View previous topic :: View next topic 
Author 
Message 
JVCSNL ...
Joined: 26 Jan 2003 Posts: 159

Posted: Thu Jul 08, 2010 3:25 am Post subject: Regarding PD analysis. 


Dear SEFI members,
I suggest the members to go through the discussion on SEFI during Sep. 2003.
The topics were PDelta analysis, Slender Column etc. This point was discussed at length under these topics.
Regards,
Jignesh Chokshi
Posted via Email 

Back to top 


lele_raj ...
Joined: 26 Jan 2003 Posts: 145

Posted: Thu Jul 08, 2010 6:09 am Post subject: Regarding PD analysis. 


Thanks for the headsup. My concern is that pdelta effect is thought to be present / effective only in the cases of long slender column. Examples I provided show that this is not the case, it's present even in latticed structures (under 100% design loads, though). And in fact, it needs to be considered in conjunction with two types of nonlinearities and the structure needs to be analysed appropriately.
I'm not sure whether these aspects were covered in the discussion of Sep03.
Best regards,
Rajendra (Raj) Lele
P Please consider the environment before printing this email or any attachment/s.
From: JVCSNL <forum@sefindia.org>
To: general@sefindia.org
Sent: Thu, 8 July, 2010 2:58:07 PM
Subject: [SEFI] Re: Regarding PD analysis.
Dear SEFI members,
I suggest the members to go through the discussion on SEFI during Sep. 2003.
The topics were PDelta analysis, Slender Column etc. This point was discussed at length under these topics.
Regards,
Jignesh Chokshi
Posted via Email 

Back to top 


hemal ...
Joined: 01 Apr 2008 Posts: 129

Posted: Thu Jul 08, 2010 1:24 pm Post subject: Regarding PD analysis. 


Dear SEFI members,
As i told earlier from PD analysis we will get global displacement (delta) and moments due to this delta caused by gravity loads (1.5DL or 1.2DL+1.2LL) for columns (no matter slender or shorter columns).
From value of global displacement (delta) for particular storey, we may determine stability index (Q) from ANNEX E IS:4562000 (page no92). From value of Q we may know weather columns are sway or non sway and we may use effective length factors accordingly from WOOD'S chart (FIG26,27) from IS:456. I don't think any software has utilize this clause regarding STABILITY INDEX. Generally we need to input eff. length factor manually. Some softwares calculate eff. length factor automatically for sway (unbraced) or nonsway(braced) columns automatically but decision of weather column is sway or nonsway is left to user (i.e. no use of Q).
For slender columns local displacements (due to buckling) will be more so secondary PD moments calculated by cl. 39.7.1 of IS:456.
SEFI members may correct me if i am somewhere wrong.
Regards
hemal mistry
Surat
 On Thu, 8/7/10, lele_raj <forum@sefindia.org> wrote:
Quote: 
From: lele_raj <forum@sefindia.org>
Subject: [SEFI] Re: Regarding PD analysis.
To: general@sefindia.org
Date: Thursday, 8 July, 2010, 4:53 AM
Thanks for the headsup. My concern is that pdelta effect is thought to be present / effective only in the cases of long slender column. Examples I provided show that this is not the case, it's present even in latticed structures (under 100% design loads, though). And in fact, it needs to be considered in conjunction with two types of nonlinearities and the structure needs to be analysed appropriately.
I'm not sure whether these aspects were covered in the discussion of Sep03.
Best regards,
Rajendra (Raj) Lele
P Please consider the environment before printing this email or any attachment/s.
From: JVCSNL
To: general@sefindia.org
Sent: Thu, 8 July, 2010 2:58:07 PM
Subject: [SEFI] Re: Regarding PD analysis.
Dear SEFI members,
I suggest the members to go through the discussion on SEFI during Sep. 2003.
The topics were PDelta analysis, Slender Column etc. This point was discussed at length under these topics.
Regards,
Jignesh Chokshi

