RAM Concept Load History Deflections to AS3600 3

[VGEngStruc]

Hi all,

Long time reader, first time poster.
I have a question about RAM Concepts load history deflections to AS3600.
I have read quite a few threads regarding this topic, but being a young engineer, still require some more advice.

I am modelling an RC flat slab, 235 thick, standard residential loading, 0.2 bending stiffness into columns.
What would be the most accurate load history criteria that would give me most accurate long term deflections?

I currently have the following;
Shrinkage restraint: 10% (Judgement call based upon geometry and layout of columns and shear walls)
Creep factor: 3.8 (2.8 + 1)
Shrinkage strain: 0.0006 (N40 Conc.)

Any advice will help...

Thanks in advance!

Regards,

VG

 

[Trenno]

Hi VG,

This a fairly complex topic with so many different factors that can influence long term deflections. Keep in mind even with best estimates of all parameters, you can still be out by +/-30%.

Firstly have a read of a few threads that relate to your questions above.

Link 1
Link 2
Link 3

10% is reasonable with a single dominant restraint (eg. lift/stair shaft) and relatively flexible columns. There is a great reply by a tech guru in the first link that explains this factor thoroughly.

Over the past few years I too have been very interested in tweaking the RAM Concept Shrinkage and Creep (C&S) factors to more align with AS3600. RAM Concept uses the ACI C&S models and it's not very straight forward when comparing with the AS3600 C&S models. How we calculate various parameters is quite different to ACI models and methods.

I'm currently in the process of developing an in-house RAM Concept template (based upon AS3600), which I hope to address the proper use of each Load History factor. However, for now... I'm just using the default values and changing the time of first loading to 7 days. I like to verify deflections (moments, shears too) using 2D programs such as RAPT. I highly recommend you do the same - it gives you a good feel for how a certain arrangement is working and gives you confidence in your design.

Another word of advice, would be where possible try and model the PT and rebar as per the final design (after detailing and curtailment), this will provide a much more accurate snapshot of the floor's effective stiffness and load redistribution (if any, after cracking) and therefore long term deflections.

 

[VGEngStruc]

Hi Trenno,

Thank you for the advice and the threads!
What makes me uncomfortable is how sensitive my results are to the factors such as shrinkage restraint, and that I am using a ballpark figure as an input.
Another model I am running is a large flat plate with columns at roughly 8m spans. For this, I found that a 5% shrinkage restraint works well.
I have also been using RAPT to verify my results, but am still hesitant to use the values RAM Concept produces.

It's a shame that RAM Concept doesn't incorporate AS3600 C&S models.

I will attempt using the default values and changing the time of first loading to 7 days to compare with my previous runs and RAPT.

 

[asixth]

Good advice Trenno.

I typically use a restrain factor of 20%, default shrinkage strain and ECR of 2.9.

I find that the contour plots give a good correlation with the concrete behaviour. I had a slab I designed recently completely surveyed along top of slab and soffit cause they thought the slab was deflecting too much. Turns out it was a support settling that was causing the feeling of "deflection". When I got in there and wrangled all the survey points it actually correlated quite well with the contour plot of ram.

Watch the long term loads or "sustained" loads. Most of ram will default as 0.6*LL whereas AS3600 requires 0.4*LL for most occupancy's.

I also ditch the 30 days initial service which uses DL+LL or DL+0.7*LL (I forgot which one) and just go with 1 day followed by 10000 days of sustained service and then 1 day full live load at the end.

 

[Trenno]

Here's a question for you all...

How long is 'long term'? 10 years? 30 years? design life of the building?

I've never actually seen this value quoted anywhere.

 

[SethGuthrie]

A lot of creep data comes from 20 year measurements. The nice thing about load history analysis in Ram Concept is that the user can enter any amount of time (along with the active loads during that time) and the program adjusts the creep accordingly.

 

[rapt]

I have not looked at RAM for years so do not know if this has changed, but previously the "restraint" check did not consider shrinkage, only restraint to elastic shortening from P/A.

Are you sure that it now allows restraint to shrinkage shortening as well now?

What about temperature, because this can be as much again as shrinkage, eg if you pour on a hot summer day at 35C and the winter night temperature could be negative you could easily have over 400 microstrain from temperature!

I have also heard from a consultant who doubted the effect of reduction in column stiffness in RAM on his deflections. Again, this was several years ago and it is 3rd hand. Have you compared the effect of the change in stiffness?

 

[Trenno]

RAPT,

Shrinkage Restraint is considered these days. A brief explanation can be found here.

I think you should be able to take temperature into account by modifying the shrinkage strain. I had a quick look at the modifications can be made using equations in the ACI209 model. LINK

In terms of column stiffness, RAM models columns as a single support node with bending stiffness. Therefore the slab can actually deflect 'within' the column. Few ways around this is to model a deeper slab section with the plan dimensions of your column or if you have a blade column to model them as walls.