post-tensioning bars

[Zouatine]

Dear all,

I hope you are doing well,

I'm trying to model a post-tensioning bar in my structure in Etabs as is shown in the figure, unfortunately, I'm not able to find the option of modeling a post-tensioning strand in the software, I was thinking if I can model them as external Tendons if no would you please give me any thoughts about how to model that?

thank you

 

[hokie66]

PT bars are just high strength steel rods. What is it you want to know about them?

 

[Zouatine]

Dear Sir,

Thank you for your reply,

about the PT bars, I want to model them in Etabs along with the height of the structure as it is mentioned in the previous figure in Etabs or Sap 2000 Software, however, when I was trying to find this option in both software I couldn't find this option to model them.

My main issue is how can I model this element PT bars in the software, would you please give me your suggestions.

 

[Trenno]

From what I understand, you're looking to induce a pretension in this element, such that it alters the initial global stiffness matrix?

 

[Zouatine]

Thank you for your reply Mr Trenno,

Yes as you said, to explain more I'm providing a PT bar along with the height of the structure in order to increase the lateral resistance against the seismic loading (Rocking Frame Structure), during the earthquake these element PT bars will get stretched as we can see in the next figure.

in the etabs i want to model these PT bars but I'm unable to find how I could design that.

 

[r13]

Does etabs has "tension only" element/member? Or can you release restrains at the ends of the element that representing the PT bar?

 

[Agent666]

Might be wrong here, but I have a feeling tendons (in terms of prestressed concrete) were only implemented in the Ultimate version of Etabs. Edit - ignore this, just realised you said you're using Sap2000

To introduce the prestressed could you add a temperature effect as an initial load case and then have all other cases as non linear using prestress case as starting point?

What type of analysis are you doing? Time history, equivalent static, modal, pushover, as that might have some bearing on the best way to model scenario.

 

[Zouatine]

yes there is an option to release the restraints in Etabs but my main task or issue is the modeling of PT bars

 

[r13]

Zouatine,

I don't understand your question on "how to model the PT bars". Maybe I don't see the issue correctly, but wouldn't be model connecting elements between notes (as the yellow line in your last sketch) with end releases be suffice? You might need to elaborate more on the difficulty/dilemma you are facing.

 

[KootK]

For some purposes, you may be able to get useful results by just replacing the PT bar with the load that it will apply at its upper anchorage. For the post-tensioned lateral systems that I've had exposure to:

1) the name of the game has really been to get that hold down / pre-compression effect while obtaining ductility via plastic deformation in other elements.

2) the post tensioned elements have been post tensioned to a fairly high percentage of their yield stress, leaving little in reserve to participate in other aspects of frame deformation.

I fully acknowledge that this simplification may well be unsuitable for many purposes but it's something to consider depending on your needs.

 

[Zouatine]

Dear r13,

I just tried to model the PT bars with a truss element and with the same Area as the PT bars however, when I was modeling I release the forces from the element and ran the analysis but the stresses coming from the gravity load some compression forces appear on the PT bars as you can see in the next figure. however, I believe that the PT bars work only in tension. would you please correct me ,

this is the option for releasing the restrains, I selected the elements and use the selected elements.

 

[r13]

Check user manual to see how it specifies "tension only" element, or spring. If it has command available, then your problem is solved by follow the instruction. If only the latter is available, then follow the instruction with equivalent spring constant calculated by k = (EA/L)[sub]ELEMENT[/sub]. Please note that I don't think the PT bars will offer much help in your case due to the very stiff nature of the braces. May I ask why you want/need to incorporate PT bars in your mode? KootK's suggestion may yield satisfactory solution, if none of the commands are available.

 

[Agent666]

Zouatine, if you're using the tension/compression limits you do need to run a non-linear static case to activate them (you can find these limits under assign frame menu in etabs (presumably the same or similar place in sap2000)). These limits won't function in modal cases or static cases. As per my previous comment, not sure on the type of analysis you are doing here to give best advice. Yes the PT strands will only be effective in tension for obvious reasons

Also you need to turn off the meshing of the PT line element as in the screenshot above it is attaching analytically to each intermediate floor as the force is changing, if you turn off the internal meshing it will only be attached at the top and ground if you draw it as one element top to bottom. My understanding was it is only attached to the top level and ground in these types of rocking systems with external restoring forces provided by the PT, so you need to achieve this in the modelling by having it effectively disconnected at intermediate levels by playing round with the meshing. Otherwise you'll have to model it offset form the plane of the brace somehow (but keep it simple by exploring the meshing first).

 

[r13]

Another way might worth a try. Test run lateral load cases without any gravity load. If still resulted in member compression, it indicates the PT bars are ineffective. Otherwise, the resulting member forces can then be superimposed on the runs with all loadings, but the PT bars deactivated. For better result, a spring constant, as suggested previously, shall be assigned to the "start" and "end" of the member for axial restrain.

 

[Brad805]

You need the ability to do a non-linear analysis for tension only member load cases.

 

[Agent666]

Running any lateral case (in this situation) without the gravity load is flawed, the PT in that case is the only thing providing the restoring action for overturning in a rocking frame. So of course they are going to be in tension and effective, I don't think your suggestion achieves anything.

The OP's problem is not to do with the PT being effective or not, it is to do with how to model the PT and prestress applied in a rocking frame.

To the OP, just noted if you are designing a rocking frame, you need to also model the uplift of the tension side columns. You appear to have fixed supports applied.

 

[r13]

I suspect the compression in the PT bars was mainly caused by the gravity load, so suggested the "test run" to see whether the lateral load case will trigger tension in the PT bars. I have no clue if there is a rocking load or not. But if it does, the uplift should be less than the gravity load, so not showing on the previous analysis output provided. BTW, I agree nonlinear analysis should provide better indication.

 

[r13]

Agent,

Do you think posttension in the center of the building is an effective way to eliminate "rocking"?

 

[Agent666]

You're not trying to eliminate or suppress the rocking. It is a rocking system. The ductile response is rocking. The PT strand provides the self-centering capability.

r13, if you don't have experience with this type of system or the analysis software being used (as appears to be the case) then please opt out of the conversation as it is not helping address the OP's original query.

They systems rock, the columns may uplift the PT strand provides the self-centering force to stop it overturning. I don't think you have experience in this type of system.

 

[r13]

I'll be silent, but stand by the sideline to learn a new system I would rather hear from the OP though.