PT Transfer 'By Others' 1

[ljk80oze]

Hi

In my region post tension slabs are undertaken as design and construct packages and engineers with PT transfer slabs provide loading plans / outputs for other consultants to undertake this component of the design. Whilst i have next to no experience at PT design I strongly dislike having such a major component of the project being handled by another consultant.

I would really appreciate some advice regarding any measures i should be taking to validate the external consultant’s design and to ensure that the rest of the building is not effected by PT slab movements etc..

The particular project that i am working on is a 5 story residential building with a basement carpark slab, transfer slab at ground and an additional 3 levels of suspended slabs above. The building footprint is about 35x50m. The client has opted for a PT transfer with the rest of the project remaining conventional insitu concrete.

Questions:
Is it appropriate to estimate the PT slab thickness (flat plate slab) by determining the punching shear thickness required for a conventional slab?

What measures are required to control shrinkage, PT shortening, when are they required and what effect does the lateral stiffness of the supporting structure have?

The transfer slab is supported by perimeter basement retaining walls and internal columns between the carparks. I have concerns that the large restraint provided by these perimeter shear walls wall result in restraint cracks in the corners of the transfer or in the wall elements themselves. The client also advised that the previous job they did with a PT transfer had double circular columns instead of conventional rectangular columns to reduce column stiffness and prevent column damage due to the PT movement.
Will I require pour breaks / expansion joints etc to control shrinkage/shortening stresses?? How do I determine whether basement walls and columns will be damaged by the movement (do I apply a deflection to the structure and determine if the induced moments exceed the cracking moments of the individual elements?).

Finally, should I request deflection, reactions etc results from the designer to verify my design assumptions (load distribution)? Is there anything else I should know?

Thanks for your time!

LJK

 

[hokie66]

LJK,

I share your dislike of dividing the responsibility of design by having someone else design a critical element like a transfer floor. I suppose it has evolved this way in many cases because PT design is still somewhat of a specialist design field, and that in itself is to be lamented.

You are concerned about the correct things. I would make my requirements for the other consultant's supplied information as extensive as possible, and hold his feet to the fire.

Punching shear is often an issue with PT transfer plates, both from column loads above and at supporting columns, but the span lengths, deflections, etc. also play a part.

For your 35x50m footprint, I would look at casting it in 2 pours with a delayed pour strip across the short direction. So you would have 2-35x24m pours, with a 2m pour strip.

Some cracking in the transfer plate is probably inevitable unless you isolate the plate from the walls. This can be done by using corbels on the walls to be later tied to the slab. Insist that the PT designer tell you what will crack and how much. Then you (and possibly the owner and architect) can make an informed decision as the consequences of the cracking. Cracking of the walls is usually not an issue, because the walls are placed into compression by the shortening of the plate. Columns, especially very short columns, can be problematic, especially near the extremities.

Where is RAPT when you need him? He is great on this type question, but haven't seen him on the forum for a while.

 

[rapt]

hokie66,

He is working very hard in UK teaching people about RC and PT design and the real way concrete works, not the PTI/FEM way, especially for crack control and deflections, and how to use a product he develops that he is not allowed to mention on this site for self-advertising reasons.

I normally only answer when I feel I have something worthwhile to contribute.

LJK,

I agree that consultants should design their own PT, but do not rely on software to do it for you or or the PTI manual to tell you how to do it. Software will not design it, just do calculations for you, the designer. You need to gain a good understanding of concrete design and load paths and flat slabs from first principals before getting into it. And you need to do it by hand to get a good feel for it. Then you might get an good undserstanding of some of the code and industry methods published around the country and their limitations. Once you really understand how it all works, then use software to do the calculations for you to help you do it faster, but you are still the designer and it is up to you get the correct software and to drive the software the way you want it to work if it has the capability to do it properly.


According to the codes, PT punching shear is better than RC in most cases (except to British codes) so estimating it this way for a transfer slab should be ok.

You also need to worry about real stresses and deflections (not average moemnt ones).

