View previous topic :: View next topic 
Author 
Message 
btgprasad ...
Joined: 05 Aug 2008 Posts: 125

Posted: Tue Nov 27, 2012 7:11 am Post subject: 


<a>Nicely explained krishnan Sir
!
tall structure have to maintain enough stiffness to be in allowable lateral drift limits both in wind and Earthquake loads hence that required stiffness control time periods.
however,If we see Tall chimneys (275m) time periods will be 4 sec.Earthquake forces seldom governs.
"In the most critical earthquake zone with zone factor of 0.36 and response reduction factor of 1.5, the earthquake response is almost matching with that of wind response but never been crossing the wind response." extracted from conclusion of a paper.
larger Time period cause so many problems like Flapping,wind resonance for which mathematical idealization would be difficult and requires wind tunnel to analyze.here there is not concept of max limit of time period ,its concept of response.
I this context I want ask is there any process to take care of tolerances (error) in time period form actual field value.
Btg
Do you know .maximum drift allowed in Burj ?
it more than half a meter. 

Back to top 


sarfaraj.husain ...
Joined: 26 Jan 2003 Posts: 90

Posted: Tue Nov 27, 2012 7:21 am Post subject: Time periodTall Building Design 


in continuation to my trailing mail..........
for h >300m "T" goes on increasing..... (in general)
..........how to caliberate such flexibility with " Sa/g" for analysis & design.....
sarfraj
dear all
consider a tall building of height h =300m
time pd. T =0.075*300^0.75 =5.4 secs
how to get Sa/g ..........as per 18932002 all formula valid for T<= 4.00 secs...
if it is correct ...please clear my doubt...
sarfraj
 ญญ
Posted via Email 

Back to top 


gautam chattopadhyay ...
Joined: 17 Feb 2009 Posts: 128

Posted: Wed Nov 28, 2012 2:41 am Post subject: Time periodTall Building Design 


Sarfraz, first of all the formula you are using in computing fundamental time period is very empirical. A building 300 m high must have 75 to 100 floors suggesting as many lumped masses as the number of floors and number of springs = number of floors. Hence the dynamic stiffness matrix and the mass matrix will be of order 75 x 75 to 100 x 100. Since natural frequency is the eigen value of the dynamic matrix ([K]/[M]), 75 to 100 eigen values will appear off the characteristic equation of the dynamic matrix. I feel now you can follow that predicting natural frequency and hence the time period is not an easy task. Time period is given by 2*(pie)/n where n is natural frequency (eigen value) computed from above. I feel it is dangerous to impose such empirical formula on tall buldings which should be analysed as a MDOF system as I described above.
On Tue, Nov 27, 2012 at 4:39 PM, sarfaraj.husain <forum@sefindia.org (forum@sefindia.org)> wrote:
Quote:  dear all
consider a tall building of height h =300m
time pd. T =0.075*300^0.75 =5.4 secs
how to get Sa/g ..........as per 18932002 all formula valid for T<= 4.00 secs...
if it is correct ...please clear my doubt...
sarfraj
 ญญ

Posted via Email 

Back to top 


sarfaraj.husain ...
Joined: 26 Jan 2003 Posts: 90

Posted: Wed Nov 28, 2012 4:46 am Post subject: Time periodTall Building Design 


Mr. Gautam
i 100% agree with your post that "T" is a function of mass & stiffness........and it can be carried out only by any software like staad etc ...( for > 3x3)......
the term eigen value,stiffness, MDOF are related to structural dynamics & are correct way to get "T".....
but T =0.075h^0.75 (independent of stiffness) says our 1893......point is how can we follow this for tall structures.......refer clause 7.8.2 to scale up the responses as against T ( T = 0.075h^0.75 as per 7.6)....
is there any convergence for this......
sarfraj...
From: "gautam chattopadhyay" <forum@sefindia.org>
To: econf34289@sefindia.org,
Date: 11/28/12 08:44 AM
Subject: [ECONF] Re: Time periodTall Building Design
Sarfraz, first of all the formula you are using in computing fundamental time period is very empirical. A building 300 m high must have 75 to 100 floors suggesting as many lumped masses as the number of floors and number of springs = number of floors. Hence the dynamic stiffness matrix and the mass matrix will be of order 75 x 75 to 100 x 100. Since natural frequency is the eigen value of the dynamic matrix ([K]/[M]), 75 to 100 eigen values will appear off the characteristic equation of the dynamic matrix. I feel now you can follow that predicting natural frequency and hence the time period is not an easy task. Time period is given by 2*(pie)/n where n is natural frequency (eigen value) computed from above. I feel it is dangerous to impose such empirical formula on tall buldings which should be analysed as a MDOF system as I described above.
On Tue, Nov 27, 2012 at 4:39 PM, sarfaraj.husain forum@sefindia.org)> wrote: auto removed
Posted via Email 

