double angles built up section

huangyhg

double angles built up section
when you have a double angle compression member (in a truss top chord) and if the top chord is reinforced with two angles (one on each angle) forming a new section that looks like this:
with this new section, would you consider the top chord as double angles or do you consider it a new built section such as hollow section?
i am debating whether to compute the new axial capacity based on double angles with torsional flexural bucking or just simply euler's type buckling using kl/r?
i can make arguments both ways? what are your thoughts?
by the way, the double angles will be connected at two locations with plate.
regards,
lutfi
the new shape you have shown is not symmetrical about the centroidal y axis. a flexural torsional mode is possible here.  
even if you ignore the angle legs that stick out past the new "tube" section, by the thicknesses represented in the picture, it is still only singly symmetrical. i feel that checking only euler buckling would not be exactly correct here.
will the existing chord be under load when the new angles are installed?
haynewp,
your point is one of my arguments as well to consider torsional flexural buckling..
yes it will be for the dead load. i am requiring that the welding be done by two (2) welders, one on each side and towards one end. the top chord is attached to composite deck with 6 inch thick concrete slab.
i read several article in aisc engineering journal and talked with several fabricators about the effects of the heat from welding. all of them agree that it will not have much effect if done in the sequence that i am asking for. i am also requiring that no live will be allowed during the strengthening process. the weld will be intermittent weld.
if you have other input, please let me know.
regards,
  
lutfi
lufti
just a hard question.
i'm agree with haynewp and you about the stages.
my point of view like yours:
it is a fact that you have two tubular section (if the weld it is properly placed). but regarding the other tube the section it is not simmetrical.
if one tube have to be checked, it has to be like a composite section by a tube and some flanges. it is not simetrical. nevertheless the member have to be checked individually in compresion, taken the distance between stitch fillers for slender.
this tubular section have to be checked by euler and  torsional.
if both tubes have to be checked like a whole section again it is not simetrical, and have to be checked taken the whole   
if the chord is braced against movement in the 'y' direction as you have labeled, then flexural torsional buckling should not occur.
this would then be a design of flexural buckling about the 'y' centroidal axis. if the dead load is not going to be removed prior to reinforcing, then you have to consider the dead load stress in the existing chords when you analyze the new section. but the slab itself will help resist buckling in the 'z' direction between panel points as well, if it is adequately attached.
but would the slab and chord at first try to act compositely to resist the new loads and break the welds due to shear flow? i think the new total joist section itself will be stiff enough for this not to occur, and therefore the slab can be considered to remain acting as a brace, but i am not sure of this.
lutfi:
i wonder if rotating the new angle 180 degree about their x axis so that the legs touch the original ones would solve the problem. one more thing, what is the nature of the inadequacy of the top chord?
since welding will be done in place, the selected configuration would make easier plus i think the proposed section would provide higher compression resistance.
the nature is that with new added loads; the top chord would be overstressed in compression. the new section and angles will carry the additional stress.
regards,
lutfi