Posted via Email 

Back to top 


Rajiv Diamond Sponsor
Joined: 04 Aug 2008 Posts: 52

Posted: Fri Jul 09, 2010 4:30 am Post subject: Effective Lengths for columns, Sway/No Sway and P Delta 


Dear All:
Please note that currently ETABS is doing all these calculations as desired by Hemal for column design. ETABS computes Story index Q and determines if a story is Sway/No Sway and calculates Effective Length factors accordingly. User has a choice of enhancing effective length factors. Many additional things for column and shear wall designs are now available.
If you are an ETABS user you can write us at support[at]csiindia.com. Just change the word [at] with @.
More on these important aspects in our support mails to you!!
Best regards
Rajiv 

Back to top 


PSRaju ...
Joined: 02 Jun 2008 Posts: 92 Location: HYDERABAD

Posted: Fri Jul 09, 2010 5:50 am Post subject: Re: Regarding PD analysis. 


Dear Er. Hemlal,
Thank you for information on stability index.
Regards,
Raju
hemal wrote:  Dear SEFI members,
As i told earlier from PD analysis we will get global displacement (delta) and moments due to this delta caused by gravity loads (1.5DL or 1.2DL+1.2LL) for columns (no matter slender or shorter columns).
From value of global displacement (delta) for particular storey, we may determine stability index (Q) from ANNEX E IS:4562000 (page no92). From value of Q we may know weather columns are sway or non sway and we may use effective length factors accordingly from WOOD'S chart (FIG26,27) from IS:456. I don't think any software has utilize this clause regarding STABILITY INDEX. Generally we need to input eff. length factor manually. Some softwares calculate eff. length factor automatically for sway (unbraced) or nonsway(braced) columns automatically but decision of weather column is sway or nonsway is left to user (i.e. no use of Q).
For slender columns local displacements (due to buckling) will be more so secondary PD moments calculated by cl. 39.7.1 of IS:456.
SEFI members may correct me if i am somewhere wrong.
Regards
hemal mistry
Surat
 On Thu, 8/7/10, lele_raj <forum> wrote:
Quote: 
From: lele_raj <forum>
Subject: [SEFI] Re: Regarding PD analysis.
To: general@sefindia.org
Date: Thursday, 8 July, 2010, 4:53 AM
Thanks for the headsup. My concern is that pdelta effect is thought to be present / effective only in the cases of long slender column. Examples I provided show that this is not the case, it's present even in latticed structures (under 100% design loads, though). And in fact, it needs to be considered in conjunction with two types of nonlinearities and the structure needs to be analysed appropriately.
I'm not sure whether these aspects were covered in the discussion of Sep03.
Best regards,
Rajendra (Raj) Lele
P Please consider the environment before printing this email or any attachment/s.
From: JVCSNL
To: general@sefindia.org
Sent: Thu, 8 July, 2010 2:58:07 PM
Subject: [SEFI] Re: Regarding PD analysis.
Dear SEFI members,
I suggest the members to go through the discussion on SEFI during Sep. 2003.
The topics were PDelta analysis, Slender Column etc. This point was discussed at length under these topics.
Regards,
Jignesh Chokshi

Posted via Email 


Back to top 


hemal ...
Joined: 01 Apr 2008 Posts: 129

Posted: Sat Jul 10, 2010 1:20 am Post subject: Regarding PD analysis. 


