Wrapping Existing Column With New Concrete - Low-Shrinkage Concrete Mix 4

[KootK]

I'm reviewing a design where an existing round concrete column is being encased with 6" more concrete all around. Roughening for composite behavior, new cage, and all that jazz.

I have this picture in my head where the new concrete tries to shrink but is restrained from doing so by it's bond with the existing concrete. That, in turn, result ins distributed cracking and puts the new longitudinal rebar in tension and compression alternately from one crack to the next, in effect, prestresing the original column.

So my questions are these:

1) Am I considering this correctly?

2) With a prestressed column that one might get from a precaster, the prestress does not add to the axial load on the column for stability purposes. The same would be true here, yes?

3) Would you be specifying a low shrinkage mix for this application? For what it's worth, the contractor would like to use SCC for obvious reasons.

4) If you were to specify a low shrinkage mix, how would you determine the performance characteristics to specify here?

5) The new concrete will be rather heavily reinforced. Should that effectively eliminate my shrinkage concern on it's own?

This is a stocky parkade column that is in need of additional compression capacity. Moments and slenderness will be ancillary concerns.

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.

 

[TehMightyEngineer]

Interesting project, how many columns do you need to do?

1) Doesn't sound unreasonable, the "composite" addition will try to shorten the length of the column effectively similar to a prestress. If it cracked though, wouldn't you loose the "prestress"?
2) Doesn't it add to P-little delta effects? Either way I wouldn't consider any "prestress" in the design unless it was unconservative to do so.
3-4) Unless I had a bunch of these to do and the cost to going with a no-shrinkage mix was large, I would definitely go with a no-shrink mix. I'd be looking carefully to make sure the mix isn't one that expands initially and then contracts down to get the "no-shrink" concrete. 100% agree with using SCC on this. I'd even encourage it with a little external vibration. Whenever we've done tight pours (around 6" to 8" thickness) with a bunch of cages of rebar and a long fall for the mix, we've had issues with bugholes and other finish issues due to entrapped air. Never did come up with a perfect solution after 14 or so pours in the same form but light external vibration at key locations didn't over-consolidate the mix and seemed to help a little bit. If you can pour in segments or avoid long falls for the mix it will help but I'd still be using SCC if it were me. I'd probably include a few witness holes as well.
5) I'd say "maybe" (but that's mostly a guess). I'd say shrinking away from the bond with the column is a concern and your reinforcement wont help you here. For longitudinal shrinkage I'd expect that the heavy reinforcement will resist it but you still might have some issues due to shrinkage at the ends where the reinforcement isn't fully developed or because of restraint where the existing column connects into the structure.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

Thanks for your input TME.

If it cracked though, wouldn't you loose the "prestress"?

I've been pondering that as well. I feel that the stuff that is the wrap (conc + bars) will be trying to shrink. End of story. Thus, cracked or not, the wrap will be in tension over all and the existing concrete will be in compression overall, whether the wrap is cracked or not. This, at least, until load is applied.

Either way I wouldn't consider any "prestress" in the design unless it was unconservative to do so.

100% Agree.

Doesn't it add to P-little delta effects?

Interestingly no. The tendency for the tensioned bits to restore the member to straight (load balancing-ish) perfectly balance the P-Littles.

I'd be looking carefully to make sure the mix isn't one that expands initially and then contracts down to get the "no-shrink" concrete.

I hadn't considered this. Are you concerned with the initial expansion because it might disrupt the composite bond? In a sense, a little permanent expansion might be nice as it would precompress the new concrete some. All for naught though if the composite bond is lost though.

I'd say shrinking away from the bond with the column is a concern and your reinforcement wont help you here.

Oh my. Another thing that hadn't occurred to me. I'd imagined that the concrete would shrink towards the existing column transversely. Your thinking is that the ties would prevent that and the concrete would shrink away from the existing columns. At present, the bond does not include any mechanical anchorage.

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.

 

[jayrod12]

In regards to the composite action/shrinkage away from the bond idea, do you really need it to act together? or are you just trying to confine the existing column to up the capacity?

I'm envisioning this akin to a FRP or steel jacket where you're providing confinement. And the 6-8" of concrete is just to protect the rebar from the elements. Is that a bad way to think of this repair?

I would think that alleviates many concerns, including connections at the top and the bottom, and the development concerns at the top and bottom as well. But maybe you need the additional punching shear capacity at the slab as well.

 

[bowlingdanish]

Hello!

