Where to use Redundancy Factor=1.3? 3

[Osama Anwar]

If I have a Symmetrical Special moment resisting frame in SDC D with 3 bays in one direction and 8 bays in another direction and I am solving it in ETABs and have enabled P-Delta effects, will I use redundancy factor=1 or =1.3. Is there any way to access redundancy factor without calculation if not then how the calculation is made?

 

[KootK]

If your S1 is greater than 0.75, then you are in SDC E and rho is 1.30 regardless.

What informs your opinion on that? I'm not seeing a one to one relationship between being in SDC_E and having ρ=1.3 in the ASCE provisions.

 

[JLNJ]

My mistake. No relationship exists.

We had a project a little while back which went from SDC E to D due to the reduction of the accelerations from ASCE 7-10 to 7-16 and I was not remembering correctly about the assignment of rho. Should have actually looked it up first! Mea culpa!

 

[KootK]

No worries. If we all thoroughly fact checked our contributions before posting them, we'd never get anything done. You can only expect so much of pro bono help.

 

[HTURKAK]

4) With your very well distributed, and highly symmetric lateral system here there's zero chance that the removal of any one beam will result in a torsional irregularity much less an extreme torsional irregularity.

Just for curious, Mr.KootK, how did you estimate zero probability , without knowing the column and beam dimensions , size of the bldg and storey ht?.
I agree with you , can say without calculation, removing one of the beams will not cause to a reduction by more than 33% of the lateral strength of the structure.

However,in this case it is hard to say the same for torsional irregularity . The common practice, outside the North America,(the architects impose ) the use of slender beams and columns restricted with wall thk.

 

[Osama Anwar]

Thank you all of you for deep insights. I did another check.
1) First of all I ran the analysis and fetched the values of "Story Max/Avg Displacements" option in the under tables tab (table > Analysis > results > displacements > Story Max/Avg Displacements) and view all the ratios.
2) Then I removed the beam from the story having highest ratio (all ratios were less then 1.1)
3) Performed the analysis again, checked the values for "Story Max/Avg Displacements", luckily none was above 1.15.
I concluded that I could go with roh=1. What's your final thoughts?

 

[HTURKAK]

Mr ANWAR,

If you performed the analysis , with considering accidental ecc.at unfavorable side 5% and with removed beam ( Y/ 2-3) and find Story Max/Avg Displacements 1.15 , yes ...you can use redundancy factor =1.0

I want to remind the minimum column and beam dim. 250 mm . and if you perform inelastic analysis, you will find more than 1.15 but should not be more than 1.25.

Good luck.

 

[Osama Anwar]

I am using column 375mm x 600mm and least beam dimension of 375mm x 375mm and I considered 5% eccentricity.

 

[r13]

I suggest to test the model with hinged beam at varies locations around the perimeter to gain better understanding of its behavior.

 

[KootK]

Just for curious, Mr.KootK, how did you estimate zero probability , without knowing the column and beam dimensions , size of the bldg and storey ht?

I examined the fundamental, torsional nature of the building in a qualitative sense and recognized that it is capable of forming an almost ridiculous number of closed loop torsion cells as shown below.

As far as torsional regularity goes, almost none of the issues that you've listed above is actually relevant, including:

1) Story height.

2) building size.

3) the use of small narrow members in developing countries.

None of that stuff matters. The only thing that matters with respect to torsional regularity is the relative distribution of torsional stiffness about the building.

Yes, it is possible that the building might be torsionally irregular because all of the beams on the left side are 2000 mm deep and all of the beams on the right side are 375 mm deep. However:

a) This strikes me as so improbable as to not be worth considering / mentioning and;

b) I have enough faith in OP to trust that, if that were the case, he or she would have the good sense to bring it to our attention.

I did ask if there were any vertical oddities that might affect the distribution of torsional stiffness and OP assured me that there were none. I've been collecting information on the problem strategically so as to be able to provide a considered response.

I agree that clause would be satisfied unless there's something weird going on vertically that you've not shown us.

No, vertically it is same.

All of the beams shown in blue in your sketch are, in fact, moment frame beams correct?

Yes blue lines are beam.

 

[KootK]

You are expected to provide two V.B. (for steel SMRF ) or SW ( for R.C. SMRF) at short direction , far ends...

I actually consider that to be poor advice given the nature of this building. If you add a shear wall at each end of the building, and lose one of them somehow, you really would have a serious torsional irregularity. I would only support that recommendation if you had multiple shear walls or braced frames at both ends of the building.

 

[r13]

Can you assume a hinged beam along line 5 between J-M, or M-P, and see what happens.

