Evaluation of unbonded tendons in "push-through" type sheaths, in beams

[ajk1]

Is anyone out there with extensive experience in evaluating the condition of unbonded post-tensioning strand in the old "push-through" (not extruded) type sheath with annular space between the steel strand and the sheath where water can collect and corrode and fail the strand?

After much investigation of a 40 year old garage we find a few totally broken tendons, some with one or more wires broken in a 7-wire strand, etc. Investigation over several years included chipping into the low point of tendons in about 150 spans, and tightness testing of the strand by crowbar prying, and screwdriver penetration test (screwdriver penetration test is not that good a test).

The whole structure is currently outfitted with an acoustic monitoring system and it reports about one "possible wire break" every 6 weeks. There have been only about 5 "definite" wire breaks reported by the acoustic system over its 3 years of operation.

The crowbar tightness testing has identified about 300 broken wires.

I have concluded that there is no way to reliably determine the number of broken wires in a strand (except for the case where the strand is totally limp, in which case we can conclude it is broken).

Based on the crowbar tightness testing, some spans have 22% loss of wires, but no midspan soffit cracking, but cracking is not a good guide to percent loss of wires, since you can have a large loss of wires without resulting cracking. However, if there is significant cracking, there is most probably also significant loss of prestress due to wire breakage.

Running ADAPT indicates that the loss of strength is close to the percent loss of strand wires. So 22% seems a lot, and I am having trouble justifying not doing anything to add support, but that would be very expensive.

Given the subjective nature of the crowbar tightness test, am I justified in recommending strengthening, or should I recommend another few years of the acoustic monitoring before making a decision to strengthen? Is there anything else that can be done to evaluate the loss of strength?

 

[Ingenuity]

ajk1,

Since 1993 I have been involved with antiquated unbonded PT tendon system evalatitions, all of which were (and continue to be) in the US, where "push-through" sheath systems are not too common (predominantly a Canadian system, compared to the more prevalent US past practice using heat-sealed sheaths and kraft-paper sheaths). However, one notable project (constructed in early 70's I think) that used the "push-through" sheath was probably the project with the highest % of failed tendons compared to total tendon population, with the most significant levels of severe corrosion along multiple locations of tendons.

I shall dig out my notes and project photos and get back with on a more detailed response, but my quick reply (and only partial answer) to your questions is that the most definitive way to determine pre-existing damage to tendons is to de-stress and fully remove those selected tendons and visually evaluate the tendon condition along its full extracted strand length.

With my building, we had 7 levels of 25'+ cantilever perimeter beams, so by definition there was no low point for water to settle, but we had 25% tendon failures! Water was 'trapped' in the "push-through" sheath between pockets of grease, and water ingress occurred during transportation and installation (small breaches open the sheath etc). We extracted hundred and hundreds of tendons on this building (to both slabs and beams) and did visual evaluation to selected tendons. We had perimeter access to stressing and fixed ends so we conducted many lift-of tests using a special collet that we developed.

It is not fair to state that tendons corrode ONLY at low points of spans due to entrapped water.

I think it is possibly premature to tell the owner to strengthen until you have extracted full strand lengths, and/or done some lift-off tests, or some other reliable field method that gives you confidence on the probable tendon failure %, compared to what your structural analysis determines is an acceptable % of failed tendons.

There is a Canadian document I shall dig up that gives you a starting point for determining sample sizing based upon tolerated errors, and acceptable tendon failures, assuming a normal distribution.

More to follow...sorry I am a bit time-crunched at present - end-of-the-year deadlines.

 

[ajk1]

Ingenuity;

Thanks very much for your detailed response that goes directly to my question.
Yes I am well aware the water can get trapped at other than low points. I have found the same thing in other structures.
I don't think this is a Canadian horror story only; since my first investigation of these issues in 1986 ±, I have been under the impression that the push-through sheath was also used in the U.S., but in the U.S. they switched to the extruded type sheath years earlier than in Canada. Had the industry association been more open and transparent about the problem, then the switch may have occurred sooner in Canada, but of course industry associations are not in the business of advertising industry problems. Even with the extruded sheath, which is definitely a huge improvement, water can still get in thru the ends and accumulate in the interstices along the wire (for a new project where we tested end caps that were advertised as water tight, we found they were not quite watertight, and we had to get the manufacturer to tweek the cap design) , or it can get in thru the many breaks that are inadvertently made in the sheath on site (by the contractor dragging rebar across it, etc.). But I am digressing.

Yes about 3 years ago tendons were extracted from this structure but unfortunately the contractor messed up when he cut the extracted tendon into short pieces to facilitate the extraction because he was pulling the tendon out into the atrium of a shopping mall. When the pieces were transported to the lab, they did not know which piece belonged with which tendon.
Anyway, we do know for sure that some tendons have definitely failed (all 7 wires) because the tendon emerged out the bottom of the midspan inspection opening at the beam soffit, as well as out the end of the beam a number of years ago. So I am not sure that we are going to learn a lot more by extracting tendons, although we are going to do so in 2017.

But the bottom line is that, irrespective of how much extraction and examination we do, what are the rules for deciding if it is time to install additional support or is it ok to just monitor it? Can acoustic monitoring (and visual crack monitoring) justify living with 22% under-strength? Or 15% under-strength? Can the remaining tendons in a beam span suddenly fail due to having to support more load from the failure of several tendons?

We could do some lift off tests as you suggest(although it will be difficult to grip the tendon stub), and that will tell us the force in the tendon, and the tendon elongation, but will it tell us how many broken wires there are, given that the friction and wobble coefficient is not known very exactly?

Do we risk breaking a tendon if we do lift off tests? Suppose a tendon has 3 broken wires of the 7 wires, then might the remaining wires break when the lift-off test is done?

Also,what are your thoughts on all the inspection openings that have been chipped into the beam soffits and the sheath sliced open, then closed with Densoptape, insulation and metal cover plate? My concern is that they may have taken perfectly good tendons and created a problem. In at least one location condensation occurred at this point, 2 of the 11 tendons failed completely. Any thoughts?

Thanks for your comments to-date. They are very helpful and I look forward to any further comments if you have time.

 

[rapt]

Ingenuity,

Still recovering from Christmas day I see ("evalatitions").

Someone proposed that we use this method in Australia in the late 1970's/early 1980's while I was head of design at Freyssinet Aus. A local consultant had designed a project unbonded and our management people decided to try it to see how it compared cost wise etc to our normal bonded PT system, against our protestations. Fortunately I saw the quality of the product being produced and the potential problems we would be responsible for down the track and wrote a scathing technical report on the project, basically casting aspersions on the engineering ethics of the construction engineers who were overseeing the process. Fortunately they saw the light and the experiment was cancelled before we could get into trouble.

 

[ajk1]

rapt;

Good for you. You deserve a special place in heaven.

 

[rapt]

Not sure it qualifies me for that! I thought the Americans had already anointed a certain American PT "expert" as the top PT man in heaven!

 

[ajk1]

If Ingenuity is on the system, I was wondering regarding your suggestion to do lift-off tests, would there be an issue of knowing which end anchor corresponds to which tendon, when a tendon that is not entirely tight is found by crowbar prying at the low point f the tendons in a span?