Wide Flange Beam - Web Yielding 1

[SlideRuleEra]

I'm examining an A36 W12x26 that appears to have failed by web yielding. A W6x15 column, 10' long bears directly on flange of the W12. Web of the column and web of the W12 are in the same plane. The column did not buckle. Since the W12 had a pair of 1/4" x 3" full depth stiffeners at the column location, web crippling appears not to be an issue. Assuming "N" equals 6", I came up with 86 kips as the critical concentrated load for the beam web. However I have reason to believe that the load at failure was much higher. Assuming the stiffeners act as an "extra" web, with "N" equal to 6" and "k" equal to zero, come up with another 54 kips of capacity. Brings the total capacity to 140 kips at failure which seems to fit the conditions. All dimensions and calcs by AISC 9th Edition, ASD. Anyone see flaws in my logic (or arithmetic) or know a better way to analyse this situation? Thanks.

 

[PEinc]

Did anyone calculate the actual column load? According to AISC, 9th edition, page 3-32, an A36 W6x15 column with an effective length of 10' has an allowable axial load of 67 kips - well below both the 86 and 140 kip loads you calculated for the W12. It doesn't seem like the problem is with your analysis. It looks like the column was greatly overloaded. But it did not fail.

Could this be a new(er)50 KSI column installed on an older 36 KSI beam? If so, this might explain why the column did not fail but the beam did.

 

[SlideRuleEra]

PEinc - Thanks for you input. The allowable load is 67 kips but the ultimate load (at yield, without buckling) should be about 159 kips (4.43 sq. in. cross-section x 36 KSI).

I did neglect to say that we do no know the applied load. Several dozen of similar W6 columns failed by buckling - when they buckled, the web and stiffeners of the W12 remained more or less intact. A few W6 columns did not buckle but failed anyway by destroying the web & stiffeners of the W12 as described. I'm trying to zero in on the loading that caused this combination of failures.

 

[PEinc]

Until someone figures out what actual loads are being applied, there is no use in trying to check any designs. Sounds like the building is being used for a much heavier application than originally designed. This is an accident waiting to happen!

 

[OzarkMTBr]

Mill material often varies in strength but must fall above 36ksi. Where the W12 failed and not the W6 could be from varying material strength.

Just a possible variable in the puzzle.

 

[ChipB]

Ahhh, puzzles, don't you just love 'em?

If I remember my strengths of materials right, Euler Load(critical)=(pi)^2*(E)*(I)/(L^2)=185Kips (W6x15 @ 10')(AISC manual has factors of safety taken into account)

Are all the W6x15 posts supporting the same thing & framed together above the supporting W12s?

Base plates two bolt or four bolt?
If four bolt, are all in the same orientation (Webs Parallel to beam web? (four bolt in this condition would at least partially "fix" the base) If all are the same, this is moot.

Failure started somewhere. If there were more residual stresses due to the welding the stiffeners at this location than the others, it starts to fail first, the supporting framework above is now deflected, and instead of having a true axial load on the remaining W6x15 posts, you now have induced moment into them, (Ughh the dreaded P-delta), which would cause them to buckle more rapidly.

Just my 2 cents and thoughts, not knowing all the parameters

 

[SlideRuleEra]

OzarkMTBr & ChipB - Appreciate you comments. The same detail, W12 with stiffeners, was used at both ends of the 10' W6 columns. The W6 was securely welded directly to the flange of the W12, no base plates or bolts (maybe not the best practice, but that is what was done). None of the welds failed on either the columns that buckled or those that did not. The "evidence" of this entire failure was in unusually good condition since the total displacement, after the failure, was about 1 foot and the industrial structure did not collapse. For the columns that did not buckle, the only damage was W12s at both ends of the column coming out about 6 inches tall with "crushed" webs & stiffeners.

What we are trying to determine is the force necessary to have "squashed" the W12s at the end of the few columns that did not buckle. That force must have been less than the 159 kip to 185 kip range that would have caused the columns either to yield or buckle.