Modal Analysis - Torsional Mode 4

[rummaan17]

Dear All,
I received a peer review about the modal analysis I conducted in a high rise building. It states the following :

"The preliminary mass participation factors indicate that mode 1 is predominantly translational in the transversal Y direction, mode 2 torsional and mode 3 translational in the X direction. It is not ideal to have torsional as mode 2".

1. What is the effect of Mode-2 on design as we usually concerned with fundamental frequency.?
2. How torsional modifier in ETABS is going to affect in this case?

Any addition would be highly appreciated. Thanks in Advance

 

[KootK]

1) if mode two encompasses a meaningful chunk of the modal mass then it, too, will impact design. There are two problems that I know of with dominant torsional modes. Firstly, it's difficult to predict their behavior accurately. Secondly, it can be tough to design ductility into a torsional response mode.

2) I believe that the torsional modifiers affect the individual elements modelled. As such their effect on the overall building response may be limited. It sounds as though you may have an issue with overall building configuration rather than individual element deign.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rummaan17]

Dear KootK,
Thanks for your response. There is a clause in UBC - 97 1630.7 Horizontal Torsional Moment. It account for added effect of eccentricity. We have already considered this in design. I don't find any clause stating which prohibit design of building with Second Mode as Torsional.

 

[Klitor]

Im not familiar with UBC, but in my code (EC8) such system would fall under "torsionally flexible system" which would add all sorts of penalties, among others cca 2x smaller "q" factor ("R" for you Yanks).
I would imagine your seismic code has a similar clause?

 

[KootK]

Accidental eccentricity is about your loads. A dominant torsional response mode is about your structural configuration. Addressing the former doesn't eliminate concern for the latter I'm afraid. I'm not sure which document, if any, recommends against this in your jurisdiction.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[MrHershey]

At least in US codes, the torsional issues are usually controlled through horizontal irregularities and their associated restrictions and additional requirements you have to follow when they're present.

The second mode being torsional in and off itself isn't necessarily a red flag, at least to me. Could be just that you're incredibly stiff in one direction in comparison to the other or your lateral system is more spread out in one direction that the other. As long as you're stepping through all the checks and doing anything extra that you may need to do, then I'm not aware of anything that prohibits you from doing it. This reviewer may think it's just poor practice, but there are certainly differences in opinion on that front.

 

[JasonMcKee71]

If you are using modal superposition for the dynamic analysis, then there is not any special need to take actions even if the first or second mode is torsional, because that effect is considered in the analysis.
On the other hand, an equivalent lateral static quake load would not be accurate if torsional mode governs the structure.
Despite these, it is preferable to have translational modes (first or second) as it results into smaller shear forces on columns and consequently less bending moments --> economical design.
To sum up, torsional modes are not forbidden but, translayional ones are better.

Jason McKee
proud R&D Manager of
Cross Section Analysis & Design
Software for the structural design of cross sections
Moment Curvature Analysis
Reifnorcement Design etc.

 

[mike20793]

Agreed with MrHershey. This building likely has a torsional irregularity or, more likely, an extreme torsional irregularity. I don't know how this translates to the UBC, but I imagine the guidelines are very similar to ASCE 7. I agree that it is not necessarily a red flag, as long as the proper penalties are applied according to the governing building code. Does the peer reviewer recommend changing or adding anything?

 

[PicoStruc]

It is somewhat difficult to change torsional behavour of building without argument with archtects because of shear wall requirement.

Anyways, dynamic analysis are required for torsional building (WTF Would you really design high rise with statis equavalent procedure, Come on !?)

In my opinion, what is really important is to extract shear force in shear wall using LOCAL pier (one pier label for each wall segment).
Torsion can be hidden in C-shapes or core wall. For example, shear component of torsion in flange of C-shape cancel each other in global response !!!

 

[rummaan17]

Thanks Alot Klitor, KootK, MrHershey, mike20793 & PicoStruc for your valuable response.

@ Klitor :- we have similar condition to account for accidental eccentricity but there is no prohibition.

@kootk :- yes its about load but from Earthquake. If I have designed the member for that added eccentricity referred as Accidental Eccentricity that result from difference of Center of Mass (COM) and Center of Rigidity (COR), I feel there is no need to change building configuration though it is undesirable.

@MrHershey :- I have similar thought like you as I have went through all necessary check to confirm codes requirement.

@mike20793 :- since my building is a part model of a hospital building separated by expansion joints. The peer recommended to add beam at expansion joint. I checked that and found the 2nd mode still to be torsional.

@JasonMcKee71:- I am using response spectrum dynamic analysis to analyze this hospital structure.

@PicoStruc :- Thanks for adding a new dimension for me to think on. I was considering something same as to replace rectangular core wall with C shape wall and rest can be block masonary.

