Precast Prestressed Concrete - Beam Cantilever 2

[bowlingdanish]

Hello

Might be a bit of an odd question but I can't seem to find much online so here we go; I am working on an office building where the contractor is looking to use precast elements for everything. So precast floors spanning to precast beams to precast columns, the whole deal.

The building has extremely long spans and at the exterior the primary beams have to cantilever out a long way past the inset columns to support the floor. Since the spans are so long, we need to prestress the concrete to keep deflections in check with a reasonable beam depth. Which is all good for single span module elements, but I'm having a hard time getting my head around how we can prestress a precast multi span cantilever beam.

Pretensioning wouldn't make sense as the strands need to be straight, so the moment reversal could end up reducing capacity and increasing deflection in the cantilever. Unless... you somehow calibrate strand location at a section depth that would counter the expected deformed shape of the cantilever (i.e based on cantilever/backspan ratio)? Seems far fetched.

Post-tensioning the beams could also be an option, as you could drape the tendons to the moment distribution. But then, if it's tensioned in a factory setting, there'd be no supports to constrain movement (i.e no hyperstatic forces), so the beam would still want to deflect up as below;

In both pre- or post- tensioning, after tensioning the concrete, the relative vertical level of beam soffit at the column locations would not be the same (as they would for single span). Which seems like a nightmare with tolerances and installation specification.

Am I missing something/is this even possible?

 

[hokie66]

Posttensioning is uncommon in precast members. And precast multi span beams are uncommon. So if your job is to be all precast, start with the assumption of simple spans, with cantilevers.

Strands in pretensioned members don't have to be straight. They can be harped, and there can be strands both top and bottom. Middle too.

I am not an expert on precast construction, not even an exponent, but it sounds like you need a lot of help with this. Will the precast members and connections be designed by someone else? Perhaps the engineer for the precaster?

 

[BAretired]

I agree with hokie66. Continuous spans would be a nightmare to handle on site, particularly if the spans are long.

BA

 

[dik]

When I did Kingsway Mall Parkade in Edmonton, the precast Ts were prestressed as well as post-tensined... typically 60' spans with 3 and 4 of them post tensioned... Post tensioned slab over... Been there for 4 decades, so, seems to have worked OK.

https://www.google.com/maps/place/K...f1cbb49c8ea9cf7!8m2!3d53.560209!4d-113.505271

Dik

 

[bowlingdanish]

So if your job is to be all precast, start with the assumption of simple spans, with cantilevers.

Sorry I should have been clear - yes it will all be simple spans to corbelled columns, except for the cantilever which will require continuity over the column. The question is how this is achieved - hence the question to eng-tips

 

[hokie66]

dik,
Was the posttensioning in the top for continuity? Bonded?

bowlingdanish,
Top bars or top strands at the cantilevers.

 

[bowlingdanish]

Thanks Hokie- But presumably when you pretension the strands you need to continue the strands the full length of the member. So you'll have bottom strands for the backspan which must continue through the cantilever, and the top strands for the cantilever would need to continue through the backspan... I'm not aware of a way you could anchor strands at the support with pretension. Perhaps debond the top strands through the backspan and debond the bottom strands through the cantilever, with grease or sleeves of some sort? Top bars alone would not provide the required camber/long term deflection requirements.

 

[dik]

Draped to reflect BMD and also pretensioned using 'harped' strand. Post tensioning was grouted. Don't remember much about it (too far back), other than it was on an E sized sheet and needed 3/32 scale to get it to fit... I have no idea of how the strands were placed, but, the pre-tensioned Ts were post tensioned in place... or if the Ts were grouted where they abutted. Someone else in the company was looking after construction review. Was done for RJC... don't know how they are now, but, at the time they were one of the foremost reinforced concrete engineers in Canada.

Single storey and I designed the columns base plates for DL only then grouted pockets solid to carry DL and LL. Prior to an explanation, the office manager thought the BPl was too thin.

Dik

 

[rapt]

If you have all except the end beams as simply supported, they can be pretensioned in the bottom. You could use strands with double hold down points if you want to get rid of transfer problems at the ends.

For the end beam with cantilever , it depends on the length of the cantilever.

if the cantilever is not too long, then you could pretension these beams as well and simply reinforce the top for the cantilever.

If the cantilever is too long, then post-tension these beams. As there are only 2 connection points, hog does not matter.

Post-tensioning of precast is not a problem as long as someone who knows what they are doing installs the prestress. I did it for a year on a multiple bridge project in Indonesia.

If you wanted to get really adventurous, you could pretension the beams sufficiently for construction, then use draped bonded PT to connect them all together and make the beams continuous, so they could be even thinner.

