WT seat (for runway girder) welded to wide flange column tips

[CTW]

This subject is similar to thread507-309829. I was starting to get questions specific to the connection detail I posted on the aforementioned thread and decided I should dedicate my situation to a new thread before the other one was further hijacked.

Now to explain the connection shown in the attachment.

The seat is a 2'-0" long WT7x45 and the column is a W8x48. The WT has a 1" bearing plate. It was designed and constructed as part of a crane runway capacity increase. The original crane was 20-tons and is still on the runway. The new crane added to the runway is 25-tons. Both cranes are class E.

The original runway girders were connected to the web of the columns with double angle bolted connections. The design for the capacity increase resulted in one end of each runway girder being supported by the WT seat and the other end supported by a new column. The WT seat has its flange fillet welded to each column flange tip. The runway girders are still connected to the column webs with double angles.

I'm currently evaluating the capacity of the structure after finding numerous deficiencies during an inspection. The deficiencies were mostly related to lack of maintenance and poor construction of the prior capacity increase. I'm investigating removing the double angle connection of the runway girder to the column web. I've just started evaluating the capacity of the seat. Ideally, I'd like to come up with a way to get rid of the seat and double angle connection and have the girders bear on the existing columns.

To throw another variable into the equation, the X-bracing was removed and re-connected to the WT seat as specified in the design for the capacity upgrade.

 

[dhengr]

CTW:

That’s still a crappy detail, for the same reasons, connection objections, Connecteng and I had to Toad’s detail. So, do reread that other thread as regards flg. tip to flg. tip welding and stress concentration problems. It does appear that you may have the WT flg. bending problem which Toad first sketched. And, you may also have the prying, on the weld root, problem which Connecteng brought up. But, you may have a few things which make your detail more easily rationalized from the design standpoint. How is the WT loaded, load magnitude, and dimensions, etc.? Add about 98% of your verbiage to the start of a sketch, which you brought over. Show the girder, didn’t you say it’s centered on the column? Where is the 1" WT bearing plate, and its connections and welding, and the double angles, and the x-bracing? You gotta remember, you can see it, we can’t without a better sketch. And, imagine all of the possible sketches that your verbal description might prompt, none of which match what you have or are talking about.

 

[CTW]

I'm out of the office tomorrow so it will probably be Monday before I can get another sketch uploaded.

 

[CTW]

Attached is a sketch which hopefully clarifies my description.
The WT is loaded in shear with an approximate magnitude of 130 kips.

 

[ToadJones]

I wouldn't do it for a 130 kip reaction.
For 130 kip reaction, you must have a decent sized crane.

I know that AIST Tech Report 13 (Mill Buildings) recommends that brackets not be used for any reaction over 50 kips.

I cant beleive that the existing connection for the runway girder was double clip angles...really bad.

Watch how you weld reinforcing to the bottom flange. You will have to investigate fatigue in those welds. If you use intermittent welds, the existing W24 bottom flange will crack...just a matter of time (unless this is a very low cycle crane)

 

[ToadJones]

Also,
your new W8x31 would be better rotated 90º.
Crane columns should be designed for the eccentric reaction induced by the girder deflection and rotation at the seat. As the girder deflects the reaction is pulled toward the center of the span. In this case your case will see a weak axis moment somewhere in the neighborhood of bf/2 x 130 kip or 8"/2 x 130k = 520 k-in.

 

[dhengr]

CTW:

Undoubtedly someone is trying to accomplish something akin to a stiffened seat for the girder. The problem I have with your WT with its outstanding stem is the moment and moment arm which works on the welds out at the flg. tips of the W8 col. Also, in Toad’s previous thread he started out asking about (actually sketched the condition) what amounts to flg. bending on your WT, and this causes prying on the welds out at the flg. tips, which Connecteng pointed out as a possibility. The shear loading of 130k sounds doable, thus 130k/(2 x 24") = 2.7k/inch, a nice uniform shear flow on the fillet welds. On the contrary, any moment type loading on the WT imparts a large T & C force at the top & bot. of the WT on a lever arm of about 22" = (24 - 2"). And these are input to the welds in a very concentrated way, over a few inches t&b, and a prying way to boot. Now you have at least a biaxial combined stress fatigue problem. The tension forces at the top of the WT will just unzip the welds over time. I would go a step further than Toad did here and say that you have a moment on the WT of about 130k(6 or 7") = 910"k, thus T & C = 38k = (910/24") and a moment on the W8 col. of 130k(6 or 7" + 4") = 1430"k. And, I would agree with him that there would be some advantage to turning that col. by 90° if you could. Also, over and above the girder loading on the WT it seems that you have an additive bracing force to deal with on the welds. Toad and the AIST have daily experience with cranes and crane girder connections which I don’t, so I would be interested in knowing why brackets are frowned upon if done properly. Certainly, a column directly under the girder is a cleaner load path, but again, you are trying to make an existing condition work.