If your only stiffness/restraint worry is the edge walls then yes, separate from them with pour strips or other methods. If the builder is happy to pour the slab in one go, I would not worry about the central joint as long as you can get stressing access on all 4 sides. Connect the edge walls in after stressing and all will normally be ok.

The actual shortening caused by the prestress is only about 10% of the total shortening. The other 90% will be caused by creep, shrinkage and temperature change. This would also happen in an RC slab but the RC slab has better crack control as shrinkage cracking in the slab is controlled by the closely spaced bonded reinforcement while in the PT slab it is resisted by the PT until it cracks and then, with unbonded PT, the cracking is unrestrained so make sure there is reinforcement in the slab to resist any long term shortening if it is significantly restrained by the walls after they are connected. A transfer slab should have bottom reinforcement everywhere anyway and top reinforcement over the columns.

For these sort of slab dimensions, shortening should not crack the columns excessively as long as the axial prestress is relatively low, 1.5 - 2.2 MPa (200-300psi). If the numbers get too high then yes, you need to worry.

Re design data from the designers, for a transfer slab, make sure they do not average the design effects over full panel widths, there will be concentrations of moments at columns above and below and make sure they consider them by breaking the slab down into smaller design widths to reflect the concentrations. And make sure for the deflection calculations, they allow for these concentratiuons as well as cracking and proper long term effects.

And remember in the end, the elastic moments in a PT slab before cracking are exactly the same as those in an RC slab with the same loadings, so if it looks wrong from an RC perspective, it probably is as the same moments have to be resisted by whatever steel you put into it and a crack caused by a tension stress can only be resisted by bonded reinforcement at the location of the stress, not by uncracked concrete or a bar 6 m (20') away.

 

[hokie66]

Very good, RAPT hasn't fallen off the earth, just gone to the other side of it.

 

[ljk80oze]

Thanks hokie and rapt

I still feel a little lost though…. If I could ask some more questions… ?

Is it appropriate for me to estimate the total shortening based on conventional RC and add an additional 10%? Why are high levels of axial prestress more likely to crack the columns - is there more movement than 10%?
I put some numbers to the situation to get a better feel. I estimate the total RC movement would be approximately 20mm. Is it safe to take 25mm as a total amount?

Do I then apply 12.5mm to the extremities with a linearly decreasing value towards the shear centre (in each direction)? This seems appropriate for the columns but what about the walls? I don’t see how walls parallel to the shrinkage direction would permit any shrinkage movements (I have precast 3m wide panels). The force required to budge the walls is beyond the capacity of the slab. I take it this means that the walls don’t deflect and the slab cracks up to relieve the stress (sum of all crack widths = 25mm across entire slab). Since the shrinkage is relieved by slab cracking does that mean that internal columns will not offer restraint to the movement and will not crack?

So my options are to relieve the shrinkage by corbelling all the perimeter walls (the client will not like that! Any other suggestions?) or by providing sufficient shrinkage reinforcement to limit crack widths (more little cracks rather than fewer large ones). My code (Australian) has minimum % of reinf. For RC restrained slab dependant on the exposure conditions. Do I have to insist that the PT slab has this level of reinf. In it? How do I gauge whether the reinf provided is sufficient (when combined with PT)?

If I nominate a pour strip, what is the appropriate time, width and reinforcing for it?

And finally what do you think of the client’s suggestion to use double circular columns (say 2 x 300dia instead of a single 1000x300) to the perimeters to provide some flexibility and reduce cracking?

Thanks immensely for your time and wisdom!!!

LJK

 

[hokie66]

LJK,
If you can't isolate the slab from the walls initially, cracking is inevitable whether the slab is PT or conventional. AS3600 gives you reinforcing percentages to use for crack control depending on the situation.

Pour strips are usually 1200 to 2000 wide, reinforced as required for strength, normally located in the centre third of the span, and left as long as possible before casting.

The flexible columns will crack less than stiffer ones.

You are on the right track. Cracking of slabs supported on concrete walls has given me the opportunity to do numerous investigations and reports. With your diligent attitude, your building won't have that type problem.