Brockjd:
I would use one pl. btwn. the two inner col. flgs., at their center lines. I would measure the distance btwn. the flgs. and cut the pls. 1/8" +/- smaller. This pl. would be on the order of the thickness of the exiting col. webs. Then, I would look at this as if my new pls. were a web of a built-up member, stitched to two flgs., which happen to be the existing col. sections. If these new pls. are 2' long and spaced 3' on center, do I have enough double sided fillet welds to make this built-up member work. It’s a funny pl. girder, right? You could add light t&b stiff. pls. to the free edges of your new web pls., stopping short of the dist. btwn. the inner flgs. Alternatively, you could use two pls. at each 3' space level. These pls. would span onto each existing flg. tip, abt. .25" - and leave enough flg. tip to allow them to be fillet welded to the flg. tips properly. The free edges could be stiffened with 2x2 angles with the horiz. leg pointing in. These angles would be a bit short of the dist. btwn. the inner flgs.
As for the analysis of these shear pls. (web pls.?), I’d look at Roark’s Formulas book, several of Timoshenko’s books or several other good Strength of Materials or Theory of Elasticity text books.