Cracked and Uncracked Section Properties In 3D Building Design Software Analysis 2

[struggle67]

Hi

I am using 3D Tekla structural designer Software (TSD) to design buildings. I noticed that for concrete members TSD assumes cracked sections (Icr) for the analysis and by default Icr = 0.2 Ig. For uncracked sections TSD uses 0.4 Ig to account for long term effect especially creep. I think both Eurocode and BS code recommend to use uncracked cross-section for elastic analysis. If there is no seismic load, should I still use cracked section?
Does other 3D commercial software do the same and why are they not following the code recommendation?
Thank you!

 

[Settingsun]

I'm not very familiar with either code, but would guess that gross section properties are given as an option for analysis provided cracking and creep are accounted for some other way, such as moment magnification based on cracked properties.

See Eurocode section 5.8.7 which gives nominal stiffness (cracked and creeped) parameters for use in second order analysis. Tekla's reduction factors are probably simplified 'safe' numbers.

 

[struggle67]

Thanks, Steveh49

P.S. I think I got it but please correct me if I am wrong

Case 1 If there is no or minimum lateral load. Then It doesn't matter either 1.0 Ig or 0.2 Ig as long as they are the same for all concrete members it will give us the same analysis result except for deflections.

Case 2 If there is a lateral load (seismic or second-order analysis) then by using 0.2 I Tekla has already accounted for both cracking and creep.

 

[JoshPlumSE]

I am using 3D Tekla structural designer Software (TSD) to design buildings. I noticed that for concrete members TSD assumes cracked sections (Icr) for the analysis and by default Icr = 0.2 Ig. For uncracked sections TSD uses 0.4 Ig to account for long term effect especially creep. I think both Eurocode and BS code recommend to use uncracked cross-section for elastic analysis. If there is no seismic load, should I still use cracked section?
Does other 3D commercial software do the same and why are they not following the code recommendation? bigears

I'm not very familiar with "3D Tekla Structural Designer". But, I spent about 16 years working for RISA and now I work for CSI. So, I feel like I have some decent knowledge to contribute.

For the ACI code (used here in the US), it is recommended to use a linear assumption for cracked sections on the order of 0.7*Igross for columns and 0.35 Igross for beams. That's for "strength level" loading, but they then allow you to multiply by 1.4 for service level loads (i.e. 1.0*Igross for columns and 0.5*Igross for beams).

I suspect that ACI requires this because of the following:
a) They want to better capture 2nd order (i.e. P-Delta effects) in moment frame structures at strength level loads.
b) They want to allow for larger stiffness when you are designing for deflection and drift restrictions.
c) There are additional factors for "creep" when you're talking about long term or permanent loads

I'm a little surprised the British and Euro codes don't do something similar. I know that both RISA and SAP2000 allow for the use of "property modifiers" to account for an elastic analysis to approximate (or linearize) the effect of cracking in concrete members. So, I don't think this is all that unusual for TSD. And, I feel confident that (if they're doing this) then they give you the opportunity to change this as well. The big difference between programs is likely what the DEFAULT behavior is.

 

[engineering_patrol]

I suspect that ACI requires this because of the following:
a) They want to better capture 2nd order (i.e. P-Delta effects) in moment frame structures at strength level loads.

That has nothing to do with 2nd order effects. Other (much different) amplification factors shall be applied as per ACI when designing slender structures. Alternatively you can perform non-linear cracked analysis and use the default modifiers (=1).

If there is no seismic load, should I still use cracked section?

Of course not! Afforementioned crack coefficients are to be applied for lateral and accidental or transient loads only and not for service loads! Service loads have to be calculated without applying crack coeffiecients.

 

[JoshPlumSE]

That has nothing to do with 2nd order effects. Other (much different) amplification factors shall be applied as per ACI when designing slender structures. Alternatively you can perform non-linear cracked analysis and use the default modifiers (=1).

Follow the flow chart in Figure R6.2.6 of ACI 318-14.

Slenderness Neglected - No
Analyze columns as non sway - No
Slenderness effects at column ends (i.e. P-Big Delta) - 2nd order analysis per 6.7 Elastic (which permits the section properties I described in my previous post)
Slenderness effects along column length (i.e. p-little delta) - 2nd order analysis R6.7.1.2 or R6.8.1.3 (which merely says that you can sub-divide your columns to capture the p-little delta effects)​

Now, these simplified methods of determining a cracking factor are just that.... simplified. In my opinion, this is the most common way of doing concrete analysis in the US.... Assuming a "property reduction factor" for concrete members that linearizes it's stiffness then relying on a programs P-Delta analysis capability to capture the geometric non-linearity of column slenderness.

 

[engineering_patrol]

ok

 

[struggle67]

JoshPlumSE
Thanks for sharing your valuable experience. Star for you.

I'm a little surprised the British and Euro codes don't do something similar.

Eurocodes does require to modify stiffness for second-order analysis and seismic. But for linear elastic analysis it is recommended to use uncracked section properties (1st order? and service?)

Btw any thought & comment on my reply earlier to Steveh49.

For the ACI code (used here in the US), it is recommended to use a linear assumption for cracked sections on the order of 0.7*Igross for columns and 0.35 Igross for beams. That's for "strength level" loading, but they then allow you to multiply by 1.4 for service level loads (i.e. 1.0*Igross for columns and 0.5*Igross for beams).

It is good. At least ACI differentiate strength level and service level.