 

[MrHershey]

The other thing to keep in mind is what is the peer reviewer going for? What is the review trying to achieve? If it's a code compliance review then I'd disagree with the comment because you're not doing anything outside of the code as long as you're taking your appropriate penalties and stepping through your appropriate checks. Perhaps that's not how the reviewer would do it, but that's the way you've chosen to do it and it's perfectly allowable by code.

Now if it's a value engineering review and this is a suggestion to help lighten up the structure, then I'd agree with the reviewer. Typically distributing your structural elements such that the building is naturally more torsionally resistant will be a lot cheaper than trying to force things with a scheme that isn't naturally torsionally resistant. But the distribution of cores and long walls is rarely just a structural decision. Would explain the comment to the architect and the owner in that context. "They're saying we could make the structure cheaper if we better distribute the structure the way it naturally wants to be distributed" and then give them a schematic example that they'll probably hate. Doesn't mean you can't do it the way you're doing it. Just means that it might be cheaper to do it a different way, at least from a structural standpoint. Then let the architect gong the idea because it messes up the layout or facade or whatever.

 

[rummaan17]

Thanks MrHershy for taking time to be descriptive. Much Appreciated. I am considering putting few walls for better stability.

 

[Klitor]

You can play around and improve the situation by splitting long walls into series of shorters ones using vertical constructions joints (I sometimes forget this simple fact)

 

[rummaan17]

Dear Klitor, Thanks alot for a very nice addition but how can I model this in ETABS. By just giving an offset?

 

[KootK]

I'm going to speak to only the "good practice" aspect of things here because, in that respect, I definitely agree with your reviewer when it comes to buildings in seismic zones deriving the lion's share of their torsional stability from single shafts.

1) If torsion is your second mode then that is a low energy mode that will be easily excited and will need to be able to dissipate significant seismic energy. I challenge anyone to describe a stable, energy dissipating yield mechanism for a closed, torsionally loaded, composite shaft assembly that would match up with anything that's been tested.

2) Combining the torsional demands in your shaft walls with the translational demands of other modes to arrive at values suitable for capacity design of elements below your plastic hinges is extremely complex. And quite unreliable in my opinion. Part of this stems from the fact that shaft torsion will require a reversal of demand in some walls relative to the translational modes that will factor into the mode combination process.

3) What even is the mechanism of torsion resistance in a shaft wall assembly that may already be plastified in flexure due to translational demands?

4) No matter how they're reinforced, concrete sections lose 80-95% of their torsional stiffness once they crack. If your shaft cracks, and you haven't accounted for this, get ready for torsion to be your first mode.

5) You may need to capacity design your shaft foundation for the torque delivered by the shaft yielding in torsion. Even if you can figure out that demand, it won't be much fun designing for it.

There's danger in subjugating your own good engineering judgment to the idiosyncrasies of that which the code does not prohibit.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rummaan17]

Dear Kootk, Thanks alot. You added alot value to this post. I already planned to change my structural configuration. Also refer any good book if you have in your mind to better understand Torsional Mechanism in building.

Secondly Can you also take some time to reply my post on Computers and Structures: SAFE
Link . I would be obliged. Thanks.

 

[sandman21]

My guess is that your building is outside the limits of ASCE 7, and are using the alternate procedures to get a shear wall core building through plan check. For that purposes of the alternate procedures the peer reviewer is code.

To add to what Kootk says, torsional modes also cause additional P-delta effects in corner columns which will need to be designed for. Also be aware that each cycle of the building the P-delta increase to the max, these typical gravity columns are now more similar to moment frame columns. Loss of strength in these columns will increase the torsional effects on the building causing additional P-delta on the remaining corner columns or the next gravity column in line.

 

[KootK]

I already planned to change my structural configuration.

Yeah, I know. I wanted to extend the discussion anyhow for both our benefit and for the benefit of those who may read this thread in the future.

Also refer any good book if you have in your mind to better understand Torsional Mechanism in building.

See the list below. You can find references on building torsion and references on seismic design principles but, rarely, references that cover both. I'll check out your other thread as well.

Building Torsion

Link Smith & Coull
Link Taranath
Link Zalka

Concrete Capacity Design

Link Paulay & Priestley
Link Moehele

Some articles that hit both topics

Link Paulay
Link Humar
Link Paulay

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rummaan17]

Thanks alot KookK. I will go through them and might discuss question if I came across.

@sandman21 : - I didnt totally understand what you means about plan check ? I checked building drift limit and increased Earthquake eccentricity to account for accidental torsion. I have also incorporated P-Delta effect and all columns designed came out to be satisfactory. If there is something I am missing do add. Thanks in advance.