Design of variable web depth bridge girder (using AASHTO 6.10.1.4) 1

[lunera]

I'm designing a deck plate girder for a railroad. We are keeping the existing piers and so I am minimizing the pier work by decreasing the depth of the girder at the supports. I've referred to AASHTO 6.10.1.4 "Variable Web Depth Members" for guidance and have calculated the Pv, Ph, fn that it describes.

I am trying to figure out what is the proper way to include the additional forces from the variable web. In the commentary it states, "At points where an inclined flange becomes horizontal, the vertical component of the inclined flange force is transferred back into the web as a concentrated load. This concentrated load causes additional stress in the web and web-to-bottom flange welds, and will often require additional local stiffening" (AASHTO C6.10.1.4) It then goes on to describe how to determine the required stiffening.

I am not so much concerned about adding stiffeners right now as I am making sure that the web is not overstressed at the point where the inclined flange becomes horizontal. I wonder how do I account for this concentrated load? Do I add it to the vertical force in the web at that point and make sure that my stresses aren't too high? Or, is the vertical force just an increased version of my dead load forces, so I take out the original dead load and add in the higher one (instead of the force being just a separate additional load)?

For example, right now the critical section is where the girder becomes horizontal on the shortest section. The Ph=1910kN, the fn=43MPa, and the Pv=669kN. Is my total shear in the critical section equal to DL+LL+Pv or is the Pv just an increased DL, such that my new total load is just LL+Pv?

I've attached a snapshot of the drawing so that you can see the geometry of the beam.

 

[Teguci]

Shear is removed from the web at the lower bend through compression and is added back to the web at the reentrant bend through tension. Design the connection between the bottom flange and the web to take 35% of the bottom flange tension force (based on geometry provided).

 

[lunera]

Teguci - Thank you for your reply. What you've explained makes sense. Do you have an example or a resource that I can refer to go over this in more detail?

 

[hokie66]

Just use a truss analogy. The flanges are the truss chords, and you need a web member to take the resultant at the offset.

 

[Teguci]

As Hokie said, use a truss analogy (thats what I did). I imagine in practice that a double stiffener (with associated weld to bottom flange) will be needed at the reentrant bend to put the shear back in the web. This is a bit of a departure from the adage "don't weld to the bottom flange" for crane girders, but I imagine the concern can be limited by only providing partial height stiffeners at the bend.