Broken Hollowcore Plank 4

[shacks26]

I'm working on an existing building, where the demolition team have broken one of the existing hollowcore floor planks - part of the impact load has caused the wires on one of the ribs to break and concrete to break off from the soffit across 70% of the width (probably over the middle quarter of the plank).

My first thought was to replace it, though the demolition contractor has roped in one of his engineers to suggest a repair (obviously there's some politics of them trying to save costs on forking out for this). They've demonstrated based on % of damaged reinforcement that the slab still has capacity in bending, and their methodology of repair is to roughen up the inside of the voids, add additional bars and grout. This seems sensible - floor capacity is ok, additional reinforcement to make up for that damaged and re-bond/re-cast.

Has anyone undertaken a similar repair before? As we do not know who installed the original floor (back in the 80s) many precast companies I've spoken to won't go near it / be willing to discuss at what point they think the integrity has been compromised and it is unrepairable. Not having much experience in concrete structures (especially precast/prestressed) I'm trying to understand some of the issues with repairing; or whether warranty/workmanship is more the controlling factor...ensuring the concrete and wires are re-bonded etc. Would the impact load from the demolition material have caused any internal damage / weakened the section in any way?

 

[Ingenuity]

Are the precast hollow-core planks prestressed with 7-strand wire (say 3/8” or 1/2” nominal diameter) strand OR straight wire (1/4” dia)?

Is there visual evidence of damage to the prestressing strand/wire?

Given that the plank damage extends to 70% of planks keep width - depending on the extent of damage over the section depth, and the plank length - the prestress to the several strands/wires may have been lost due to longitudinal slip.

I have repaired several damaged prestressed planks (solid and hollow-core) and each repair always addressed the prestressing strand capacity such that planks ultimate strength AND serviceability was reinstated.

It appears that the contractor’s engineer is proposing to simply use passive reinforcing steel as replacement for possibly damaged prestressing steel. I would place an emphasis on the contractor/engineer to prove (by calculation or load test - and then PE seal the submittal) that the plank capacity has been reinstated, or has sufficient capacity for any revised applied loadings.

 

[JLNJ]

If the plank is keyed into the adjacent planks, you are probably getting some load sharing. This might make someone sleep better at night, but if I'm the Owner I'm not sure I'd want to sacrifice any capacity when a contractor broke something.

Full disclosure - I have not had to repair planks damaged to the extent you have described.

I think there is some merit to adding mild reinforcement if at can be shown that the reinforcement brings full flexural strength back to the slab. I'm not thinking that re-bonding the existing reinforcing strands is particularly viable.

I would also worry about the remaining shear strength of the planks if the impact was so great as to rupture the tensile reinforcement. Shear strength might be a little more difficult to replace if the concrete is "broken". Depending on the size of the cores, maybe grouted cores can restore some the shear strength.

I'd make sure the Owner is on board with the repairs before proceeding.

 

[KootK]

1) Load sharing would have been my goto if I were fortunate enough to know the design / capacity of the existing planks. With a decent span and a few like minded neighbors on each side, you've got at least two options:

1) Say the flexural demand is shared over four or five planks such that you've lost one strand of 20+ strands.

2) Imagine that the damaged plank is worthless and spans laterally to the neighboring planks as two edge load conditions for the purpose of load sharing.

3) Was the impact from the bottom or the top? If it's the top, as is usually the case, I worry that the load sharing might might have resulted in several strands slipping as Ingenuity mentioned. You might consider some NDT to find out or, at the least, do some sounding / banging around the damaged area to see if you can encourage any other, easy delamination.

I have repaired several damaged prestressed planks (solid and hollow-core) and each repair always addressed the prestressing strand capacity such that planks ultimate strength AND serviceability was reinstated.

4) Could you share the high level details of how you did that? I don't know how to restore the prestress in these situations, short of external repairs maybe, and have always considered that a pretty tall order.

 

[shacks26]

Definitely some things to think about here...I'm in the UK so maybe I should have described them as hollow core floors...I've attached a picture for reference. The wires are 4mm (1/6"?) in diameter, 2 per strand/rib. I don't think that they are keyed to the other floor slabs.

 

[shacks26]

The previous photo was of the top, and this is the soffit.

 

[Ingenuity]

4) Could you share the high level details of how you did that? I don't know how to restore the prestress in these situations, short of external repairs maybe, and have always considered that a pretty tall order.

