Slab Bands 3

[slickdeals]

I know this item has been discussed before in this forum, but I wanted to ask a few more questions (specifically with how Ram Concept analyzes it).

I have a grid of 11m x 8m with some really heavy loads. I am using a 300mm thick slab with a 500mm x 2700mm slab band (in the 11m direction).

According to PTI and Bijan Aalami's recommendations, as long as the thickness of the band is <= 2t and width >= 3*overall thickness of band, then the behavior generally remains two way.

My question pertains to the design strip in the 11m direction. Ram Concept has the option of choosing either full width or Code T-beam for these strips.

For one way shear checks, the program is using only the width of the 2.7m wide band. Is this appropriate or should the one way shear check be based on a 8m wide strip?

For flexure checks, I think (conservatively) the band beam should be analyzed as a T-Beam and not a two-way slab.

Any other thoughts/suggestions are welcome.

 

[slickdeals]

@rapt:
I apologize for asking so many questions, but I do really want to try to get a better understanding of how things are done in countries outside the US. I am currently working in India and would love your input on any issues (code/construction related) that you have encountered here.

In the PT flat plates that you design, do you provide reinforcing similar to that of a mild reinforced slab i.e., reinforcing in the column and middle strips in both directions?

How do you provide your cables? I am sure you don't provide the banded/distributed pattern? Do you provide concentrated bands in the column strips and spread out tendons in the middle strips?

If its not too much trouble, can I ask you to post a typical flat plate floor designed in order for me to see how the tendons/reinforcing get laid out?

I presume you carry on the same logic for flat slabs (with drop panels).

I know your answer for band beam systems, you essentially design them as a continuous beam supporting a one way post-tensioned slab.

Thanks for all the help.

 

[rapt]

slickdeals,

Answering out of order,

2 With bonded PT, I would always use a column/middle strip pattern in both directions. About 70% in the column strip and 30% in the middle strip. I would only use banded/distributed if the column arrangement required it (lined up in 1 direction only).
Unbonded PT (I would never use it) has to be banded as weaving of tendons is very difficult. THis results in a different response to Q1, which is why I reversed the order.

1 C/M strip tendon layout
- over the column - always provide top reinforcement in the column strips, concentrated over the column
- bottom - only if the stresses are high enough to require it for crack control or if there is significant restraint to shortening (multiple cores etc) - AS3600 limit fior this would be .25 root fc. me countalways on a bottom mat of reinforcement.
Banded/distributed layout
- over the column - always provide top reinforcement in the column strips, concentrated over the column (at least ACI minimum but bonded PT area can be subtracted)
- bottom - always provide a bottom mat independant of stress level (ignore ACI .2 root fc limit)
In all of these cases I would ignore the ACI L/6 termination points for the topm and bottom reinforcement. Bottom reinforceemnt should extend to the support.

3 it is too much trouble. I do not design any more so have nothing available. Look at the moment distribution across the slab. That is how the tendons should be layed out, with tendons near to column for shear transfer to the column and then the spacing increasing towards the middle of the bay. Look at the C/M strip layouts in TR43 second edition. It is acceptable to have the column strip tendons equally spaced in that strip width and the middle strip tendons equally spaced in the middle strip with, but I would normally close up the spacing in the column strip at the column to ensure there is a tendon either side of the column close to the column for transfer. Once you do this, it is normally to then increasing spacing towards the centre.

4 Flat slabs with drop panels are exactly the same as flat plates except that the column strip tendons drape to the edge of the drop panels at internal columns (not at edge columns where they still drape to the centre of the column). The middle strip tendons still drape to the centre of the columns. The tendon distributioon in this case will not be 70/30 as the column strip tendons are much more effective than the middle strip ones as they have the extra depth of the drop panel in the negative moment zone. M/BD2 logic as well as load balancing logic will both show this.

 

[rapt]

Sorry, Had some keyboard problems (batteries now replaced)

End of 1 C/M should have been

"some countries always insist on a nominal bottom mat of bottom reinforcement, irrespective of stress conditions."

 

[slickdeals]

Just to revive this thread again.....

In Australia, it seems like metal deck forms (Bondek) are used in lieu of traditional formwork. I am looking at an option of being able to use them on a project in India. Columns are on a 11m x 8m grid.

I came across a presentation in PTIA (

http://www.ptia.org.au/Documents/Current Trends in Post-tensioned Building Design.pdf)

There is a line that suggests that these bondek systems can be used as a two-way system also.

