Cover seal for post-tensioning inspection openings

We have used natural rubber (not synthetic rubber = neoprene) to act as a 'perimeter compression seal' to an end cap for a 3-strand PT anchorage. It was attached to a slab/wall edge.

We used a 1/4" thick mild steel custom circular end cap (in a top-hat shape) and secured the cap plate annular flange to the PT bearing plate with 3 threaded studs at 120° around the circumference. The natural rubber was only 1/8" thick but quite stiff, so we torqued the studs to compress the end cap against the rubber seal (which was against the bearing plate. The end cap was grease filled.

We removed the end cap about 15 years after installation and no moisture was present. Grease was in excellent condition. The exposure was exterior, tropical, with lots of rainfall and humidity.

I realize the above is not your situation, but it may have some relevance.

I assume your cover plate is probably only a thin gauge sheet metal square or rectangular plate, and probably secured with 4 corner tapcons, or similar.

I presume you are proposing double-sided peel-and-stick neoprene tape to obtain some sort of seal.

Maybe it is best to provide a perimeter polyurethane caulking too, especially at the corners where tapcon screws would interrupt with the neoprene tape. I know caulking the plate each time after inspection is a pain, but caulking is cheap.

PU 'sticks' well to prepared concrete and sheet metal - like your would detail a 'flashing' etc.

To Ingenuity - thank you for your very clear and detailed response. Was the rubber perimeter compression seal something that you made by cutting a hole is a square piece of rubber, or were you able to find an off-the-shelf rubber ring that fit your situation? How did you specify the natural rubber, as to its composition (or perhaps a reference standard) if you can recall? Was the hat shaped assembly something you found off the shelf, or was that a specially made assembly for the project?

My beam soffit plates will be 16 gauge steel, about 12" x 20" and I am specifying that they be fastened along 2 edges(the other 2 edges being too close to the bottom corners of the beam to install fasteners, but they will have a vertically lip to stiffen them. The fasteners will be the Hilti 3/8" diameter drop-in anchor type so that they can be periodically removed to allow inspection of the tendons. I will send sketch in a few moments...that will be easier than trying to describe.

Most of the inspection openings I will specify to be grouted up i.e. no removable cover plates, where no looseness of the tendons was found. We have an acoustic monitoring system in place now for about 3 years to detect tendon wire breaks.

I am still interested if anyone on this site has experience with the neoprene gasket with a rectangular plate, and if so, how did you seal the corners, and did you remove it say 3 years later to see if there was any tendon corrosion?

Attached are 2 preliminary sketches illustrating what I described above.

...And here is the 2nd file, I hope!

ajk1:

The top-hat end cap was custom fabricated in our shop (very remote from the project site - different country actually - then hot-dipped galvanized before shipping. See 3D image below. Everything on the project was custom - it was an external PT strengthening project to an existing PT slab with extensive corrosion damage. Internal deviators, end plates, some weird skewed angles, etc, everything was custom.

The rubber seal maybe called "custom" too, but in actual fact is was rather crude. Think truck tire inner-tube cut in the field in fit the annular shape of the end plate. This was in a tropical island setting where resources were limited. There were no specifications for the end-cap and seal - it was the early-mid 90's!



In 2009 (15 years after installation) we did an inspection (removal of cap etc) and it was in good visual condition.

In some respects the end-cap above replicates what is used today in multi-strand bridge tendons with grouted/bonded tendons - but the end-caps are now a HDPE or PP polymer with a compression seal material.

ajk1:

Your two details look good to me. Very detailed and well thought out. Mock-up is a great idea.

1/2" thick neoprene is pretty thick - I thought my 5 mm thick wetsuit was thick - which reminds me: I use a wetsuit (foamed neoprene) glue to repair nicks, cuts and seam failures - so maybe this contact-type cement is the type of glue you need to the corners.

Maybe a few things to consider:

ajk1:

Not sure if this is important to you or not. The two PDF files you uploaded above, when someone hovers over the text it brings up a dialog box and has you name on it. I assume you use Bluebeam or another PDF file maker and it automatically adds such info to its text boxes.

