Total deflection of PT Elements

[Aet20]

Dear professionals,

I have a question of designing PT elements let says PT beam using RAPT, will the software consider the total deflection of beam including deflection at column support in the design? As what I can think of by changing the column stiffness can only control the moment taken by the beam and hence affect the deflection results but in actual the deflection might be more than the analysis result when there is rotation of column support.

I am not certain on this concept, can someone please enlighten me on this? Or how can we estimate a result which is closer to actual as the PT elements are supporting some brittle facade at the cantilever end which meant to be controlled with limited deflection.

 

[Retrograde]

RAPT does take into account the rotational stiffness of the columns. You can adjust the percentage of stiffness you want to use by adjusting the %Izz value in the "Columns" tab.

I do not think it accounts for axial shortening of the column.

 

[rapt]

Are you asking about deflection at the column support or rotation at the support?

Maybe you should just contact me directly if you are a RAPT user!

 

[Aet20]

Hi Retrogade,

I understand that the stiffness can be adjusted at the column tab, but on my point of view that will just control the moment taken by the beam itself and we still don't know what is the deflection at the column?

Hi rapt,

I'm looking at the deflection at the column which may result in higher total deflection of beam by adding up the beam deflection in the design model. Yes, I'm using RAPT(under EC) and was wondering whether RAPT can capture case where there is situation when the beam has no backspan and hence the column may deflect and result in the higher deflection. I'm not sure how can I contact you directly?

Thank you for all replies

 

[rapt]

Via the RAPT support!

Are you worried about the rotation at the end column causing an increase in the beam deflection, or shortening in the column. You wording is suggesting a mixture of the 2 and I am not sure which you want.

Which is it?

 

[KootK]

Improbable as it seems, it sounds as through we're talking about a beam that is lacking a back span and, rather, is truly cantilevered from the face of the supporting column such that any column rotation will have a large effect on cantilever tip deflection.

 

[rapt]

If the case being discussed is the one Kootk is suggesting, with 1 or 2 cantilevers but no main span, RAPT gives a warning in the output results that deflection caused by rotation at the top of the column is not included and the designer has to determine a value for this.

In all other cases RAPT will include the effects of rotation at a support.

 

[KootK]

If I've guessed the setup correctly here, a common design approach would be:

1) Use something like rapt to get your beam deflection, ignoring the column rotation.

2) Model the column separately, using whatever tool you have for the purpose, and determine it's rotation.

3) Apply the column rotation to the beam as a rigid body motion and add the result to #1.

 

[rapt]

Kootk,

Yes, unless it is a single cantilever and the tension in the column is low so it is uncracked and the inability to model axial compression does not matter, in which case you could actually define it as a span of the height of the column.

 

[Aet20]

Yes, this is the case that I'm referring! Was wondering if rapt included rotation at top of the column in the deflection results. Now I understand that it actually is included and designer have to calculate separately when there is a warning.



Thank you for suggesting the idea! I will continue study with these suggestion, much appreciated.

 

[rapt]

Aet20,

It is NEVER included in a case with only cantilevers and a single column!

 

[Aet20]

:It is NEVER included in a case with only cantilevers and a single column!

Thank you! Understand it now, will do a separate checking for beam without a backspan. But now I know for those with backspan it is already included. Thank you again!

 

[KootK]

At the risk of muddying the waters, I think that rapt was proposing that you model the column as a fictitious, equivalent, horizontal backspan. That's clever if it's appropriate to your situation and may allow you to better exploit the capabilities of your software.

I struggle to imagine a backspan-less cantilever that would ever be long enough to effectively post-tension unless the tendons run back into a conventional slab as shown below. Is that your situation?

 

[rapt]

Kootk,

Yes, you can model it as that but without the column. Use a knife edge support.

I can imagine a double cantilever and no back spans with PT, as a bridge pier.

Not sure how PT would help a single cantilever as any uplift effect would simply rotate out so agree on that one.

 

[KootK]

Yes, you can model it as that but without the column. Use a knife edge support.

My sketch wasn't meant to represent the fictitious backspan concept but, rather, the real situation... whatever that is. Good call on the bridge pier.