Weld Stress in Built Up I-Beam

[Meshak]

I have an I-beam that was fabricated by welding both flanges onto a web. Flanges and web were fabricated from strip steel. There are two fillet welds joining each flange to the web (one on each side of the web). I want to perform a stress analysis on the welds for my loading. Can anyone refer me to a published stress analysis treatment for this case?

 

[WARose]

I don't get the question....stress from what exactly? Residual (from welding)? Stress from applied load? Or something else?

 

[BridgeSmith]

What you're describing is a fairly typical plate girder. Shear force on the flange to web welds from loading is calculated as VQ/I.

 

[Lomarandil]

Agreed. As far as references, you'd not do any better than to pick up a copy of Blodgett's Design of Welded Structures.

 

[Meshak]

To all,
Thank you for your responses!

Please forgive me, I was trying to keep my original posting brief without TMI.

The beam is on an incline with vertically downward transverse loading. I’ve accounted for both bending stress and axial stress (because of the incline). I’m aware of shear flow (VQ/I) stress at the weld. However, I didn’t know if I should combine bending stress, axial stress, and transverse shear along with it and then resolve that entire combination into principal stresses or maximum shear for either VonMises or Maximum Shear failure theories. Combining all four stresses seems a bit much.

Thank you for the reference: Blodgett’s Design of Welded Structures.

 

[Lomarandil]

Whether to combine bending and axial stress will depend whether your transverse loading is external or internal (inertial). In most cases you'll combine it, as often an external load is applied to one flange but resisted/reacted at the opposite flange.

There's no need to get into shear failure criterion and principal stresses as long as you orient the problem/calculation along the axis of the beam.

 

[Meshak]

My concern is the stress in the weld caused by the external applied loads. My beam is simply supported and I have no external applied shear. My apologies, after posting I realized my use of the term “transverse shear” is misleading and probably incorrect. I was referring to the internal shear(V=dM/dX)evaluated at X=a that puts a section of the beam (0–>a) in static equilibrium.

Incidentally, I just ordered Blodgett’s Design of Welded Structures. Should be here in about a week. Is there a particular section/page that addresses this particular case?

 

[Meshak]

PS,
The bottom flange of my I beam is not attached to anything. The beam will span two walls and its attachment to them I’m considering to be simply supported.

 

[BridgeSmith]

Does this beam have an odd shape that doesn't allow for the use of a rolled beam (wide flange)?

 

[Meshak]

The beams already exist and are being repurposed because of the cost of new replacements. Intuitively I know they’re strong enough for the new application. However, the beams are of a welded construction and I don’t know the previous loads the I-beams were designed for. So I thought it prudent to get an idea of their suitability for the new design loads (SF=?).

The beams are 10” deep. The one-piece flanges are 3.5” wide and are 0.23” thick. The web is 9.54” tall and 3/16” thick. I’m sure weight was a concern in the previous application and consequently drove this non-standard configuration.

 

[TLHS]

It's unclear what you're asking.

That weld has to be designed for a few things:

-Shear flow this ensures tension and compression get into the flanges for composite bending action between the elements

-Transfer of loads between elements at load application points if there isn't a prepared bearing load path or stiffeners that do the same thing. If you have a point load on the beam from above, this needs to get into the web somehow. If you have a bearing point below, the load needs to get out of the web and out of the beam somehow. Usually this is pretty straightforward, but it depends on what kinds of loads are happening

-Sufficient connectivity between elements so that any buckling modes are of the whole section rather than independent elements and that other weird failure mechanisms don't happen. This is usually covered by dimensional minimums in the codes, but you don't want the compression flange buckling in it's weak axis or the web starting to act in bending because you have welds every five feet.

If you're talking about welding stresses induced by the welding, this can be ignored in most conditions.

I'm assuming this is all covered in AISC but haven't ever done a US Plate Girder. In Canada the clause is:

"14.2.3 Fasteners or welds connecting flanges to webs shall be proportioned to resist horizontal shear forces due to bending combined with any loads that are transmitted from the flange to the web other than by direct bearing. Spacing of fasteners or intermittent welds in general shall be in proportion to the intensity of the shear force and shall not exceed the maximum for compression or tension members, as applicable, in accordance with Clause 19"

Clause 19 has a bunch of dimensional minimums and criteria for general built up sections.

 

[centondollar]

There is an additional term to the shear flow for beams with stepped changes in flange inclination. Should be found in e.g., eurocode design guides. VQ/I applies to prismatic members.

PS. Weight does not seem to have been a concern if the beams really are 10 inches (250 mm) deep with roughly 75 mm * 6mmm flanges (flanges compact) and a 5 mm web (h/t = 50, shear buckling checks likely not required and a compact web for bending).

 

[Meshak]

I sincerely appreciate everyone’s input! You’re great!

These I-beams are intended to be the members spanning the tops of two walls of a small building. There will be two purlins (@ L/3 & 2/3L) crossing over the tops of the I-beams which will carry the roof load. While I’m at the analysis, I added a point load (@ L/2) to account for the remote possibility of a hoist. Since the flanges are welded on, support for the point load would be over the top of the I-beam, not hung from the bottom flange like a trolly does.

I’ve already obtained equations from the California building code and calculated the beams are compact and non-slender. From my obviously limited knowledge of structural engineering, I was only envisioning failure of the welds from VQ/I shearing. However as I considered the loads further, I could envision loading in this differential coupon of the weld. As I pointed out earlier, this loading seems excessive. I can envision both shears but don’t know if section shear should be included since VQ/I comes from it.