Back to top 


KABIRDASGUPTA SEFI Member
Joined: 24 Aug 2010 Posts: 2

Posted: Wed Nov 28, 2012 5:26 am Post subject: Time periodTall Building Design 


Subject: [ECONF] Re: Time periodTall Building Design
From: forum@sefindia.org
Date: Tue, 27 Nov 2012 16:39:05 +0530
To: econf34289@sefindia.org
dear all
consider a tall building of height h =300m
time pd. T =0.075*300^0.75 =5.4 secs
how to get Sa/g ..........as per 18932002 all formula valid for T<= 4.00 secs...
if it is correct ...please clear my doubt...
sarfraj
Dear Mr Sarfraj,
All Values of Sa/g are same for T more than 4.0 sec.
Regards,
KABIR DASGUPTA
 ญญ
Posted via Email 

Back to top 


swamikrishnan EConference Moderator
Joined: 28 Jul 2011 Posts: 18

Posted: Wed Nov 28, 2012 11:30 pm Post subject: Re: Time periodTall Building Design 


To add to my earlier comment on long period ground motions, the response spectra for the Central Business District of Christchurch in the Feb 22, 2011, Christchurch earthquake are given in the following document:
http://www.ipenz.org.nz/christchurchspectra.pdf
Note the conspicuous bump in the spectra between periods 2.75s and 3.75s; compare this against the NZS1170 design spectrum. The observed spectral acceleration at 3.25s period is about four times the design spectral acceleration for a 500year event (Figure 3) and about twice the design spectral acceleration for a 2500yr event (Figure 2).
Clearly, the observed spectrum does not exhibit the smooth 1/T drop in the design spectrum. Similar longperiod "bumps" can be seen in response spectra of nearsource records such as the Rinaldi and Sylmar records observed in the 1994 Northridge earthquake.
Swaminathan Krishnan
California Institute of Technology
krishnan at its.caltec... wrote: 
(ii) Dynamics: The more flexible longer period structures are more
susceptible to longperiod longduration ground motion. Such ground
motion is produced by large magnitude earthquakes (moderate
earthquakes can also produce longperiod pulselike ground motion in
the nearsource region) and get amplified in deep sedimentary basins
and/or soft soil deposits. Unfortunately, there are not enough number
of records from such earthquakes collected in basins around the world.
As a result, the global dataset on which all design spectra are based
upon is deficient in longperiod motion. Ground motion from
relatively small magnitudes tend to have much less longperiod content
and the 1/T dropoff is reasonable for such conditions. The demands in
the longerperiod regime from these rare large events may be much
greater than what the code spectra across the globe seem to suggest.
The 1985 Mexico city earthquake demonstrated this; more recently, the
Christchurch event reinforced this (although this was not a large
event, the longer period ground motions were greatly amplified by the
soft soils). As engineers we must decide, whether to protect our
structures against such rare events or bet against the occurrence of
such events during the lifetime of the building.

Posted via Email[/quote] 

Back to top 


Manoharbs_eq General Sponsor
Joined: 17 Jul 2012 Posts: 422

Posted: Thu Nov 29, 2012 4:21 am Post subject: 


Higher the time period lower the frequency, hence low frequency structures attract lesser seismic forces. beyond time period 4 there is no significant change in acceleration value, critical is low frequency structures whose time period is less than 0.6 or so.
we can follow the value for 4 for more than 4 since further there is no significant change.
however how is that structure of 300m high is bare frame?.
if the structure is with infill we can use 0.9h/sqrtd.
Rgds
Manohar 

Back to top 


