deflection in ADAPT PT/RC 1

[struct_eeyore]

Hi everyone,

For those who have used ADAPT to do 2 way slab design. We are having some excessive deflections, on the order of 2-3x what was predicted by the software. There is some settlement, but still not enough to account for the deflection. Understandably, there is no way I could present the entirety of model details in this brief post - I am just wondering if anyone has encountered any discrepancies in deflection calcs as done in ADAPT. If so, what was the nature of the problem?

 

[JAE]

Sort of similar to this thread: thread507-424281

I haven't seen that degree of difference in the past - haven't used ADAPT in a couple of years so can't say what the current version does.
How do the slabs compare with ACI's recommended minimum thicknesses for two-way slabs?

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[JoshPlumSE]

Could this be a difference between instantaneous and long term deflection?

 

[Ingenuity]

User rapt will undoubtedly reply.

Were you using ADAPT 2D equivalent frame or their 3D FEM software?

In my personal experience of using the software from late 1980's (the 2D equivalent frame version only) it inadequately accounted for creep and shrinkage effects in estimating long-term deflections. By at least a factor of 2, when compared to other software that we had 'calibrated' in the field from a 25 year old in-service PT structure that had excessive defection.

What magnitudes of deflection are you taking about in your structure?

In the one that I was involved where we calibrated the software, we had in-service deflections of 3+" in a 27 foot continuous spans.

 

[struct_eeyore]

Yes, our slab is within the depth/span limits set by aci. Interior panels are 12" thick, with maximum 30' panels.(No caps/beams) The slabs were just recently poured, the remainder of building is still under construction - this is not a long term creep issue. We actually put more steel than ADAPT specd. and bars were developed in excess of provided calcs. (Shearheads were also over designed). The deflections at the moment are between 1-1/2 inches (so we are pushing l/240 in some places) - however, the building has not yet seen close to the maximum load either... EDIT: just to clarify - this is a two way framed slab, not a PT slab

 

[Ingenuity]

If engineers design PT structures using the book "Post-Tensioned Concrete: Principles and Practice" by Bondy and Allred, then you will NOT check deflections. The text dedicates ZERO pages to the subject. Link

This book is destined to become the definitive work on the design of post-tensioned concrete buildings for many decades to come.

 

[Ingenuity]

Interior panels are 12" thick, with maximum 30' panels

So it is a RC slab, no PT?

 

[rapt]

Sorry. Ingenuity, I cannot add much. I know their long term estimates are wrong for PT as we have discussed previously, using long term factors which is not allowed by any code in the world for PT. But for RC, design codes allow it but it is very inaccurate.

But this is a short term problem. There have also been doubts about how it handles cracking for PT as well, but nothing I can confirm.

Structee
If you either post (or email to me privately) the slab span lengths, concrete strength, column information etc plus loading the slab has experienced, I could give you an estimate from RAPT for expected deflections. Based on the information you have supplied to date and assuming SW only, I would not expect that much deflection for an internal span. But it is very dependant on the missing information.

 

[Settingsun]

"The slabs were just recently poured, the remainder of building is still under construction - this is not a long term creep issue."

How recently? You can get half of the total shrinkage and creep in the first month for thin concrete elements.

I can't comment on ADAPT and have only used the ACI standard for structures that were never going to have deflection problems (container wharfs), but I understand from books that the ACI deflection provisions are pretty basic and not always conservative. I wonder whether ADAPT adopts ACI with all its issues.

 

[struct_eeyore]

rapt - I'm a little embarrassed, but I don't know how to pm in this form. I will have to get specs for the slabs monday or tuesday from the office..

 

[nonplussed]

rapt - I'm a little embarrassed, but I don't know how to pm in this form. I will have to get specs for the slabs monday or tuesday from the office..

Hi structee, would you kindly post the slab specs and continue this discussion on the public forum as I (and I'm sure many people here) would like to follow this and know the outcome. It's always interesting when we can have direct comparisons between real world deflections and calculations. Thanks.