 

[Osama Anwar]

I removed the beam at story 1 at J-M and observed the ratios were more or less same as before.

 

[Osama Anwar]

Can anybody guide me how to quote someone in this thread?

 

[KootK]

The simplest way is shown below. The function basically builds an HTML string like the one shown below which you can just type for yourself if you prefer.

[ignore]

No, vertically it is same.

[/ignore]

 

[r13]

Osama,

Thanks for the feedback. I guess the columns are quite rigid. Actually I am a little puzzled that the results are close without noticeable difference. Just my noise.

 

[Blackstar123]

Osama,
I don't think removing a beam outright is the correct way to go. By removing the beam you are disturbing the vertical load path which, I don't think is the intent of code. Code only says to consider the possibility of the "Loss of moment resistance at the beam to column connection at both ends of a single beam..". You may easily achieve this by releasing the end moments of a beam.

Furthermore, modal shapes do not represent actual displacements of a structure. It may look like they are giving a value of displacement but in fact they are only relative magnitudes just to give a sense of possible story displacements when the structure vibrates at a certain frequency. How much these relative magnitudes will be depends on the story mass and stiffness. Maximum response due to a seismic event is the combination of all such modes. If you want to determine if the structure will develop a torsional irregularity when you release a beam end moments, you need to check it for the final response and not for the individual mode shapes.

 

[HTURKAK]

None of that stuff matters. The only thing that matters with respect to torsional regularity is the relative distribution of torsional stiffness about the building.

If you have a computer program that can model inelastic structures, i will suggest you , perform inelastic analysis to a similar bldg, rectangular in plan, see the contribution of of frames around the center and see what is happening to the far end, corner column.

I actually consider that to be poor advice given the nature of this building. If you add a shear wall at each end of the building, and lose one of them somehow, you really would have a serious torsional irregularity....

I have not the same opinion. For this case ,the SW's added at 2,3 -4,5 at A and Y axis will be very efficient to reduce the torsional effects and drift control. Moreover, the lose of SW is less probable than beams and columns.

Learning has no age , no limits ; and i learned after 40+ years of experience ,not to comment when the full picture is not clear .

 

[Osama Anwar]

You may easily achieve this by releasing the end moments of a beam.

I released the moments of 2 beams one at plinth level and 2nd at 1st story from Y/5-4 and analyzed result w.r.t. this method

I ran the analysis and fetched the values of "Story Max/Avg Displacements" option in the under tables tab (table > Analysis > results > displacements > Story Max/Avg Displacements) and view all the ratios.

And none was above 1.139
I analyzed for Earthquake forces not modal shapes.
You can point out any flaw in my process you're welcome. Its my first time at ETABs. There is high probability that my method for computation of torsional irregularity may be flawed.
P.S. I was wrong in the previous post, it should have been "luckily none was above 1.1189" rather than "luckily none was above 1.15"

 

[KootK]

If you have a computer program that can model inelastic structures, i will suggest you , perform inelastic analysis to a similar bldg, rectangular in plan, see the contribution of of frames around the center and see what is happening to the far end, corner column.[/b]

I don't need to run a plastic, FEM analysis to know that this structure is torsionally regular. I can determine that just fine using my own judgement, experience, and intuition regarding structural behavior. The ETABS modelling is a prudent validation step for the person actually designing the structure but I am not that person.

I'm also fully away that the inner frames will be less efficient torsionally than the outer frames. That invalidates none of my previous comments. In the first sketch below, is the red torsion loop really that much less efficient or more lightly taxed than the black?

Moreover, the lose of SW is less probable than beams and columns.

1) What makes you say that? Lateral buckling of the compression block of shear walls is actually a very common failure during seismic events.

2) Regardless of the probabilities, ASCE7 clearly means for the loss of shear walls to be considered when shear walls are used as the lateral system.

For this case ,the SW's added at 2,3 -4,5 at A and Y axis will be very efficient to reduce the torsional effects and drift control.

3) I agree that would be very efficient but I submit that it's completely unnecessary in a building where every single column is part of a bi-directional moment frame.

4) Your proposal makes most of both end walls nearly solid concrete. That costs money and may compromise architectural intent regarding fenestration etc.

earning has no age , no limits ; and i learned after 40+ years of experience ,not to comment when the full picture is not clear.

And, in the decades that I've been practicing structural engineering, I've leaned that:

5) A better understanding of fundamental structural behavior often allows one to assess situations accurately with less information and;

6) Computer modelling is an aid to, but no substitute for, genuine intuition regarding structural behavior.