Are column strip middle strip assumptions required to model 2 way slabs that are post tensioned? 3

[thejonster]

Column strip middle strip assumptions concentrate moments & shears in ultimate, and stresses in service for mild 2 way slab design.. should this also apply in prestressed/ Post-tensioned design?

I was very interested in ACI 318-14 researching this question because the re-organization put mild and prestressed 2 way slabs in the same section.
I found in ACI 318-14 8.4.1.2 that prestressed 2 way slabs shall be designed by the equivalent frame method, and R8.11.2 that the equivalent frame consists of 3 parts: column, joint, and slab beam strip. The slab beam strip is shown as 'full width' or same width as the tributary area for loading. Same as ACI 318-11

the equivalent frame section of 318-14 (8.11) has no guidance for concentrating the moment or stresses in the column strip, and the direct design section where those forces and stresses are concentrated is prohibited for prestressed design.

RAM Concept support has a page online: Frequently Asked Questions
When should design strips be modeled with column and ½-middle strips and when should they be modeled as full-width column strips?

https://communities.bentley.com/pro...esign__wiki/4507/ram-concept-design-strips-tn

They advocate that col & mid strips be modeled, concentrating the moment and stresses to the col strip.

I spoke with an Adapt designer who says that 95% of 2 way slab designers model as 'full width', and he pointed out that ACI 318-11 chapter 18 says ch 13 does not apply, and he pointed out that the equivalent frame has nothing to do with col strip mid strip, but rather the vertical frame including columns floor to floor.

Applying csms assumptions will make the design much harder and less economically competitive vs soil treatment under a basement. There is no life safety risk in our particular application as this is a residential basement floor on pier supports.

I have primary experience in 1 way garage pt slabs and limited experience in 2 way design.

Edit:
This is for a 2 way flat plate design

 

[rapt]

There is a difference between "should" and what ACI allows and what other codes allow.

Historically, ACI designers have designed upbonded PT flat slabs using a whole panel width concept. This stemmed from the use of unbonded PT which is very hard to lay in a weaved pattern, so designers basically adopted a one way PT pattern, with banded tendons in one direction over the columns and distributed tendons in the other. This can sort of justified by yield line design principles but designers have not used yield line analysis for the design, they have used flexural design. While this assumption holds for rectangular grids of columns, it does not for random column grids but we cannot let good theory get in the way of a large industry groups interests. The further problem with this logic, as is always the problem with yield line) is that to get realistic service stress results, the designer needs to look at the effects of local stress concentrations. For both crack control and deflections. For this reason ACI does not allow partial prestress design using this method, but the allowable stress limits it has set are unrealistically high and result in cracked members even though ACI suggests that will be class U. This method also requires some large minimum reinforcement requirements in support regions to allow for the inaccuracy in the design method.

The other approach, as used exclusively in Australia and Asia for many years, and often in Europe, until recently where some designers have been playing with PTI methods due to the spread of ideas on the internet and in software, the analysis is based on the same distributions between column and middle strips as for RC flat slabs. Thus the correct stress distributions are accounted for and partial prestressing and reasonable deflection estimates can be made. This requires basically that half of the tendons in each direction are equally spaced and the other half are concentrated over the columns creating strong bands in the column strips. To achieve logical profiles, these tendons often have to form a weaved pattern. You cannot simply lay one direction and then the other over the top.

It is interesting that RAM Concept are now suggesting column/middle strip, because when we were having discussions with them in the early 2000's there was no concept other than banded/distributed. It is nice to know that they have finally agreed with us. I will not say what that tells me about Adapt designers!

 

[DETstru]

rapt, Bentley still suggests full width strips for PT design in Concept along with the banded and distributed layout. That's how it's done in the US (as far as I know). Adapt has the same stance. I'm not sure where thejonster heard that.

I've only once weaved tendons once and that was a special case where I was forced to in a small portion of slab. I've never seen it done in the US otherwise.

 

[thejonster]

Thank you guys, this is really helpful. I started looking into this because concept has options to model the slab strips as full width or "code slab" (column strips and middle strips) which made me think we weren't following the code by modeling full width. Bentley's own page in my first post said that col strips were appropriate to account for concentrated moments and stresses, if I had read closer I would have seen that they emphasize full width is appropriate for post tensioned slabs

 

[hokie66]

rapt is too polite. The US way of building unbonded PT slabs, banded and distributed, is stupid. In my opinion, of course. Just speaking from the experience of a US engineer who has worked in Australia for a long time.

 

[rapt]

Ah well, we can only hope for a day in the future then! On many fronts the way the world appears to be going.

They initially only added the column/middle strip approach to appease other markets, along with an attempt at real long term deflections. From thejonster's comment I was hoping they had finally seen the light.

Though people in US are wanting to do partial prestressing in flat slabs! They cannot do it until they accept column/middle strip at the least. But that means moving to bonded PT which the PTI will never do and Ingenuity will continue to be able to fund his retirement repairing the unbonded ones! Until the PTI comes up with another fudge design method to justify partial prestressing based on average moments. That one will be interesting to see.

 

[rapt]

Hokie,

I thought they had enough on their minds over the last year or so without being my normal blunt self so thought I would take it easy on them and give a controlled overview. I am sure they can search for some of my more expansive dialogs on the topic.

Unfortuntely, from what I am hearing, the disease is spreading. People who do not understand design and who cannot read codes are using software that encourages it so are starting to do it here. We are trying to figure out how to stop it. Just putting it in the code seems to have no effect any longer. What is written is NOT what is practised. Not sure if they cannot read or are simply unethical. Someone telling me about some cases yesterday and today suggested it was deliberate and therefore the later.

 

[DETstru]

I seem to be lacking the international experience in PT you all seem to have but to perpetuate the USA jokes...

http://www.post-tensioning.org/Uplo...y Slabs Using a Dual Banded Tendon Layout.pdf

If you don't like banded, how about banded in both directions!

 

[hokie66]

Well, yes. If you want to build flat slabs or flat plates, the tendons should be banded in both directions. But that in practice leads to bonded prestressing, as the weaving of strands is cumbersome, but the weaving of ducts is simple. If you want a one way system, then use band beams in one direction.

 

[rapt]

Dual banded is OK, better than banded/distributed) as long as the slab is reinforced in the bottom away from the bands for the full positive moment with normal reinforcing. (I have assumed this is dual banded only, no distributed tendons in either direction like there are in the column/middle strip logic).

You cannot make the assumption that the bands are providing the positive capacity in the middle strip.

I like his deflection comparison looking at elastic deflections, in a slab that can only work based on massive redistributions from the elastic state to the cracked state.

 

[IDS]

rapt is too polite.

You don't see that very often

But seriously, interesting discussion.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rapt]

Thanks Doug. For that, you can buy the drinks in Adelaide next month!