If anonymity if important to you, RED FLAG your two file upload posts and the moderators will delete them. For future files, consider uploading JPEG or PNG copies of you sketches using the UPLOAD PICTURE or UPLOAD IMAGE icon on the Reply To This Thread "menu".



If the above is not relevant, feel free to RED FLAG this post and the moderators will delete my post.

To Ingenuity - thank you for your thoughtful comments. Very much appreciated indeed.
Regarding hitting the tendons, I will add note about over-drilling.
Regarding field drilling the cover plates, I am reluctant to do so as that would leave an ungalvanized edge. Perhaps I could specify a slightly over-sized hole, and double washers.
I will consider adding a PU sealant as you indicate, although I suppose that it should be applied at all edges.

I am thinking now that the idea of a gasket may not be as good as I at first thought. Perhaps my other option (not posted here), of a 1/2" wide wide bead of sealant applied to the plate before it is bolted in place may be simpler and better if it is consistently done right, although it will make any future removal of the plate more difficult.

I did not realize about my name being revealed. Thanks for the heads up. I use Bluebeam. I will see if I can understand how to red flag an attachment next time I do an attachment.

Your top hat details that you used 15 years ago look very good. Too bad not all people designing repairs put the thought and care into them that you do. The fact that it has worked for 15 years is confirmation indeed of an effective solution.

I tried "flattening" the bluebeam file to see if this gets over the name reveal issue.

How can I calculate the force (even just approximately) that a corroded erupting 7-wire high strength post-tensioning tendon would exert on the cover plate? I expect that it would involve equating the energy released when a tendon breaks, to the work done when the cover plate deflects.
But what would the equation be for the energy released when the tendon breaks?
And what would be work done when the plate deflects? If I assume that the tendon strikes the plate at midspan of the plate, then the work done is P L³ / (48 E I) (where L is the span of the cover plate between fasteners), or is it half of that?

I see that no one has responded to my last post. I wonder if BARetired or Kootk is on the system. They might be able to help me!

ajk1:

Sorry, I am not BAretired nor KootK, but I shall offer this.

I have intentionally cut several thousand INTERNAL PT tendons over the last 30 years, and with the exception of a handful of special cases, the vast majority are not very 'exciting' - meaning we don't see the 'urban-myth' cases of tendons flying out of building perimeters, or chopping of limbs etc.

Any meaningful/real-world/approximate calculations of release energy/force of a failed strand will have to consider the damping due to the grease, extent of corrosion to individual wires, tendon curvature, tendon wobble, distance from failure point to cover plate etc, and these will present significant differences/unknowns from any theoretical values to field-validated values.

Williams did his PhD on this related subject back in 1988: Williams, M. S., "The Cutting of Unbonded Prestressing Tendons During Demolition" PhD thesis, University of Bristol, 1988, 303 pp.

There was an ACI Structural paper by him and Waldron in 1989 entitled: "Dynamic Response of Unbonded Prestressing Tendons Cut during Demolition" Link It was laboratory-based testing of severed tendons, looking at 'exit distances' from anchorages, and has some relevance to your situation.

There is also this ACI Structural journal paper from 2012 Link entitled: "Simple Equations for Tendon Failures" by a couple of WJE engineers, Perry and Estenssoro.

To Ingenuity: Thank you so much for your comments. It would make my life much happier if I could simply say that it is not an issue, but the following militate against such an approach:

1. A failing tendon on this structure which erupted thru the beam soffit did knock down one end of the metal cover plates that was installed (bolted in)under the inspection opening, and left the plate precariously hanging.

2. As far as I know, all the references that you give are for tendons within a monolithic concrete member (where whether it erupts up or down out of the structure depends on the concrete cover), not for a bolted plate or doweled-in grout fill, which are the 2 closure methods we are considering.

3. On Place Bell Canada garage, in the 1980's, the tendons most definitely did erupt out of the floor. That resulted in the longest trial in Canadian civil litigation history up to that time (I was one of 70 who gave expert testimony at the trial).

Thank you for the links. They are very much appreciated. I will try to look at them and see if any of them can help answer my question. I may have some follow-up questions.

Thanks again.