Specifying Concrete Shrinkage

[njkeng]

I work in Australia and one practice that confuses me, is the specification of 56 day shrinkage microstrain on concrete on the drawings (specification notes). I am interested to know how these shrinkage values are derived, what are appropriate shrinkage strains for different applications etc. I have tried to look up the AS3600 commentary but there is no commentary available for the latest version (2009) of the code, so can anyone provide advice or point me to a reference which discusses these points?

Thanks.

 

[IDS]

The commentary for the current code was published at the end of last year, and is available from SAI Gloabal.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[asixth]

Yup. 650 microstrain get's specified for post-tensioned slabs and 750 microstrain normally for most other applications. It is nice when the contractor actually undertakes and submits the drying shrinkage results. Usually it doesn't get submitted.

I'm just looking over some of the specifications from the previous consultants that I have worked at and found a 450 microstrain specified for columns and walls in multi-storey buildings.

 

[rapt]

The test method is defined in the concrete materials standard. Reading it will probably tell you all about it!!

The important point in all of this, explained in the Commentary mentioned above, is that the 56 day shrinkage value from the tests is not equivalent to the codes "final drying basic shrinkage strain" with default values of 800 to 1000 micro strain. The code values are 30 year values. The 56 day test value needs to be converted to a 30 year value as shown in an example the new Commentary. 650 will probably really be closer to 800-850 and 750 closer to 950 as a starting value for design.

So, if you are using RAPT for design, you do not input the 650 value in the materials properties, you need to convert it to a 30 year value. We are adding a converter function to do this for you.

 

[njkeng]

Thanks guys for your input. I did get the commentary to AS3600-2009. It does provide an example to calculate the 30 year shrinkage strain from the 56 day shrinkage strain which is helpful. Usually in your design though, you would incorporate the shrinkage into the design prior to getting shrinkage tests performed, so you would typically determine your 56 day shrinkage value using the code and specify that as a maximum limit to be met in the specification?

I was also interested in why the shrinkage values would be less for columns than for beams or slabs.

Do we approach creep values in the same way as shrinkage i.e. get testing done to identify creep which verifies the creep values determined in the design? I personally haven't seen this approach in specifications, but it does seem to follow the same logic as shrinkage.

Both creep and shrinkage do seem important as they effect the long term behaviour for deflection checks and do inform reinforced concrete detailing particularly at restraints and changes in geometry of concrete members.

 

[asixth]

Don't know why they use to specify a different shrinkage value for columns and walls. I think it would have to do with high-rise columns where differential shortening of the vertical elements can be a problem.

I think the materials properties written into AS3600 for shrinkage, creep, E-modulus etc. were all based on materials relevant to the Australian suppliers and what get produced in the industry.

 

[rapt]

Most designers will simply specify Standard Concrete and place no limit on shrinkage. So the values in AS3600 would then apply for different areas. Unless you know that those figures cannot be achieved from the suppliers in your area, in which case you should use local values.

Creep tests can be performed. If you want to. Boral has done this with its new Envisia concrete in Sydney. But most additive suppliers simply test shrinkage and do not worry about creep. Not sure why, other than that consultants probably only ask for control over shrinkage properties.

 

[IDS]

I guess people don't specify creep tests because:

1) Shrinkage tests are easier/cheaper
2) No-one else does it
3) Creep is seen as being less variable than shrinkage.

I don't know that 3) is true. Certainly Bazant doesn't seem to think so.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rapt]

I hope no one is requesting 750 in Brisbane/Sydney and then using that in the design, or even worse specifying 750 and then using the code default 30year value of 800 in design, as the real value will be closer to 950 at 30 years for a 750 56day test value!

 

[asixth]

I have.

 

[rapt]

Asixth,

Since 2009, the code in clause 3.1.7.1(b) has specifically said that the design shrinkage strain can be determined from tests after 8 weeks of drying "modified for long term value".

 

[njkeng]

Thanks for your input guys. It seems to me that this topic is understood and practiced in different ways. My experience has been that whilst the shrinkage limits have been specified on drawings (56 day max shrinkage), the reasons for it and the calculations used in the design are seldom if ever put together. I remember having asked a senior engineer previously where these values came from to be told that they weren't sure, but that they had been determined by someone else previously and that was good enough. Based on this discussion I think that would indicate a disconnect with the design of the structure which is a bit concerning given the specifier is responsible for this, and is in the best position (in most cases possibly the only position) to make a judgement on this.

I have appreciated this discussion and it does give me some direction in considering shrinkage and creep in my design and especially reinforces the point that if you request a certain shrinkage limit, you should understand why and know how it effects the structure you have designed.

 

[rapt]

njkeng,

When designing using L/D ratios and simplified deflection acceptance criteria, there is no way to account for shrinkage directly. But consultants will often nominate a shrinkage value that is lower than the code default value so that they have a little up their sleeve.

The misconception in the industry has been that the drying basic shrinkage strain is the 56 day test value. That has been addressed in the current version of the code and the commentary. The problem is getting industry in general to understand this. The 2001 code did not specify the difference, just left it to engineers to figure out that the basic shrinkage should be a 30 year value, not the 56 day test value. That is why the wording in the code was changed.

Many designers using RAPT do define the shrinkage in RAPT to correspond to what they are specifying on the drawings. Many PT companies specify lower shrinkage concrete and allow for it in design as do many of the larger consultancies with RC concrete. The problem is to get them to define the correct value, the 30 year estimate from the 56 day test value! That is why I am changing the RAPT input to do the conversion to 30 year value from a nominated 56 day test value for them.

 

[IDS]

The problem is getting industry in general to understand this.

Or even to know that the commentary has been published!

I forget where I found out that it had been published, but it certainly wasn't from SAI Global. From my brief examination so far it seems to be an excellent document, and certainly not one that should be ignored.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[Trenno]

And for only $225.06 it can be yours!

 

[IDS]

Trenno - you might want to check out:

http://www.concreteinstitute.com.au/Australian-Standards-Online.aspx

This gives you on-line access to all Australian Standards for $220/year (for Institute individual members). You can download pdfs of any standard, but these expire after a short time (2 hours I think). There is no limit to how many times you connect or download though.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[njkeng]

I am grateful that Doug let me know that the commentary had been published. I have got it and I agree that it is a good source of information in understanding the standard. I am surprised it took 5 years! I would have loved to of had it earlier. It gives a bit more info on the Fire Design requirements too, but that is a discussion for another post.

 

[IDS]

Converting the test shrinkage to a basic drying shrinkage is not difficult, as long as you remember what to do, but I have added a function for it to my Concrete Design Functions spreadsheet anyway:

https://newtonexcelbach.wordpress.com/2015/03/21/getting-concrete-shrinkage-right/

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[Trenno]

Doug, finally got my hands on the commentary - certainly a great reference indeed. I like included visuals too.