that the 7.2 kip tension in the bolt is resisted with compression of equal value between the nut/washer plate and tube wall. When I analyze the tube wall as a flat plate with a 7.2 kip force at the 5" location the 'flat plate bending moment' of the tube wall is 21 kip inches. But, ASD quick check shows that a 1/8" plate that is 6" wide and 12" long is only good for .656 kip inches (that is using 100% yield strength). The FY is 42 ksi. It seems the tube wall should have yielded long before they completed their torque.

The problem - according to the shop they have used this connection on HSS 10x2x1/8 millions of times in the past with no problem. I know the 10" tube will allow a larger force to be applied but not enough to account for the extremely large discrepancy I'm seeing between my analysis and the actual results in the shop.

My question - Am I analyzing this correctly?

We are going to have them add a steel plate (3/8" or 1/2") between the tube wall and the angle to stiffen that section. Then those torque forces and the applied torsion when the lifting lugs are being used will be able to distribute to the top and bottom flange of the HSS section. The problem is that I don't know how to design this plate. If I'm correct that the torque force alone is 7.2 kips then I'm going to end up with a 1" thick plate or more.

 

[KootK]

Your analysis makes sense to me. Effective plate width is always a tricky thing. Some things to consider:

1) Could you avoid pretensioning your through bolts?
2) Could you use blind bolts that would engage only one HSS wall?
3) You'll either have to extend the inner stiffener plate full height or do a different analysis for that side.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[SAIL3]

It does not make much sense to pre-tension bolts on such a thin HSS section and not a good idea to pre-tension thru-bolts on HSS sections in general without addressing the overtightening problem...one could instead use double nuts or lock nuts....

 

[racookpe1978]

You have two 5/8 dia through bolts, the lower one very close to the steel tubesteel wall. The seond near the middle of the wall.

Yes, absolutely determine the correc torque. But it doesn't matter: I'll bet you dollars to donuts that the idiot who put his entire 250 lbs weight on the end of a two foot long cresent wrench didn't know what he was applying. And problably didn't care either --- until the tube steel wall collapsed.

But he will do it again. Or be scared up and not tighten the next 240 enough.

Drill your pilot holes for the two 5/8 bolts. (Or mark them, if you're going to use a mag-based hole saw. Drill the lower hole as you do now - it doesn't seem to be giving you problems.

Drill the middle hole to fit a tube 3/4 ID and at least 1 inch OD on one side, and 5/8 dia on the other. Length of pipe to be 1/8 less than tube wall distance: you want a slight recess on the welded side. Slide your 5/8 bolt through, then the 3/4 IS tube over the bolt so it touches the far side, and is in the larger hole on the near side. Fillet weld the near side.

The short round tube will brace the rectangular tube steel walls from collapse, and will be much cheaper than an internal "wall" or plate.

 

[jayrod12]

^^ that is a great solution racooke

 

[KootK]

X2. That's going in the KootK vault. Curious though: is there any reason that it couldn't be done without the welding?

Detail coolness aside, simply not pre-tensioning must be the cheapest way to go, right?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[tstructural]

Thanks to all above for your help.

 

[RWW0002]

It should be noted that the HSS connection manual recommends that thru-bolts be installed to a snug-tight condition only (for this exact reason)

As far as the numerical discrepancy between the calculated plate bending force and the actual applied load, you might want to take a look at AISC 360 Chapter K and AISC design guide 24 for concentrated loads applied at the face of HSS sections. Effective plate width is not the only thing at play here, as the sides of the tube tend to stiffen the tube face(a simple-span plate similar to the above example is often too conservative).

 

[racookpe1978]

X2. That's going in the KootK vault. Curious though: is there any reason that it couldn't be done without the welding?

Detail coolness aside, simply not pre-tensioning must be the cheapest way to go, right?

If the 5/8 bolt will be loaded in shear through the tubesteel, it can be done without welding: But you have to be very careful with the final length of the 1 inch OD x 3/4 ID tube so the compression of the outside bolt tension is carried trhough the two outside plates to the tubesteel wall to the 1 inch tube to the far tubesteel wall to the other side lifting clamp.

Assume the 5/8 bolt is loaded in tension between the two lifting lugs on both sides. Cut the 1 inch tube about 1/32 less than the outside wall -> outside wall width of the tubesteel. When the 5/8 bolt is torqued down, the two outside lugs are pulled together compressing the tube steel walls by that 1/32 (0.030 or thereabouts) but this locks the 1 inch tube in place by friction and compression, and limiting the "bow" of the tubesteel walls to 1/32 over the whole side.

By the way, welding the 1 inch tube on both sides increases stress on the tube steel walls because the weld shrinks as it cools, and so will pull the two tubesteel walls into tension through the 1 inch tube.

 

[dhengr]

Tstructural:
That’s not a very good detail however you cut it, analyze it, or whoever designed it. Put some more dimensions on your sketch, along with loads and their directions. What are the 4" high cross members btwn. the side rails, and how are they attached to the side rails, and how can they be brought into play. Show a plan and side views of the unit base frame with all framing members. What does the AHU weigh and does it cause the same lifting loads at all four lifting lugs? I assume that the angle lifting lug was also intended to act as a bearing support (base pls.) and hold down point at each corner of the frame. Can the frame bear on the side rails near the 4 corners? Racookpe1978's suggestion is a good solution to part of the problem, but I think you should redesign that detail if you have that option. The lifting and base pl. loads should not be put into the side rails the way your detail shows. The bearing areas btwn. the bolts and the 1/8" side walls of the 2x12 HSS are just too small for good permanent support.

 

[kingnero]

Would the following detail be helpful in your situation?

 

[SAIL3]

agree with dhengr....it is a soft mushy conn, especially for a lifting lug which is a one-path load...what strikes me is the lack of rotational restraint for the lift lug based on the info supplied sofar...