POST TENSIONED CIP LONG SPAN CONSTRUCTION

[tbone73]

Post-tension design experts, when you model your structure with your software of choice, do you somehow represent the dead load balancing effects of the post tensioning in your model? When you have long span (60 ft) beams with substantial dead loads, parking garages for instance, you get large column moments and at the top level minimal axial loads so your column reinforcing could get rather large. Or do you just model it as regular reinforced, and design for full moments?

Comments/recommendations are greatly appreciated.

 

[rowingengineer]

For axial loads below about 20% of capacity I prefer model these as beams (about 0.40Ig).

A clear conscience is usually the sign of a bad memory

 

[rapt]

tbone73,

You seem to be implying that the column moments will be larger with the prestress!

The prestress effects will normally reduce the column moments. The effects are real so why not allow for them.

If you add in shortening effects, yes, the column moments from shortening in a prestressed structure will be larger than for an RC structure, but normally only by about 10% (depending on climate and environment) as shrinkage and temperature shortening are the same for both structures, and, as the amount of prestress is normally fairly low in building structures, creep shortening could actually be higher in the RC building. So the extra is for axial P/A. The rest occurs in RC structures and is often ignored!

 

[tbone73]

rapt I was not trying to imply that the prestressing increases column moments that is why I used the term dead load balancing. What I was/am trying to say is that when you ignore the effects of PT you get large column moments, do you typically accept these moments and design accordingly or do you somehow model the effects and design for the reduced moments? If you do incorporate these effects, by what means?

 

[rowingengineer]

tbone73,
You have lost me in the mist of the night, can you try another explanation angle.


A clear conscience is usually the sign of a bad memory

 

[slickdeals]

@tbone73:
When I model transfer girders in the lateral analysis program, I include an upward force (balancing force) in the model. The PCI Manual has a chart that gives you the equivalent line load that can be applied to the beam in order to simulate the effects of post-tensioning.

As Rapt said, these effects are real and you will end up with a more realistic column moment rather than something obnoxious in the case of transfer girders.

See attached.

 

[BAretired]

If you want to understand the behavior of the structure, you would consider the upward force of the tendons. If tendons are draped in a parabolic configuration, the upward force is a uniform load throughout the span. If tendons are harped, they provide a concentrated load at the harp point.

If you are interested in the ultimate load the structure can carry, you combine the positive and negative moment resistance and compare it to the simple span moment.

BA

 

[csd72]

It is definately not a silly question, I have often wondered what the general practice was myself.

When I was a young engineer in australia I designed a few buildings with reinforced columns (by us) and PT floors that were designed by others. I took what I believed to be the conservative route and designed the columns as if the whole building was reinforced concrete.

 

[tbone73]

Thanks all, slickdeals that is exactly what I was trying to get at. Thanks for the attached figure. This project has a tight budget (don't they all) and the architect is trying to get us to reduce costs anywhere we can. If I ignore the pt effects the reinf. ratio in some columns is 3.5%, if I include them they reduce down to 2.5%. A decent savings. I was just wanting some assurance that this is accepted practice.