Help Welded Structural Tee With SHARP Flange to Stem Joint

[JBR87]

Hi guys,

I have a structural steel construction project I'm working on and need some input/ Ideas. There is a section of this project that calls for a T-section to be manufactured using two 3/4, or one 1 inch and one 3/4 inch plates. The only problem being that these members are exposed and the architect wants the flange-stem interface to have "a hard 90 degree angle" (no visible signs of the welds). I am currently struggling to think of a way to accomplish this. The sketch sk-1 (attached) is what the drawings call for.

One possible solution although very costly I thought of is in sketch 2 (attached). In this situation the flange would be cut in two and sandwich the stem. The 2 pieces of the flange would in turn be welded to the stem on the back side (non-visible) of the tee via a groove weld. Although I think this might work the fab cost would be very high and thus undesirable.

Another solution that a colleague of mine had suggested is to use a stitched square weld as shown in sk-3. This if possible would allow us to use one plate for the flange

Any help or insight would be appreciated.

Thanks,
Jeff

 

[dhengr]

JBR87:

What is the Tee used for, how is it loaded, how is it stressed, etc.? You have to know the horizontal shear stress and shear flow at the welds to design them, and you do not want to have a prying action across the weld roots, that is a tension stress perpendicular to the weld and across the root. I would opt for Sk. 3, and push the stem up into the two flange halves .125" or .1875" and then just apply fillet welds btwn. the stem and each of the flange plates. No need for complete fill of that square groove. They will penetrate about .125" at the roots of the two fillets, almost to the underside of the flange plates. You will have to pay some attention to distortion in fabing this section, but a good fabricator should be able to account for most of that, or straighten the finished member. I prefer a smaller continuous fillet weld, properly designed, to the layout and starts and stops of stitch welding, although the stitch welds will settle the distortion problem down.

 

[hokie66]

Architects! Does the client know how much of his money this moron is wasting?

 

[Teguci]

Use a standard tee shape and have the architect box out the exposed area in sheet steel.

Main issues are:
1. The local shear stress at a 90 degree corner will be significantly high for a well loaded beam (why did you think they had fillets there?)
2. Fabrication as your post is getting at.
3. Cost. Your cost is the steel member. The architects cost is the sheet steel. Makes it simple when the owner hears about this and value engineers.
4. Variability. What happens when you have to upsize the beam?

I'm with Hokie on this one. Tell the architect to stick with golden columns and leave the real work to us.

 

[JBR87]

The tees are being used as window stringers so the loading is not our main concern and I higly doubt they will ever be upsized (I guess I should have mentionned that).

I work for a fabricator (I deal mostly with connections) and we have never dealt with any requests like this before. In a perfect world we would have and wanted to use a standard T section.

A couple of test sections are being fabbed right now using sketch 3 with slots in the flange to weld to the backside of the stem. We're going to go this route to save having two plates for the flange, hopefully the distortion can also be kept to a minimum this way.

The building is ironically enough an engineering building at a university.

 

[Teguci]

Have you considered cutting the fillet out of a standard tee?