I use 'GRABB-IT' strand splices. Basically a custom turnbuckle with LH and RH threaded rods connected to strand splice chucks at each end. They are very cumbersome and the manufacturer (PRECISION-SURELOCK - now GTI) has tabulated torque-to-prestress values but - similar to bolt tensioning - I have found it not sufficiently accurate, so I hook up a strandmeter (vibrating wire strain gauge) to my datalogger and PC and record strain over a fixed 8" gauge length and use strain as an indicator of tension in the strand during the torqueing process. Never miss an opportunity of turning structural repair into a research project

Successful repairs are dependent on where the strand damage occurred. Ideally the strand repair location should be far enough away from the plank ends that strand development lengths are not a concern, so 1/3 to midspan of plank length works ok, unless the planks spans are very small.

I can set the datalogger to record strain at any time intervals, so I typically leave the strandmeter on the repaired strand for 12 to 24 hours and continuously record strain losses over that duration, so basically I get a measurement of immediate prestress loss (mainly strand slip).

Cover to strand is usually only 1" or so, so at the splice location the soffit often has to be lowered locally to provide fire protection and durability.

Here are some details/photos from a solid plank project repair where the coring contractor decided to sever several strands over a few planks:

Procedure:

GRABB-IT strand splice:

Torqueing a GRABB-IT splice:

A GEOKON vibrating-wire strand meter attached to strand and connected to datalogger and PC:

 

[KootK]

Thank you very much for sharing that Ingenuity.

 

[Ingenuity]

shacks26: Are these planks UNtopped?

What is the plank span?

 

[Ingenuity]

Thank you very much for sharing that Ingenuity.

You are most welcome, eh!

 

[jayrod12]

How did the damage happen? More just curious than anything, it really isn't important but that's a lot of damage on the bottom for what appears to be relatively minor damage on the top.

I share some of the others concern that for that level of damage seen, it's possible that they have potentially caused bond loss for other strands that aren't exposed.

Lastly, any idea what that line of holes is running down what I assume is the centre of all of the planks?

 

[KootK]

How did the damage happen?

Something similar happened to me a couple of years back. A lift of plywood slipped off from an upper roof and landed on the plank. The top of the plank was unscathed; the bottom looked like it had be carved away with a giant ice-cream scoop.

 

[Agent666]

You'd definitely be wanting to send a camera (boroscope?) into the voids to check for the extent of any web cracking (reviewing for damage beyond the immediate area that broke out). You don't want the rest of the underside of the unit to drop out over more of the span once some real load comes onto the system for example.

https://engineervsheep.com

 

[Agent666]

Lastly, any idea what that line of holes is running down what I assume is the centre of all of the planks?

It's not uncommon in some regions to specify drain holes to the soffit/voids. I'm guessing that this might be the case here?

https://engineervsheep.com

 

[jayrod12]

Yeah I was thinking drain holes too, but normally see those as close to the bearing locations as possible. But in looking closer there does appear to be a matching row at the far end of the planks.

Regarding the damage, that's an interesting anecdote Koot. Never would've expected that level of soffit damage from a top level impact. Guess that's my new thing for today.

 

[Ingenuity]

How did the damage happen? More just curious than anything, it really isn't important but that's a lot of damage on the bottom for what appears to be relatively minor damage on the top.

OP stated it was by a demolition contractor and due to impact.

Lastly, any idea what that line of holes is running down what I assume is the centre of all of the planks?

The holes at midspan appear to line up with the plank cores/voids - maybe drain holes to allow water to escape. Guess on my behalf.

EDIT: I was too slow - others beat me to the punch!

 

[shacks26]

Not drainholes on this project at least - I've read that they are sometimes used for services - essentially it's one of these (though not by this manufacturer) -

https://www.bison.co.uk/products/hollowcore-floors/

The impact was caused by the bit of rubble/slab you can see in the top picture. Plank span is 6.5m.

Yes, I think that is an important thing to consider - two wires broken...but is there bond loss elsewhere and therefore how can you prove what capacity it is potentially working at...

 

[shacks26]

The current proposals involve:
- Removing the loose sections of concrete,
- Crimp snapped wires with crimped sleeves,
- Drill grout flow holes at either side within each core,
- Roughen exposed concrete,
- Vacuum clean cores, clean with water and allow to dry,
- Form grout stops,
- Shutter, add new reinforcement to centres of cores (with spacers to achieve cover) - coat internal concrete with a bonding agent,
- Infill with grout,

There's a factor of safety of nearly 2.0 based on the plank capacity and proposed loading, and ignoring the two wires that have snapped - then adding additional bars to match the ultimate stress of the wire - using the grout and bonding agent -I guess the other problem as noted is clarifying any unknown damage.