My first question - is this a philosophical question rather than practical? I haven't actually strengthened a column before but assume it is done routinely by this method, so wouldn't have thought it an issue. Next question - is the typical detail just to scabble to column surface and pour concrete up to it or drill/epoxy some small diameter hooked bars at regular vertical centres in addition? On to my attempt at the questions (which I may not understand)

1) I can't imagine the shrinkage forces of the new poured concrete would ever affect the existing rebar in the column. It would shrink in two ways, radially and longitudinally.
Radially, the existing column provides edge restraint to the new pour - causing tension in the outer surface, leading to tension in the radial bars and possible fine cracks. Longitudinally, it would try to compress the existing column, but the tensile strength in the concrete would be exceeded well before it did anything, and it would crack and dissipate all its nasty power.

2) I would not consider any prestress effects

3) I don't think I would mind. I would add enough rebar to control cracking.

4) -

5) Yes. The concrete will crack before doing anything of significance, reinforcement to control crack widths

Gooodbye!

 

[KootK]

Welcome Jayrod & Bowling Green

In regards to the composite action/shrinkage away from the bond idea, do you really need it to act together? or are you just trying to confine the existing column to up the capacity?

The overstress is an axial thing. So the goal has been to transition future load out of the original column and into the jacket concrete through bond. The jacket could probably handle the load on its own but it still has to get there somehow.

Is that a bad way to think of this repair?

It's not the approach that's been taken but that could be changed if need be. The original concrete had its own cage and, thus, some confinement. Could such a system benefit meaningfully from the addition of a reinforced concrete jacket of a similar scale? Or would we need ties at 2" o/c?

My first question - is this a philosophical question rather than practical?

They'll be executing the repair Wednesday. I'll have to rely you guys to tell me if my concerns are impractical.

Next question - is the typical detail just to scabble to column surface and pour concrete up to it or drill/epoxy some small diameter hooked bars at regular vertical centres in addition?

So far, the program is scrabbling but no dowels.

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.

 

[hokie66]

Too much thinking. Just get on with it. It is a "stocky" column, so I wouldn't worry about shrinkage. And I wouldn't allow them to get too robust with the scabbling. Just ensuring a clean surface would be enough.

How will they cast and compact the concrete? Those would be my only concerns with your scheme of reinforcing the column.

 

[bowlingdanish]

Green

Well I never. hokie66 seems to agree with my sentiments though, when do I get to change colour?

So far, the program is scrabbling but no dowels.

Should be a high score then!

 

[TehMightyEngineer]

Too much thinking.

Probably

How will they cast and compact the concrete? Those would be my only concerns with your scheme of reinforcing the column.

If they use properly batched SCC then they shouldn't need to worry about compaction. This isn't much different then much of our precast forms which are all poured with SCC and unless we have a major fault in the mix we never see honeycombing or other large voids. Though, see my note about long falls and reducing bugholes.

I hadn't considered this. Are you concerned with the initial expansion because it might disrupt the composite bond?

More worried about it trying to expand longitudinally, pushing against the top and bottom of whatever the column is connected to causing longitudinal compression, and being unable to push these relatively rigid supports up; buckling outward away from your existing column putting high tension loads on your bond. I'm not sure if this is even what would happen but there are definitely mixes out there that have balanced non-shrink properties so that they nether shrink nor expand. You're right that if it didn't "buckle" outward then it would be beneficial to expand longitudinally.

Oh my. Another thing that hadn't occurred to me. I'd imagined that the concrete would shrink towards the existing column transversely. Your thinking is that the ties would prevent that and the concrete would shrink away from the existing columns. At present, the bond does not include any mechanical anchorage.

Here's the two issues I can see happening with shrinkage:

The longitudinal shrinkage pulling away from one end would cause high shear stresses on the bond. Maybe not an issues but worth considering.

The radial shrinkage might cause the concrete to shrink towards the reinforcement in the added ring of concrete and thus pull on the bond to the existing column. I will say that I haven't really seen this in our precast forms; there's definitely some radial shrinkage but even with a smooth surface and form oil there's still a definite bond to the steel form. I'm not sure if this is just the concrete shrinkage creating a vacuum or if it's an actual concrete to steel bond, though. Regardless, PCI recommends ~130% of the self-weight for precast stripping loads from steel forms so that force is definitely real.

Overall, the uncertainty on what the shrinkage/expansion might do to your bond was why I felt a non-shrinkage mix was easily worth it to avoid loosing sleep. Might be worth using a fancy bonding agent too as I almost never see anyone prepping an existing concrete surface properly for proper bonding and to not steal water from the new concrete.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

This article's decent: Link. They mention shrinkage control as desirable bu ton't say much about why.

Well I never. hokie66 seems to agree with my sentiments though, when do I get to change colour?

Sorry about that. Brain failing...

Too much thinking. Just get on with it. It is a "stocky" column, so I wouldn't worry about shrinkage.

I appreciate the nudge.

