Cover Plate Across Flanges + Bracket

[ToadJones]

Check out the attached sketch.

I have a bracket welded to a cover plate as shown.

other than checking bending in the cover plate, what else should I look at?

Note that the bracket is wider than the column.

Any literature on checking plate bending in a situation like this?

I thought I remember seeing something in Blodgett for this.

 

[dhengr]

Toad:

Do you actually have that bracket design and have to live with it, or are you designing that bracket and have the option to make some changes? Right now you don’t have the bending problem you show in the WF column cover pl., the bracket top flg. prevents that. But, you have an impossible weld condition in the area where the bracket top flg., the column cover pl. and the W10 column flgs. interact. The welds btwn. bracket top flg. and cover pl., and the cover pl. and the column flgs. will be very highly stressed at an unforgiving detail btwn. the three elements, that’s a hard spot.

 

[connectegr]

I have to second dhengr comment. Do you have to live with this connection. The eccentricity is very large. The fillet welds at the toes of the column may be in prying, which is not advised. If a stiffened seat this large is necessary for the shear load, I would suggest connecting to the column web. I see similar details for joist seats, but rarely for beams. I think I have a design example somewhere. I will let you know.

http://www.FerrellEngineering.com

 

[ToadJones]

Yes-
This is existing (I wouldn't do something like this)

I was thinking that since the bracket top flange was wider than the column flange there wouldn't be much bending in the cover-plate.

Good point on the welds dhengr...there will be high forces in the vertical and in the horizontal direction on the cover-plate welds & to make things worse, the cover-plate is welded intermittently to the column.

I was thinking about removing the existing bracket & bringing gusset/bracket plates off the W10 column flanges. The bracket is a W16.

My other option might be removing the column coverplate and putting another widefange (with a narrow flange) into the web of the W10 column and using it as the bracket.

Any additional input is welcome.

Thanks guys

 

[dhengr]

Toad:

From my perspective you’re still keeping to many secrets. This bracket supports a crane girder, therefore a potential fatigue problem? The cover pl., stitch welded, is full height of the W10 col.; or maybe 2' high as part of that awful bracket detail? What is the cover pl. intended to do, the stitch welding is awful in the region of your bracket?

If you have the option, I would remove the cover pl., but pay attention to its intended purpose, and make the bracket WF as deep as I practically could to lesson the T & C forces on the top and bot. flgs. for a given P, loading. I would cope the top and bot. flgs. on the bracket WF to fit btwn. the flgs. on the W10 column, that is {10" - 2(flg. thickness) - 1/16"}; and clip the corners of the flgs. to miss the col. flg./web radius. Now you can stuff the bracket back into the web of the column; and weld the bracket flgs to the col. flgs. and web; and weld the bracket web to the col. web for vert. load shear.

In your plan view, draw a 2" circle at the tips of the col. flgs. In a FEA that area would be flashing red, so hot it would burn a hole in your monitor screen in a few seconds. All of the flg. force is taken to that weld area until that weld area yields (fails, in a fatigue situation), then it will just unzip the bracket flg. weld to the cover pl. The cover pl. isn’t strong enough to take that kind of loading, in that orientation; the bracket flg. is very stiff and the stress at the 2" circle will be very high, and then infinite as the crack propagates to the middle. Connecteng called this same thing prying, and I’m trying to describe it a bit differently. And, you will have a fatigue problem right there in short order; tri-axial stresses, relatively high residual stresses from welding, through pl. tension on the cover pl., not a good thing on thicker pls., or with larger welds. I called this a hard spot, not unlike the beam flg. to col. flg. and web region in our old CJP moment connections, pre Northridge. They failed because of the high stress concentrations and that (hard spot) area’s inability to redistribute any high stresses without a cracking failure.

 

[ToadJones]

dhengr-

you are correct.
This a bracket for a crane girder and fatigue is always an issue.

Someone wanted a crane runway put on columns that weren't intended to carry a crane.
I was going to simply tell the owner "this design is garbage and you need to redesign".
I was not going to waste my time analyzing something that is this poorly done even if some how my numbers said it was "okay".

An old timer told me once "design what you know".

I dont care for brackets for cranes at all.
I if I can add another column into the web of the W10, I can essentially design a step column of VERY deep bracket.

I am assuming the coverplate was added because the column was not oriented strong-axis for the bracket load.
Good idea in theory....not carried out well.

 

[connectegr]

Sorry, no luck finding an design example of your condition.

I would suggest using horizontal stiffeners at the bracket flanges and eliminating the cover plate.

http://www.FerrellEngineering.com

 

[CTW]

What did you end up doing? I have a similar situation except this bracket is a tee with the flange welded to the tips of the column flanges.

 

[ToadJones]

CTW-
I'm not sure I follow you.

I haven't talked to my client yet.
I am leaning towards suggesting that they remove the cover plate in the area of the bracket and cope a new bracket to fit into the web of the column and do a 3-sided C-shape weld on the top of the bracket and possibly something different on the bottom to make the welding easier.
Then they can place the over plate back on the column and run it right to the bracket above and below.

Can you post a sketch?

Connect- you out there? My idea ok?
I think it is what you suggested.

Of course my numbers will have to check out.

 

[connectegr]

ToadJones -
I definitely like your proposed alternative better. Much better moment transfer to the column.

CTW -
Why not connect directly to the column web. And extended plate with top and bottom plates for stability or moment, if necessary.

http://www.FerrellEngineering.com

 

[CTW]

ToadJones-
See attached. The centerline of the column is also the centerline of the runway

 

[ToadJones]

My problem now is, the column fails without the cover plate.

Even if I replace the cover plate right up to the bracket above & below, there is still technically no cover plate at the bracket.
Not sure what to think of this.
I supposed I could have them weld the cover plate to the bracket also.

 

[ToadJones]

CTW-
How are you girders connected?
Are they shear connections to the WT!?

 

[azcats]

Can you get a plate in line or on top of the bracket flange that is coped to fill the 'void' and drag load into the column flanges and web?

 

[connectegr]

ToadJones -
Is the entire column cover plated? I have had a similar condition, and was forced to put the connection material on the cover plate. In my condition I had to CJP with cover plate locally for the connection forces. With you wide-flange bracket the forces will be concentrated at the bracket flanges into the cover plate, then to the column. The forces are not evenly distributed for the entire height of the bracket.

CTW
I have attached a sketch of my recommendation. I am not a fan of welding connections at the toes of the column flanges. Some cases cannot be avoided. For ToadJones, with hindsight a heavier column might have been more economical.

http://www.FerrellEngineering.com

 

[ToadJones]

yes...the column is cover plated from about 6" above fin. floor to just below roof.

I did not design this.
Existing structure some shop guys decided to toss a runway on and ask about it after the fact.

 

[CTW]

I don't want to further hijack ToadJones' thread so I've started a new one to explain my situation plus to have a more accurate thread title. I didn't intend to get into specifics on my project, but since I'm getting some inquiries I'll post some details. It may help someone else who runs into this oddball situation.

All new posts related to the connection detail I attached previously should be posted to the following thread.
thread507-310301