Return to index: [Subject] [Thread] [Date] [Author]

Re: Post Tension (SAP Question)

[Subject Prev][Subject Next][Thread Prev][Thread Next]
T. Eric Gillham:
Thank you for your response.
The reason why I want to introduce PT cables in these walls is the flexural forces at the end of the cantilever OR at when the walls are supported by the fdn mat. This cantilever wall is taking considerable part of the gravity loads on three stories. At the beginning of the analysis I tried several systems other than using walls at the ends of the building which induced excessive deflections at the corners of the building. This point called the idea of using walls to reduce the deflection and also enhanced greatly the distribution of the gravity loads to the columns at the perimeter. Back to PT, I then preferred using PT to make use of the complete wall section in serviceability (considering also construction sequence which will expose the walls on various levels to tension. There is no possibility to install shoring underneath the cantilever during construction). Therefore PT will be done on phases during construction to eleminate tensile forces on the cantilever part of the walls during construction too.
I am running Elastic analysis and I did not account for PT during analysis however I want to locate the profile which will be better done if I include the PT in the model to see the impact of moving the profile on the output. The suggested profile will be curve staring from one end of the cantilever to the other end. I am thinking in three cables, one for each floor. Vertical cables only will be placed at the center wall to transmit forces to fdn mat. This way to implement procedure #3 as suggested by you I think I have to break the cable into pieces to form that frame or the curved profile.
I am not sure whether Push over will be necessary in this case or no.
Thank you,
----- Original Message -----
Sent: Wednesday, January 31, 2001 12:17 AM
Subject: FW: Post Tension (SAP Question)

Although I have heard of PT being used for masonry as well as precast R/C walls, I (like Nicholas Blackburn) am wondering why you would put P/T in a CIP R/C wall, ESPECIALLY if it already has axial load due to gravity (and you stated that it is a cantilevered wall, so presumably there will be no net tension problems like those encountered with heavily coupled walls).
Be that as it may, I can see at least 3 approaches you could take if you are running an ELASTIC analysis:
1)  If you need the axial loads on the walls for design purposes, simply account for the P/T during the design, not the analysis, phase.  With an elastic analysis, it shouldn't matter if you don't have the P/T in there during your analysis.
2)  You could also put some horizontal dummy frame members at the level of the live ends of the P/T and load the wall uniformly (using frame distributed loads) to account for the P/T induced P/A.
3)  If you want to get fancy, model the P/T strands individually as vertical frame members integral with the shell elements.  Set the temperature dependent properties of all other elements (besides the strand frame members) to 0, then use a temperature load case to "shrink" the P/T strand frame members the desired amount to induce compression in the shell elements.  I have done this on RISA3D (but, I admit, not on SAP2K) succsessfully, and although it requires a fair amount of interation (since the "solution" is a certain P/A) it can be done.
All 3 of these answers are predicated on your performing an elastic analysis.  If you are performing a SAP2K pushover, let me know the specifics of your setup and I can give it some thought.
Hope this helps.
T. Eric Gillham PE 
-----Original Message-----
From: Wael Shabana [mailto:waelshabana(--nospam--at)]
Sent: Wednesday, January 31, 2001 5:07 AM
To: seaint(--nospam--at)
Subject: Post Tension (SAP Question)

I am designing a structure whose system includes Walls which act as part of the gravity system and shear wall as well. The walls act like cantilever (cantilever span is 43' , worst span). I modeled the walls as shells on the SAP 2000 however I want to locate PT cables in those cantilevered parts of the wall. My question, is there a way to introduce the PT forces to the shells in the SAP or it is only available for Frames. If so, how do I model this wall as a frame? The wall height is 55' and rests on a fdn mat on a length of only 42'. The walls total length is 104'
Thanks in advance.
Wael Shabana, P.E.