Does anyone have any additional information on this?

 

[rowingengineer]

I didn't see it? I wouldn't use it as a two-way system slab only system. normally a band solution would be used.

How could you do anything so vicious? It was easy my dear, don't forget I spent two years as a building contractor. - Priscilla Presley & Ricardo Montalban

 

[slickdeals]

In Page 3 of the pdf under "The use of metal deck formwork"

Yes, using it in conjunction with a band beam with bondek spanning the short way.

 

[rowingengineer]

so a band and one way slab solution, not a two way as i would describe.

How could you do anything so vicious? It was easy my dear, don't forget I spent two years as a building contractor. - Priscilla Presley & Ricardo Montalban

 

[hokie66]

Maybe they mean you can use the steel deck as form for the bands as well as for the slab. That can be done. Still a one way system.

 

[asixth]

This was a long thread.

11x8m grids, are the 11m spans band beams? Then yes, design as one way but put enough flexural reinforcement over the support in the secondary direction because the structure will naturally want to behave that way.

I heard a saying once when studying strut-tie that the structure will want to behave the way it want's to behave or something similar.

I can post photos of a cracked slab similar to what you described with no reo in the secondary direction, but it's the same photo which got me into troule the first time.

Anyways good to be back.

T-beam for flexure checks.

Also check the beam shear (one way shear) but take the width of the band, it will be ample.

 

[AndyBWilson]

Slickdeals - I would suggest using a different software to those who utilize the finite element method (FEM) in post-tensioned slab analysis. The boundary element method (BEM) diminishes almost all the mishaps that appear in the FEM.

For example, columns are represented by their true geometries in the BEM, thus no fake singularities around columns in contour results of the slab. Same goes to beams, the actual beam width is a factor. The BEM requires you to place your elements on the boundaries too, no internal meshing, saving time on the effort put into choosing a suitable mesh as done by users of the FEM software.

A software that uses the BEM is the PLPAK. It is a new developed software that supports the analysis of shear deformable plates. Post-tensioning has been added to it around 3 years ago, and the results have been compared to those of the FEM software since then. What I love about the PLPAK is the accuracy it has at solving punching shear, much better than any software I have come across.

There are much more advantages to using the PLPAK in post-tensioning than other software like ADAPT and Ram....check out all the details on its website:

http://be4e.com

 

[rapt]

Slickdeals,

You get some 2way action in all beam/slab systems. There is always some stiffness in the slab in the transverse direction. The amount of 2way action you get will depend on the relative stiffneess of the slab to the beam in that direction. Thius can be allowed for in any design as long as the detailing of the system is consistent with this in providing the proper 2way load path. In doing this, the main slab design must allow for a concentration bear the support to allow for the "column strip" created in the other direction as the slab is not considered to be carrying load in both directions, so the transverse load has to be carried to the support by this "column strip".

Once metal decking is added, the transverse slab stiffness will reduce markedly, depending on decking type used and slab thickness. You could easily lose 1/3 of the slab thickness resulting in a stiffness reduction of the cube of this, so lots! So the amount of 2way action would be minimal, and you have complicated the detailing a lot.

I do not think it isd worth the effort!

AndyBWilson,
Your posts would appear to be bordering on the inappropriate! I hope you can justify your comments regarding punching shear solutions. We software developers are not supposed to be "marketing" in out answers!

 

[rapt]

Slickdeals,

Apologies for the mistakes in the last. Fixed below (it is early Monday morning!!)

You get some 2way action in all beam/slab systems. There is always some stiffness in the slab in the transverse direction. The amount of 2way action you get will depend on the relative stiffneess of the slab to the beam in that direction. This can be allowed for in any design as long as the detailing of the system is consistent with this in providing the proper 2way load path.

In doing this, the main slab design must allow for a concentration near the support to allow for the "column strip" created in the other direction as the slab is not considered to be carrying load in both directions, so the transverse load has to be carried to the support by this "column strip".

Once metal decking is added, the transverse slab stiffness will reduce markedly, depending on decking type used and slab thickness. You could easily lose 1/3 of the slab thickness resulting in a stiffness reduction of the cube of this, so lots! So the amount of 2way action would be minimal, and you have complicated the detailing a lot.

I do not think it is worth the effort! Normally it is ignored in beam/slab systems because we want the simplicity of obe way systems to construct. It is even more the case with slabs with metal decking where the benifit is so much less.