Dear Mr. Rajiv,
Thanx for the information on ETABS. It has many gifts for indian standards for frames and shear walls.
I have gone through concrete design manual of etabs for Indian codes (I think it is first time released). It is mentioned that "Etabs by default considers all columns as sway columns and calculates eff. length factors accordingly (varies from 1 to infinity, sometimes unusual and very high, needs overrides as mentioned there). But it is mentioned that if u perform PD analysis, it will assume that further sway is prevented and will take eff. length factor 1 conservatively.
However, i have tried PD analysis in a trial version but it takes eff. length factor for sway columns even after performing PD analysis.
I think there is nothing mentioned about STABILITY INDEX (Q) which decides sway or nonsway column.
Regards
Hemal
 On Thu, 8/7/10, Rajiv <forum@sefindia.org> wrote:
Quote: 
From: Rajiv <forum@sefindia.org>
Subject: [SEFI] Re: Regarding PD analysis.
To: general@sefindia.org
Date: Thursday, 8 July, 2010, 11:30 PM
Dear All:
Please note that currently ETABS is doing all these calculations as desired by Hemal for column design. ETABS computes Story index Q and determines if a story is Sway/No Sway and calculates Effective Length factors accordingly. User has a choice of enhancing effective length factors. Many additional things for column and shear wall designs are now available.
If you are an ETABS user you can write us at support[at]csiindia.com. Just change the word [at] with @.
More on these important aspects in our support mails to you!!
Best regards
Rajiv

Posted via Email 

Back to top 


Rajiv Diamond Sponsor
Joined: 04 Aug 2008 Posts: 52

Posted: Sun Jul 11, 2010 3:39 am Post subject: ETABS and Concrete Design Issues 


Dear Hemal:
Originally we had adopted the method that a user would be doing a PDelta analysis always, since it is what our code requires and moreover second order analysis is now common and recommended in most advanced codes. Since our code gives no specific guidelines and rather intimidates a new user (See Explanatory Notes published in 1978), we had noticed that most users were NOT doing it due a sense of fear.
ETABS' original method was to assume that the user would be doing a PDelta analysis and then K factor would be taken as 1.0. This is similar to ACI code. But ETABS would be required to magnify these moments to meet the demand of PDelta (local) deformations. So ETABS would always be adding additional slenderness moments based on an effective length factor of 1.0. This was considered conservative.
However after conducting many seminars all over the country we realised that very few engineers were actually doing a PDelta analysis correctly, either due to lack of interest or lack of guidance or simply due a very short time allocated for design. So to safeguard the building against a possible misuse of ETABS, we decided to update the program to have a 100% codal compliance.
In ETABS 9.7 we made computation of effective length factors the default method and PDelta was optional. When a PDelta analysis is done then the user should use overwrites to change the effective length factor to 1.0 on his own.
Our new update 9.7.1 has a more refined and a very robust design for concrete elements for Indian code. In that we have added Q factors, effective length factors, Sway/NoSway issue and we have even improved our section designer for a more optimized search for steel in column sections which can be of any shape. This has been completed and is to be released soon.
Our code wants to include PDelta effects for story sway as well as PDelta due deformations within the column length within a story. But somehow our code writers have not paid much attention to write the specification of such an analysis for the reasons best known to them.
So what is needed is a comprehensive review of PDelta analysis and the underlying issues and we need to educate our design engineers and design managers at all levels on how to include it in design correctly.
Best regards
Rajiv 

Back to top 


hemal ...
Joined: 01 Apr 2008 Posts: 129

Posted: Sun Jul 11, 2010 4:51 pm Post subject: test 


Dear Mr. Rajiv,
Thanx for the really fruitful explanation on PD analysis, Q, k related to ETABS.
I have observed that due to consideration of column as sway column by default in v9.7, main reinforcement of column is veeery high (many times @ 2 to 3 times) as compared to v9.5. However overwriting k values to 1 in v9.7 gives 100% same results in v9.5 & v9.7. But, if we have 100 columns in a 100 story building, how much time it will take for overwriting k to 1 for all columns?
As mentioned in previous post, it is mentioned in manual somewhere that if PD analysis is performed k will be taken as 1 by default. But in v9.7 even after performing PD analysis, it takes k for sway columns (>1). Can u explain is it so or i am making any wrong statement.
I would like to suggest that in "preferences" u may put one variable for k as below
(1) calculate k based on Q values (sway or nonsway). (if PD analysis not performed)
(2) Take k=1 (sway or nonsway). (if PD analysis is performed)
(3) calculate k for sway columns
(4) calculate k for nonway columns
option3 and 4 are not logical as user can not decide weather column is sway or nonsway. but, may software gives user this choice. Actually it is somewhat dangerous because many times people takes 0.65 or 0.85 or 1 factor without considering storey height, support condition, beam framing etc.
Option1 will give k<1 if non sway column (Q<0.04), which will be advantageous. which we are not having in present version.
If all above suggestions are incorporated in v9.7.1, then i think it will be the first software to address this critical issue.
Regards
Hemal
Posted via Email 