 

[struct_eeyore]

Ok guys, the specs on our slab (typical strip) are:

3 spans north to south: 13'-0", 31'-0", and 24'-0". 22'-0" trib width. 12" thick slab. interior columns are 16x24", exterior are 12x24" (long dim. parallel to span. Currently, the slab has experienced a +/- 20 psf construction load, and the self weight. There are 10'-0" high cmu block walls at slab edges ( @ 0'-0" on the 13'-0" span, and @ 24'-0" on 24'-0" span) - these have been partially build, but we do not anticipate them to prop the center slab up much at this point. For simplicity, these 3 spans can be copied to the east and west (x3 in each direction) to create a representative building layout. The deflections we are getting are anywhere between 1 to 1-1/4" to maybe 1-1/2" inches on the 31' span (these are are somewhat in question due to the method the contractor is using to obtain the measurements - pulling a string). Let me know if this description is not sufficiently clear.

 

[rapt]

Structee,

Concrete strength, and how long has the slab been poured? And when was the shoring removed?

And column height!

 

[struct_eeyore]

Slab is 4000, but I believe our breaks were closer to 5000. Columns are 5000 (and again, I think the breaks were closer to 6000) There are 3 elevated floors, plus the roof which is the same system. the first elevated floor was started to be poured sometime in the middle of December. Each floor consisted of 3 individual pours separated by about 1- 1-1/2 weeks. I cannot give you exact times as I was no longer on site or following the project for the second half, but I believe they were done sometime by mid march with the pours. As far as the shoring - I believe each floor section was supported for at least a month; however, I was onsite only for threshold inspections, and have not way to know if any re-shoring was done. I would submit an rfi to the contractor, but things have more or less settled down by now, so I do not thing our office wants me to go stir up any more controversy, unfortunately... Thanks for looking into this![/b]

 

[rapt]

structee,

It is one of those line ball ones.

Under the loading you have suggested above, the slab is right on the point of cracking, but is uncracked. I would estimate a deflection of about .6" in that panel based on those factors at about 3 months after the pour.

If the loading is as you suggested and that loading was applied when the tensile strength of the concrete is at the 28 day strength and there is no restraint to shortening the above figure is what I would expect.

But if the tensile strength was 20-25% lower at the time of loading, or the slab was loaded early while stripping/backpropping or when supporting floors above a higher load was applied, because it is on the point of cracking, any of these effects could have easily resulted in cracking in the slab and an increase in deflections to 1 1/2". This cracking would not necessarily be visible and could have partly closed when load is removed.

If I reduce concrete tensile strength by 20-25% I get about 1 1/2" deflection. If I use full tensile strength but increase construction load to 60psf, I get about the same.

It is one of those cases where a small increase in load or reduction in tensile strength has a disproportionately large effect on deflections because the increase in deflection is not linear when the slab goes from uncracked to cracked (this is one of the major failings of the Branson method of allowing for tension stiffening).

 

[Settingsun]

Hi rapt,
When you say the slab is on the point of cracking, is that using the full cracking moment F'ctf*Z, or have you also included restraint to shrinkage? (ie would the AS3600 cracking moment reduction have resulted in a prediction of cracking?)

Also, what's the reinforcement? I'm wondering whether the 0.6*I maximum effective stiffness would have applied and given a 1" deflection prediction instead of 0.6".

 

[rapt]

steveh49,

Restraint stress from reinforcement have been included because they actually happen. Just because ACI does not mention them does not mean that they do not happen in the USA, or in concrete members designed to ACI. The ACI deflection calculations and especially the use of Branson's formula and long term multipliers are so out of date I can understand why no-one trusts them! The reinforcement ratios were not high so the stress caused by restraint would have been fairly small, probably .2 - .4MPa (about 30-60psi), but it was included.

The reinforcement required for ultimate strength was used in the calculations. Not sure how I could have used anything else. And yes, the .6Ig limit was used because tests have shown that you will never get more than that if the section is cracked. That is one of the major problems with Bransons formula. It predicts no reduction in stiffness and increase in deflection at the instance of the first crack. Tests (and simple logic) show otherwise!