Long Term Concrete Deflection due to Shrinkage - ACI 209 and RAM Concept? 1

[sticksandtriangles]

I am trying to get a good estimate for my long term deflections on a concrete podium slab I am designing.
From my brief reading on the topic, long term deflection calcs can be a mixed bag in accuracy (so many variables at play that it is hard to accurately predict).

Anyways, I am noticing that the shrinkage strain input within RAM concept has a big impact on my long term deflection calculations.

I am calculating this shrinkage strain based on ACI 209 (assumes a standard condition of shrinkage strain and apply correction factors) and I have 20in thick slab that has will be curing in a high humidity environment. From my understanding of calculating the adjustment factors per ACI 209, these items (thick slab and high humidity) push my adjusted shrinkage strain very low (0.000237). ACI 209 recommends states "normal ranges from 415X10^-6" to larger numbers.

I am not a long term concrete deflection expert and I want to ensure that utilizing my calculated shrinkage strain is not too low for practical limits. 0.000237 is about 1/2 of the "typical normal ranges" and this low parameter really helps my long term deflection (trying to limit edges of my podium to 0.3" max due to brick being supported above).

Any thoughts on the topic would be appreciated!

Thanks,
S&T

 

[IDS]

On one hand, all my experience has indicated that you should be conservative in every respect in calculating deflections; and that includes not using the ACI methods for calculating the effect of tension stiffening or the effect of shrinkage.

On the other, it is quite possible that the brick wall will interact with the concrete in such a way as to reduce deflections. That's outside my area of experience, but I'd suggest finding someone with long term experience in this area to discuss what works in practice.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[sticksandtriangles]

Thanks for your thoughts IDS.

Curious what makes you state:

that includes not using the ACI methods for calculating the effect of tension stiffening or the effect of shrinkage

 

[IDS]

sticksandtriangles - the two issues I had in mind are:

At moments just above the cracking moment the Eurocode gives a much faster increase in curvature than the ACI equation, especially for lightly reinforced sections, and this is backed up with many research results.

For sections with equal top and bottom reinforcement the ACI method does not give any increase in curvature due to shrinkage before or after cracking, but shrinkage does have a significant effect after cracking, even if the reinforcement is symmetrical.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rapt]

IDS is correct on this.

Branson's formula suggests no sudden loss of stiffness on first cracking. There is a kink in the rate of deflection, not a sudden jump on first cracking as you would expect.

it is pretty good for reasonably heavily loaded members but unconservative for lightly loaded members that are only just cracking.

And the compression reinforcement effect is a joke in many cases for both creep and shrinkage.

 

[sticksandtriangles]

Rapt,

I appreciate your input.

I've got to do some more reading on long term concrete deflection clearly, a lot of this "lingo" being thrown around is like a foreign language to me. Any suggested reading would be appreciated. I have read the ACI 209 document but would appreciate references to the other publications as well.

Thanks!
S&T

 

[Ingenuity]

Time-Dependent Behaviour of Concrete Structures" Gilbert and Ranzi: Link

Or, a bit of an older edition of the above: "Time Effects in Concrete Structures" also by Gilbert.

 

[Settingsun]

I think you should get some local data which concrete suppliers may provide to you. It varies widely eg the Australian code gives a 25% difference between major capital cities for the basic strain before you apply any design-specific factors. Then you dig into the commentary and find that aggregate can cause a 250% range. (I think everyone replying so far is Australian so you're not getting US knowledge.) Once you have the data, specify a realistic shrinkage limit and design based on that.

237 microstrain does however seem low for a worst-case analysis but I can't say wrong because maybe you have good aggregate. A couple of British bridge textbooks I have give low-200s typical numbers for your situation while the Australian code is more like 400~500. Your ACI code probably has a warning somewhere that shrinkage will vary +/-30% from sample to 'identical' sample.

I will ask whether the environment will always be humid, or will air conditioning be installed next month and suck all the moisture out of the air?

I would also suggest a risk-based decision. What are the consequences for you/your client if the brick wall cracks? That should guide how conservative to be if you're at the mercy of an unpredictable parameter. As an example, I once had precaster data for nominally identical bridge beams (about 30 "Type-D" beams so same geometry, prestressing, concrete mix etc). The hog deflection measured at 90 days varied over a 2:1 range. A refined analysis using the methods given by Gilbert & Ranzi (Ingenuity's reference above) gave a number near the average. Playing with Young's modulus, creep, and shrinkage over +/-30% could account for the maximum and minimum measurements.

 

[SethGuthrie]

Here is a post on Bentley Communities regarding long term (load history) deflection and AS 3600 in case you have not seen it:

https://communities.bentley.com/pro...iki/36046/load-history-parameters-and-as-3600