Back to top 


Rajiv Diamond Sponsor
Joined: 04 Aug 2008 Posts: 52

Posted: Mon Jul 12, 2010 5:05 am Post subject: PDelta Analysis in ETABS 


Dear Hemal:
Fixing overwrites in ETABS takes a very small time of about a few seconds only. Select all column and assign them an overwrite.
All other things have been taken care of in ETABS 9.7.1, and it offers an improved shear wall design too!
Just wait for a few days only.
Best regards
Rajiv 

Back to top 


bijay sarkar ...
Joined: 14 Dec 2009 Posts: 314

Posted: Mon Jul 12, 2010 12:31 pm Post subject: 


Dear all,
My observation on the subject matter are as follows for your comment :
Analysis which we normally do, is called an elastic linear analysis which does not take care of the deflections of the structures, be it from sway or from slenderness/buckling.
PDelta Analysis is called a nonlinear analysis which incorporates the effects of deflections of the structure.
Now there are two types of deflections : (i) Structural deflection or Global deflection or what we normally call structural “sway”. (ii) Member deflection or local deflection or slenderness deflection. Additional moments comes from these deflections which are also to be considered for design.
Now, this is pertinent to be confirmed whether the frame is a “sway frame” or a “nonsway frame” for deciding which Fig. (26 or 27) is to be used for effective length factors for columns. For identifying a “sway” or “nonsway” frame, IS 456 has directed to calculate an index Q which depends on the elastic linear analysis of the structure.
If the structure is a “sway frame”, then only we may go for PDelta analysis, otherwise, elastic linear analysis is enough.
But PDelta analysis of Staad does not cover the deflection due to slenderness effect of the members which requires a buckling analysis of the frame. Therefore, we may say that after PDelta analysis, additional moments due to slenderness/buckiling are yet to be added, if any.
For this slenderness/buckling moments, we require to calculate the effective lengths of the members. For “sway frame”, effective lengths shall be in accordance with Fig. 26 of IS 456 and for “nonsway frame”, Fig. 27 shall be used. But as we have already analysed the structure for PDelta effects, this is my personal opinion that, now, we shall go for Fig.26 only i.e. for nonsway frames to calculate the effective lengths to find out the additional moments for slenderness/buckling. If PDelta analysis is not carried out and the frame is a sway frame, then only we will go for Fig. 27 to cover up both the deflections as we normally do.
I am submitting my observations stepwise for comments of the sefians :
a) Conduct elastic linear analysis what we normally do.
(b) Find out Q to decide the type of frame (sway or nonsway).
(c) If nonsway frame, PDelta analysis is not to be carried out. Effective lengths of members shall be based on Fig. 26 of IS and additional moment due to slenderness/buckling shall be added to the normal elastic linear analysis to find out the design moments.
(d) If sway frame, we may carry out PDelta analysis or may not.
(i)If PDelta analysis is carried out, calculate effective length factors based on Fig. 26 of IS 456 which is for nonsway frame for calculating additional slenderness/buckling moments. Add this moment with the PDelta analysis. Is it right or Fig. 27 is yet to be used?
(ii)If PDelta analysis is not carried out and the frame is a sway frame, calculate the effective length factors based on Fig. 27 of IS 456 which is for sway frame and calculate the additional moment based on the buckling/slenderness equation provided in the IS. Consideration of Fig. 27 of IS will cover both the deflections i.e. from sway of structure as well as slenderness/buckling of the members to find out the additional moments.
bijay sarkar 

Back to top 




You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum


