compact section, LRFD plastic stress distribution, deflection calculation

[shaker00]

Can someone say why we continue to use the Elastic Modulus, E, when we're doing deflection analysis with compact sections. It seems to me that with p-delta moments, especially on a beam-column, you'd have a modulus less than 29000 ksi because you'd be on the flat portion of the stress-strain curve.

Wouldn't the p-delta moments for a beam-column have deflections greater than what's predicted, and then you would also be underestimating your moment forces? The same would be true for a regular compact beam that's limited to Mp, right? I'm sure it's relatively simple, but what am I missing?

 

[Jerehmy]

Deflections are service loads. You're member won't be yielding for these loads. The design loads you're talking about aren't an everyday occurrence for the member. It maybe happens a handful of times during its lifetime.

 

[shaker00]

Yes, I agree, deflections are most important (and accurate) during service loads. But if we consider wind loads, then strength design is for the 50 year wind event, which may be once in a lifetime. And this design wind may control the overall design of the member, rather than service load deflections.

Thanks for the response, I just haven't come to an internal agreement on elastic modulus for the entire design process, and how this is correct.

 

[Hokie93]

The safety factor, or equivalent safety factor if using strength design, required by building codes is intended to keep the loads actually applied to the structure within the elastic range so using the full value of E for deflection computations is appropriate. Say you have a structural steel beam that is compact and suitably braced such that Mp can be developed. Per the AISC steel specification, that means you have an available capacity that is approximately 10% to 15% greater than capacity based on yielding. Yet, the safety factor on the beam is 1.67 (in terms of AISC steel design using the ASD method). So, the maximum load the beam should ever see based on code-required load combinations is less than the load that would initiate yielding of the cross section. That's not to say the beam couldn't be subject to an extreme overload situation that is outside of the bounds of the building code but I would think in that scenario, if it existed, you would be more concerned about strength than deflection.

 

[BAretired]

We use the elastic modulus, E to calculate service load deflections because service load stresses are less than yield stress unless we have done something wrong in our design.

BA