 

[bowlingdanish]

I see it happening like this:

 

[KootK]

Thanks for that BowlingDanish. I see it the same way.

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.

 

[Ingenuity]

With jacketing of dissimilar materials, if there is a mismatch with your elastic modulus between new and old concrete materials (assuming the new materials aged-modulus is less than the existing), AND/OR, if you assume no bond between new/old contact surface (and it is therefore contact-critical) then the jacket provides a MINIMUM of circumferential confinement (assuming it is designed/detailed accordingly), and therefore you can utilize triaxial compressive strength/strain on your original (core) concrete capacity.

ACI-440.2R on FRP confinement to columns goes through some typical capacity calcs.

 

[oldestguy]

Kootk: Have you considered wrapping the existing column with cable which then gets an epoxy coating followed by ordinary plaster?
Failure potential of the column would end up with a cup and cone or diagonal shear fracture which would be considerably less likely with the perimeter well reinforced. This reduces all the questions about concrete shrinkage, placement problems, etc.

 

[BUGGAR]

Check out pile jacketing. It's done to both strengthen piles and restore deterioration. This is done in tidal zones so exposure is severe. As expected, we used special concrete mixes and a lot hoop reinforcing. Gerwick came up with a formula for the reinforcement that we still use today. We typically cleaned the piles by power-jetting, usually with a Contractor built rig. Bond was not attempted.

 

[Ron]

KootK....you're a smart guy, but I agree with hokie66....you're overthinking this. Don't worry about the shrinkage stresses in this application. Just do it.

 

[KootK]

ACI-440.2R on FRP confinement to columns goes through some typical capacity calcs.

Thanks for this. Initially, I scoffed at using provisions developed for FRP. Turns out that the provisions developed for FRP were originally developed for steel. Based on the amount of tie reinforcing provided, it seems that little would be gained from a confinement approach.

Kootk: Have you considered wrapping the existing column with cable which then gets an epoxy coating followed by ordinary plaster?

Check out pile jacketing.

Neat ideas guys, thank you. The state of the union here is this:

1) It's a peer review engagement.

2) Material is on site and to be installed this Wednesday.

3) The repair has clearly been specified assuming that the jacket will function simply as more column. Ties spacings are based on the usual code rules.

My immediate path here needs to be evaluating whether or not anything about the original design needs to be changed. I can't venture into better, more innovative solutions until I've established that what's already been specified is truly deficient (it's probably fine).

KootK....you're a smart guy, but I agree with hokie66....you're overthinking this. Don't worry about the shrinkage stresses in this application. Just do it.

Thanks for the feedback. I may do just that. I also may have pushed this further into overthinking territory than I intended with the faux pre-stressing concept. While I think that there's truth to that, it's really a much simpler issue at play here. Low-shrink mixes are often used in repairs and I wonder if there might be merit in doing the same here. With the repair assuming that new longitudinal bars will aid in compression, one wonders at the value of those bars if they'll start their lives off actually in tension. Perhaps the small amount of shrinkage cracking will be easily offset by eventual creep and elastic compressing once the superstructure is in place.

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.

 

[SlideRuleEra]

KootK - Per Hokie66 and Ron's comments, this is really not a big deal. Below is a photo from a bridge widening project we performed (as the Contractor). The photo shows cast-in-place reinforced concrete "crash walls" adjacent to railroad tracks. The crash wall encases 18" square prestressed concrete piling. Nothing special about the crash wall concrete - type 1 cement, 6 bags or cement per cubic yard (nominal 4000 psi), and air entrainment. Hard to judge the scale in the photo - the wall shown is about 30' long x 30" thick x 10' high (including the underground portion).

This type crash wall design has been in use for decades. For this overpass, it's now 40 years later, it is still in service, doing well... I drive over it about once a week.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[KootK]

Thanks for the photo SRE. I agree, I see a lot of similitude between you case and mine.

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.

 

[Compositepro]

I have little experience with concrete but I do deal with problems caused by shrinkage frequently. Thermoset resins shrink when they cure. Some basic principles that I have learned along the way include:

1. Shrinkage is volumetric not linear. Restraint of shrinkage in one direction will increase it in the other directions. Restraint can be due to reinforcement (fibers or rebar), hardening of the resin/cement so it is no longer fluid, and mold surfaces. Shrinkage problems are more pronounced in closed mold cavities than in molds with one open surface.

2. When aggregate particles are touching, they cannot get any closer together and bulk shrinkage cannot occur. Stress will build in the cement between particles.

3. Bubbles and air-entrainment are stress relievers for shrinkage. The bubbles simply grow very slightly larger to accommodate shrinkage.

4. Your drawing depicting radial shrinkage is not correct in that radial shrinkage will cause the concrete surrounding the column to pull tighter to the column.