 

[Ingenuity]

A somewhat common 'local' floor system in my area throughout the 1980's was pretensioned single-span 'shell band beams' incorporating continuous post-tensioning along the full beam length. These shell beams were acting as 'permanent formwork' and also forming a permanent structural function. Typically pretensioned one-way planks spanned orthogonally.

The soffit of the shell beam was relatively thin - enough to accommodate the straight pretensionjng strands and beam stirrups - and often, to maximize the PT tendon drape, a 'trench' parallel and between the pretensioning strands was created within the shell soffit/bottom flange.

The system 'died' when double Tees became prevalent in the area.

For your situation with significant cantilevers, a precast shell will just be expensive permanent formwork!

It may be best to engage your local precaster early on in the project development - sometimes they have significant capabilities beyond the usual planks, inverted t's and DT's.

 

[Ingenuity]

If you wanted to get really adventurous, you could pretension the beams sufficiently for construction, then use draped bonded PT to connect them all together and make the beams continuous, so they could be even thinner.

You have pretty much described the pretensioned/post-tensioned system I described above - and posted whilst I was typing...

 

[KootK]

- pretensioned harping can indeed be done although, in my area, it would limit the pool of bidders.

- pretensioned debonding bonding can be done but probably isn't necessary.

- When you get into it with these things, you'll usually find that straight pretensioning on the "wrong" flexural side really doesn't impact ultimate flexural strength all that much. The undesirable force acts over an inconsequential lever arm.

- A simple design option might be pretension top and bottom in roughly equal amounts. You'll have ultimate bending capacity where you need it, near zero cambers resulting from prestress, and pretty good deflection characteristics owing to the likely availability of the uncracked concrete stiffness in most locations along the span.

- I have some good tools for exploring pretensioned, precast beams. If you'd like to explore things a bit numerically, I can probably spare a little time to help out with that, just let me know the details.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rapt]

Pretensioning both sides is possible but very inefficient.

You are effectively using double the amount of strand that is required. A lot of the strand is wasted as it is not required where it is in the member. As well,

As the is no effective uplift from the prestress, a lot of prestress would be required to keep the section uncracked for deflections. You are simply getting benefit from P/A, without help from prestress moments.

And with no uplift, from the prestress, you have lost the beneficial effects of the prestress on deflection from the upward curvature normally associated with prestress.

For ultimate strength, while the direct effect of the strands at the compression face will not cause a large reduction in ultimate capacity, they will increase the neutral axis depth due to the increase in the concrete compression force and thus cause a further reduction in the ultimate capacity from the reduction in lever arm of the tension face strands.

And shear capacity will be reduced significantly.

 

[KootK]

Pretensioning both sides is possible but very inefficient.

Inefficient in a material usage sense. The precast world revolves around simplicity and my proposal would be about as simple as this could get. I'd think that you'd have to look long and hard to find a precaster that wouldn't prefer to toss in some top strand if it avoided all of the extra complexity of introducing PT.

And with no uplift, from the prestress, you have lost the beneficial effects of the prestress on deflection from the upward curvature normally associated with prestress.

Unless we're heading down the road of staged PT, I'd think this to be a moot point. The cantilever deflection of interest is the net displacement occurring after install. Preinstalled tendon curvature won't help with that excepting the cross section stiffness improvement that you mentioned. Also, as OP alluded, camber is not necessarily your friend in this situation as it may complicate erection.

And shear capacity will be reduced significantly.

So tighten up the stirrup spacing. Again, simplicity is king. Adding some stirrups is going to be more appealing to your average precaster than is introducing PT. Even harping the prestress tendons could limit the number of bidders on the project. In my market, precasters need additional qualifications and QC procedures to go from straight tendon work to deviated tendons. Two of my precaster clients can't currently do it.



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rapt]

"The cantilever deflection of interest is the net displacement occurring after install. Preinstalled tendon curvature won't help with that excepting the cross section stiffness improvement that you mentioned. Also, as OP alluded, camber is not necessarily your friend in this situation as it may complicate erection."

Who is after precamber? It is also necessary in design to limit the total deflection. And upward curvature provided by the PT to overcome Self Weight deflection will help a lot with this. Otherwise you need an even deeper section again.

I know ACI318 only talks about LL deflection, but that is rubbish. Total deflection is also important.

 

[Settingsun]

Sounds like a conversation with the client is needed. Post-tensioning could be the complete wrong path for them. Reading between lines, they just want to do some crane work and bolting at the site. PT off -site might be OK if they're capable of it. Or might not; have to ask.

Re precamber: could you cast it into the cantilever rather than relying on prestress to pull it upward? Might cause trouble for straight strands though.