It does seem that you have resolved most of my concern about unzipping the welds with a top plate on the WT, as part of your girder bearing detail. This allows you to get the moment components back into the flgs. and web of the 8W col. with welds. I would put a similar pl. into the col. at the bot. of the WT for the same reasons, although somewhat less critical from the fatigue standpoint. Obviously, you have to do the analysis, and design the columns, connections, and welds, but keep in mind that you have a fairly fatigue sensitive situation here so try to make those load paths, weld lines and details as direct and clean as possible. Any intermittent welds in tension, in a fatigue environment, are a potential problem, all of the starts and stops are stress raisers. Alternatively, most fab shops will run a lighter continuous fillet (same total strength) for less cost and less welding consumables used; given the savings in layout time, over welding a larger and longer weld to be sure they’ve got enough, welding time, and inspection time, etc. And, the continuous weld looks better, is easier to do, and is less fatigue sensitive.

On second thought, you haven’t resolved my concern, you can’t get in there to weld that top bearing pl. on the WT unless you apply it first to the W8 col.; then apply the WT to it and the W8 col,; then apply the bot. pl. to the WT and the W8 col., and this is all out of position field welding. It might be better if you fabed a C shaped bearing bracket; with 24" high side pls. which welded to the outside faces of the W8 col., 4 or 5" t&b and 24" high; then put what Toad called a cover pl. (like your WT flg., but no stem) shop welded to the two side pls., along with the lower top 1"x7'x22" brg. pl.; the side pls. might be 24" high x about 8" wide (4 or 5" on the W8 col. flg. + 2 or 3" to pick up brg. area); and this whole C bracket with lower top brg. pl. would be shop fabed and then welded to the W8 col. flgs. You might have to move the two bolt holes left an inch or two. The moment is still there but it’s put into the W8 col. with much less stress concentration and fatigue potential.

While the double angle bolted connection was probably flexible enough for the original girder connection condition, your new W8 reinf. col. introduces a slightly new problem that you may want to give a little thought to. And that is, that when the girder on the right deflects it causes a small displacement but a potentially very high shearing force on the double angles; due to that girder’s rotation over the new col. Can you (your double angles) tolerate that?

 

[ToadJones]

I am at a loss as to why the bearing plates (or whatever they are) are so wide? I'm referring to the 22" wide plates.

For crane loading (moderate to heavy), you should never hang the girders with a double clip shear connection. Over time as the connection rotates over and over again, cracks will begin to propagate from the upper most holes in the girder web. Usually the crack shoots down the web a few feet and renders everything above that crack useless for your shear loading. Also, there is repeated prying forces on the clip angles and they will crack too.

With regard to bending moments in the columns, I was only referring to the W8x31 column. I agree that the bending on the other column will be worse.

As for AIST, the 50 kip shear reaction is a recommendation for new design (As far as I know)...ie. "my shear is 60 kips, it's time to use a column instead of a bracket".

I have analyzed many a bracket that had reactions higher than 50 kips.

This design looks like one that will have problems down the road.

 

[CTW]

I just noticed that the line types in the sketch I attached did not translate to PDF very well. Everything shown in the sketch is existing.

The original structure consisted of the W24 runway girders, double angle bolted connections, W8x48 column and X-bracing. The original structure is approximately 30 years old.

The WT seat, W12 reinforcing girder and W8x31 column were designed as part of an upgrade approximately 10 years ago.

At the moment, I'm trying to determine the rationale someone would use when designing the WT seat. My opinion is that the intent was to take out the prying effect on the clip angles. But why not just add another column like they did on the other side of the W8x48 column? Why even risk using a detail that puts tension on vertical lines of weld in a fatigue situation that has a good chance of unzipping? Why add a W8x31 column and not a W8x48? The depth of the two W8’s is different so the tie plates are not flush with the flanges.

dhengr:

I agree with your comment that someone was “trying to accomplish something akin to a stiffened seat for the girder”. But I just can’t understand how or why someone would use that particular seat detail for a crane bracket along the weak axis of the column and not the c-shaped plates you described or another column.

ToadJones:

My guess is that the 22” wide plates are that long to allow the plate to be welded to the bottom flange of the existing girder and then bolted to the plate on the seat. I’m not sure why they didn’t use field drilled holes in the girder bottom flange though. Had to be some reason for the engineer to do it this way but the detail exists so I have to deal with it.

 

[ToadJones]

Well, IMO the original design was poor and the upgrade ever worse.

Are you trying to further increase capacity here?

 

[CTW]

No further capacity increase required. I'm only determining the capacity of the structure.