CSA S16 - cap channel sections

[canwesteng]

Table 2-8 in the commentary calls out all channels as class 3 - does this hold true for the case of cap channel as well? This seems to be quite punishing compared to AISC for crane runway design. On the one hand, the channel will bend about its strong axis when subject to side thrust, so it seems no different to a regular channel section, but on the other hand, for strong axis bending it seems unreasonable to punish the use of a channel over a symmetric cover plate.

 

[EngDM]

I just ran through this, and I don't believe they are class 3. In my opinion, the cap channels function is actually to increase Iy, which allows the W-shape to keep quite a high Mr value for longer and you get much larger unbraced lengths before the Mr drops off a cliff so they can actually span from bay to bay.

I don't have my copy of S16 at home, but the equation for monosymmetric shapes depends heavily on Iy.

 

[dik]

Cap channels can be Class 2 due to the single axis of symmetry. They cannot be Class 1 due to the lack of two axis symmetry. A discussion a few months back whether welding round BAR in the fillet, on one side, changed a steel W section from Class 1 to Class 2. Because the effect was minimal, I considered the composite section as Class 1 as was the original beam.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[canwesteng]

Struggling to rationalize this. It isn't well stated in the code, but it seems they want class 3 for anything asymmetric to the axis perpendicular to bending, but will allow asymmetry otherwise if it not symmetric parallel to the axis of bending. Can I call it class 2 in one direction and class 3 in the other?

 

[dik]

I don't know why I have a different name...

I would suggest you could, but the main axis is for flexure and the difference between the Class 2 and the Class 3 is that with Class 2, you can use the plastic section modulus rather than the elastic section modulus which gives you an added 15% (approx) strength.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[canwesteng]

Well the main axis for flexure would be called class 2 in this case. For asymmetric sections it's quite a bit more of a bump than 15%, closer to 100% in the case of a capped channel.

 

[EngDM]

For cap channels welded to I beams (in strong axis) aren't you forced to use 13.6 e) anyways, which is for class 1, 2 and 3 shapes? Does it even matter whether we treat it as 1, 2 or 3 in this case?

And then for weak axis the channel can't buckle about it's web since it's welded to the beam? Could you not check flange slenderness off of the combined thickness of beam flange and channel web since one cannot buckle without the other? Or at the very least reduce the h in the section class equation to be clear distance between welds if the channel is muuch deeper than the beam is wide.

 

[canwesteng]

13.6.e.i will let you get to Mp, at least in the 2019 code. I seem to remember that being different in older codes but I could be wrong

 

[EngDM]

It goes to Mp for 2014 and 2009 S16.

 

[canwesteng]

Update - based on some discussions internally and review, I'm going with Mp here and allowing channels to be class 2. Since the effects of asymmetry (torsion from shear center being offset) are included in the calculation I believe this is correct.

 

[dik]

That's what I did... and the EOR accepted using round BAR in the flange-web fillet (it was not possible to weld the channel in place due to the depth. Reinf was on one side only.

thread507-511413

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dik]

Canwest... misunderstood... the 15% is just the increase from Sx to Zx...

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[EngDM]

I put some more thought into this. What if you treated it as "Elements supported along two edges mainly under compressive stress due to flexural bending". Take bel as the distance from edge of wide flange top flange, to inside face of the toe of the channel. t would be channel web thickness. This would be relying on the toe of the channel as a "support" per say, but I'd imagine the web would buckle before the flange bends.

So channel web would be checked as a cover plate for 525/sqrt(Fy), and the overhang would also be checked for that same ratio. The only thing I am fighting with, is that the toe may not support it as well as a beam web, which is what the ratios are based off of.

Out of plane it's just automatically class 3 since it is not symmetric parallel to direction of load.

Other option is to neglect the toe and check it as a cap plate for 170/sqrt(Fy).

 

[canwesteng]

I'm calling it class 2 out of plane, since the effects of asymmetry are considered.