welding plate to existing column
 

welding plate to existing column
i have a wwf350x176 beam-column that requires strengthening. i will be adding a plate along the entire length of the column about its strong axis, this increase in stiffness was verified to be adequate. the plate will be 25 mm shorter than the width of the flange to allow room for a fillet weld on each side. i also have to specify the required size of the fillet weld, but remain unsure as to what force the fillet weld be resisting. would really appreciate any hints & tips on the matter.

are you adding these plates to increase the column area or radius of gyration?
if there is no moment to consider, at the top and bottom i'd add enough weld to develop the compression strength of the plate. in between, i'd add stitch welds such that the kl/r between the welds is less than the kl/r of the column.
if there is moment, there's shear flow to calculate, which i forget how to do and i don't have my books here at home.
you have not mentioned the thickness of the added plate, but this is a critical factor in selecting fillet weld size. see page two of this paper
watch out for "nothin' welds"fff"> at this link
there are many things to consider:
welds to limit kl/r of reinforcing
shear flow for bending using vq/i for intermittent welds and mq/i for end welds
welds to limit width-thickness ratios of reinforcing
end welds to allow reinforcing to carry its proportion of axial load
special welds for corrosion, fatigue, etc.
jike, what the heck does mq/i give?
you can check roark (page 95 in the 6th edition), which gives a shear stress at the interface as:
v*a'*z'
-------
i*b
where v is a vertical shear at the section, a' is a cross sectional area above the interface (or below if you are adding plate below the bottom flange), z' is a distance between the centroid of a' and the neutreal axis, i is the second moment of inertia and b is the breadth of the section at the interface. this shear will need to be taken by the weld in order to ensure that the added plates act compositely with the original section.
whenever a section is reinforced for bending, the stresses in the main section have to be redistributed throughout the total reinforced section. the end welds allow these stresses to be transfered.
mq/i is simply the average stress in the reinforcing times the area of the reinforcing. there is an explanation of this in the commentary of the green aisc manual.
bkal-
that formula you give in roark is the same as vq/(it).
jike-
where in the commentary of the green book is the discussion of mq/i?
a few things to think about:
a narrower plate might reduce the weak-axis radius of gyration. the column may have a lot of stress in it already. the welds to the thick flanges may need special preheat.
blodgett has a nice section on cover plating.
structuraleit: i am not currently at the office but i will look up the page number when i go back to the offie on tuesday.
as an alternate to mq/i for bending, i would probably just take 0.6 x fy times the area of the reinforcing for end development. this should work for both axial and bending.