AndyBWilson,
Your posts would appear to be bordering on the inappropriate! I hope you can justify your comments regarding punching shear solutions. We software developers are not supposed to be "marketing" in our answers!

 

[AndyBWilson]

rapt,
Thank you for the heads up, I really appreciate the support from you since I am new here.

I was just pointing out that the boundary element method can solve the near exact punching shear around any columns in slabs. That is possible because the whole geometry of the column is taken, not represented as a node like in finite element software.

I just think you could have a go at using the PLPAK once and giving me your opinion on it. I believe it is more economical to use it since no singularities appear in the slab region when presenting moment contours and so on.

Most people here are involving the finite element method in their discussion. You could try to consider the boundary element method too since it competes with the previous mentioned method.

If you still see this as inappropriate, please explain. Thank you.

http://be4e.com/

A more realistic approach to structural modeling!

 

[AndyBWilson]

rapt,
I will check in more for why boundary element method developers think that punching shear is more accurate using their software and reply as soon as I can.

http://be4e.com/

A more realistic approach to structural modeling!

 

[rowingengineer]

Andy,
I can't see it myself, as far as most codes go the shear design is based on empirical methods, so all you have is the M* and V* applicable to the design. Punching shear in my opinion should always be done by hand. Sorry Rapt that goes for your program too.

How could you do anything so vicious? It was easy my dear, don't forget I spent two years as a building contractor. - Priscilla Presley & Ricardo Montalban

 

[AndyBWilson]

rapt and rowingengineer,

The codes use empirical methods because the FEM is deficient as you know. I check with the developers and found this:

The shear calculation is ignored in the basic derivation in the finite element analysis, then it is calculated backward using simple equilibrium equations for every element.

However, this is not the case in boundary element formulation, where the shear variables are solved in the boundary and internal solutions directly.

The main difference is that the finite element works on the derivative of the elasticity equations (when u deffrentiate (x2+2x+3) u get (2x+2) so u r loosing terms) , but in boundary element we work on the direct integral form, so all the variables are kept.

 

[hokie66]

Andy,
I had the same initial reaction as rapt about some of your posts, as they seemed to be promoting the be4e software. We are naturally suspicious of salesmen. Perhaps you can clarify if you are an employee of that company so we can know how to take your posts. We have quite a few software developers on the site, and their technical input is valued, but full disclosure is necessary.

 

[AndyBWilson]

hokie66,
Thank you for the feedback, I am not a developer, just a user (4 years until now) and I prefer using the PLPAK because of some things it facilitates, and I have done several practical examples such as foundations and post-tensioned slabs with it. I was asked to specify the reason why punching shear would be better, and I asked the customer support on their website yesterday and replied with the answer I got from them today.

 

[rapt]

RE,

Agreed. RAPT provides a calculation, but the engineer needs to look at it and possibly expand on it for each case, but that SHOULD be the same for all calculations done by all software!

AndyBWilson
I am very doubtful about your punching shear answer. Punching shear is basically based on the column reaction and moment. Are you/they saying the FEM results for those are that wrong? Sounds like propaganda to me (and I am the last person on this site to defend FEM as most would know). Punching design is normally the application of an empirical code formula to a shear head affected by a moment and a reaction as RE said.

 

[rowingengineer]

Andy,
I still don't see it, you may get better moment distribution, but this may help for punching shear, but again it is an numerical formula that is the basis of the calculation. I know that Mohammed Nazief, et al has published a paper suggesting the moment parameters are better defined than the ACI code, but not the complete punching shear calc as such, personally I thought his paper was very poor on the subject as he didn’t include any punching shear test results only did a paper comparison. In his paper, I think he he was relying on a numerical analysis for shear component.

I am not saying that BEM doesn’t have it good parts, as you are reducing a integration but don't be fooled into thinking that all the problems with FEA are removed by BEM. The post processor part of the equation is generally where all the errors are hidden. The M*,V* and N* are generally not all that different for all analysis methods (ignoring a few doggy ones form the ACI) if used by an experienced person that has experience with the method.

Rapt,
Agreed, I like to do a separate hand calc check of all answers (a few rules of thumb), but punching shear is the only calc I always do completely by hand.


How could you do anything so vicious? It was easy my dear, don't forget I spent two years as a building contractor. - Priscilla Presley & Ricardo Montalban