"Extra" or "reserve" PT strands in beams/transfer beams 3

[DETstru]

I'm reviewing a set of PT shop drawings for a building I designed. It has a handful of transfer beams with considerable PT in each (unbonded).

The detailer added a few "extra" strands in each beam just in case there is a stressing issue with some of the strands. I have no problem with this; actually it seems like a really good idea since fixes are painful if a strand breaks during stressing. They noted these won't be stressed unless there is an issue with other tendons.

Just wondering if anyone else sees or requires this and if I would get pushback if I started noting this on all my projects.

 

[hokie66]

Is there staged prestressing of your tranfer beams? Having the ability to decide as you go how much prestress to apply at the various stages can sometimes help.

But as to the extra strands, sounds similar to reserve conduits for unanticipated electrical work. Just good engineering, I think.

 

[DETstru]

Yes, I'm staging the stressing.

For example, one beam has 60 strands:

Stage 1: stress 40 strands
Stage 2 (after many levels built): stress the remaining 20 strands

They provided 63 strands, 3 of which won't be stressed unless there is a problem.

I like a lot it since a strand break is really annoying in a transfer beam. In future projects I think I'll try requiring an extra strand for every 20 strands (or something like that) in transfer beams. Just wondering if anyone else does this. It's the first time I've seen it done.

 

[MrHershey]

We see this fairly frequently in stage stress scenarios. We'll sometimes include an extra tendon or two even in single-stage beams that have a lot of tendons for a similar reason. We'll often even stress those, so long as it doesn't overstress the beam or overbalance.

There's certainly merit to trying to be economical, save money, and not use more than you need. But if you typically see about 1 in every 200 strands snapped, for example, then it probably is more economical overall to include an extra strand or two in beams with a lot of strand because the odds are fairly high that one of your snaps will be in that beam. Helps if elongations come out low too.

 

[rapt]

I cannot believe that any responsible engineer would detail a transfer beam with 60 or more single unbonded strands to support a multi-storey building.

Support of the whole buildings is relying on the friction grip of each of those strands for the next 50 years or more and hoping that none of the long term decay problems experienced commonly on other unbonded buildings occur. As well, if it is an an earthquake zone the friction grip is going to have to survive the event. I will not go into the logic of having 60 or 100 or more individual plastic coated strands and their anchorages in a beam compared to 3 or 4 bonded PT ducts and anchorages.

I did not really believe this was common practice so I asked a friend who has extensive experience with PT in the USA over a long period and his response (modified in "..." to remove incriminating details as requested but otherwise his words) is repeated below with his permission.

Surprising common and exceptionally stupid.

I first saw it on a "PT Company name removed" job in "Major capital city name removed" and then on transfer beams in "an extreme seismic zone" with over 100 staged-stress tendons supporting a 40 Level hotel - "name removed" - so don't stay there!


I think he agrees with me.

 

[DETstru]

Can't have a PT thread without rapt pushing bonded!
Just kidding, I always appreciate your responses, thanks.

We've explored bonded in the past but nobody does it (or wants to do it) around here. Just seems like unbonded is the way it's been for so long. I don't really know the history behind that though; that's beyond my years.

 

[271828]

We used bonded on a gigantic hospital project. I'd use bonded for a transfer girder.

 

[rapt]

DETstru

After 40 years of it, you would think I would give it up, or maybe someone would change to shut me up! Some history for you from someone who actually designed 2 unbonded buildings in the 1970's!

Building PT in the USA started as unbonded in the 1960's and became basically a factory operation without specialist PT installers on site. This resulted in some really bad site practices which probably contributed a lot to the problems in the long term. But it is then difficult to change to a bonded system which is very dependent on specialist PT installers. The whole industry would have had to be transformed. VSL tried it in the 1990's, but they tried a very expensive bonded solution which was completely uncompetitive. 2 strand tendons in plastic ducts. Most bonded slab PT in the rest of the world is 4 - 5 strand tendons in metal ducts. I can understand the plastic ducts in high snow/ice areas in the north, but not in the south where de-icing salts are not needed. But 2 strand tendons were never going to be competitive.

Building PT in Australia started as bonded with specialist PT companies offering full service from design to installation, stressing and grouting and this spread through most of Asia. As I said elsewhere, we looked at unbonded in the 1970's and decided we could do things more economically and provide a better product with bonded, because that is the way the industry was set up, as a full service operation, and ice is used for other things here and in SE Asia.

A lot of my discussion for slabs is not so much regarding unbonded PT, it is the construction method that was adopted to make it work on flat slabs (one way banded/distributed tendons)and the resulting design method that was adopted to justify it. And the watering down of the limits of these requirements over time by allowing much higher tension stress levels than were originally used and that result in bad serviceability problems in the buildings.

Unfortunately this design method is spreading to countries outside the ACI zone and is also being used for bonded PT as well but it is being used without the added minimum reinforcement requirements and protections that are built into the design method in the ACI code. Part of the reason for this is that the limits in the ACI code are deemed to be for unbonded PT, whereas some of them are specifically for the "average moment design" method used for flat slabs in the ACI code. This fact is missed when people start using the design method to other codes that do not cover the logic of the "average moment design" method and assume that designers will be using column/middle strip logic in both directions. THis applies to both Eurocode and AS3600. BS8110 covered it in a Technical Report (43) which specified the much more restrictive tension stress levels I mentioned above (0 tension at supports!), but no other code has considered the possibility of average moment design.

RE members that require large PT forces, and are structurally very important, like transfer beams, unbonded single strand is really a joke! Bonded multi-strand systems are available everywhere (used in most larger bridges) and should be used for this type of structure.

 

[dik]

Can't have a PT thread without rapt pushing bonded!

Joking aside Rapt is correct, in particular, for a heavily loaded transfer girder in a high seismic zone... You should see what can be done to change this.

Dik

 

[KootK]

1) I don't understand the seismic concern other than a general distaste for things juggling around. If an unbonded PT anchorage in a horizontal, gravity element shakes loose under the vertical or horizontal accelerations typical of an earthquake, I would submit that there was something seriously wrong prior to the seismic event. Self-centering PT shear walls etc are a different animal of course.

2) This is one of those areas where I wish that codes would step up prescriptively. Rather than adding five new load cases in the next cycle, I'd much rather see a provision to the tune of "members supporting more than XXX square feet of tributary area shall utilize XXX% bonded post-tensioning". What is allowed to be done will be done by somebody. And I don't think that it's realistic to expect individual designers to fight their own markets on this kind of thing. It results in the bad engineers winding up with all the work which helps no one but those bad engineers. I like the codes to serve as a surrogate spine when mine can't get the job done due to market pressure.





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.