AISC 341-05 Technical Inquiry - Seismic Provisions 1

[MattLuce]

According to AISC 341-05 section 8.2b, when required by the provisions members shall not have elements that exceed the limiting width-thickness ratios of table I-8-1.

However, I believe it is reasonable and acceptable to use a section that would otherwise be considered seismically non-compact provided the compressive resistance is determined using
(a) an effective area and section modulus calculated using reduced element widths meeting the maximum width-to-thickness ratio requirements of Table I-8-1; or
(b) an effective yield stress determined from the width-to-thickness ratio meeting the requirements of Table I-8-1.

For example: I believe it would be acceptable to use an ASTM A992 W8X24 beam provided that in the seismic design the yield stress is reduced from 50ksi and instead taken as Fy = E*[0.3/(b/t)]^2 = 29000*(0.3/8.12)^2 = 39.58ksi.

I believe that in doing this one has satisfied the intent of the standard mentioned above.

Please note that the reason I believe this is reasonable is because it is not unlike how CAN/CSA-S16-01 Section 13.3.3 treats Class 4 members in compression.

I would appreciate any thoughts on the subject.

 

[JoshPlumSE]

I don't agree. The issue is really related to ductility and hysterisis (sp?) loops. You have to have adequately ductile performance to ensure the energy absorbtion and force reduction implied in the R value used for that frame.

I think the CSA technique is reasonable for typical member design. But, seismic design really is a different animal.

 

[MattLuce]

If I specify A992 (Fy=50ksi) a W8x24 is not seismically compact, but if I specify A36 (Fy=36ksi) then the same beam is seismically compact. Now it just doesn't seem right to tell a contractor or a supplier that they can not substitute a stronger material for a specified beam.

Additionally, ASTM A992 can actually vary from Fy=50-65ksi. If I'm wrong then reasonably one must design for the minimum FY=50ksi, but in checking width-to-thickness ratios one had better use the maximum Fy=65ksi as this would eliminate more beams. Now I'm wondering what AISC used in the beam tables. I'll check tomorrow.

As an engineer I'm used to specifying minimum yield strengths. Do I also have to specify a maximum yield strength to insure we remain seismically compact? How can I even know that a mill did not produce a material with an Fy higher than the minimum specified?

 

[JoshPlumSE]

Fmulce -

It does seem odd that the contractor cannot substitute a stronger material for a specified beam. But, that's the way it is... at least how I understand it.

Think about it this way:
What is the difference between the ASTM's for A992 and A572? The main difference between the two was the upper limit on yield strength for A992. Therefore, calling out this material makes the specification of an upper limit very easy.

Take a look at the commentary for AISC 341:
Although the limiting width-thickness ratios for compact members given in the specification (AISC 360) are sufficient to prevent local buckling before onset of strain-hardening, the available test data suggest that these limits are not adequate for the required inelastic peformance in several of the seismic lateral force resisting systems

Therefore, it appears that one of the main reasons that AISC 341 has the more stringent element slenderness ratios is to account for the difference between minimum yield and the maximum probably stress (considering expected yield and strain hardening).

 

[MattLuce]

Thank you for the feedback. It is greatly appreciated.

One more question:

In the Seismic Design Manual Table 1-2 a W10x26 is considered seismically compact for Fy=50ksi. However, when I calculate the limit-to-thickness ratio for A992 (the upper limit being Fy=65ksi) it is not seismically compact. Same thing happens for W8x21 and many other beams.

Now my coworkers think that these beams are seismically compact. So, it seems to me that everyone around me is placing a cap on Fy=50ksi for checking the limit-to-thickness ratio even though they are using A992. Am I crazy, or isn't everyone already doing what I originally suggested without even realizing it?

Again thanks for the feedback.

 

[JoshPlumSE]

The question is how much of that 50 ksi vs 65 ksi difference is already accounted for when AISC 341 requires "seismically compact" sections? I can't say that I'm truly knowledgeable enough in how these slenderness ratios are defined to answer that question. My guess is most of it is already in there.

If the difference in yield and strain hardening is the main reason behind these slenderness ratios. Then, conceptually it would be nicer if the seismic slenderness ratios were defeined based on sqrt (E/(Ry*Fy*Cpr)). That way, materials with a wide range in the expected yield stress would be more penalized than ones without a wide range.

 

[RFreund]

You should ask AISC, I'd be interested to hear their response. What Josh has said does make sense though.

EIT