Smart questions
Smart answers
Smart people
Join Eng-Tips Forums

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

SkiisAndBikes (Structural) (OP)
7 Apr 11 21:54

I have a project to design a 75ft. long, 12" thick curved wall for a cenotaph in Canada. Crack control is essential as no joints allowed and it will be an exposed, polished concrete surface. I have had a little experience with water retaining structures where minimum reinforcement ratio may be as high as 0.004, to reduce the chance of cracking. Very rough sketch attached. I would love to hear from any recommendations related to minimum reinforcement ratio for the horizontal reinforcement.

Helpful Member!  JAE (Structural)
7 Apr 11 23:24
Reinforcement doesn't reduce the chance of cracking.  It only keeps cracks more tightly closed.

75 ft. long, restrained from shrinkage at the bottom but not on the top.  Sounds like cracks are going to have a party.

The curve doesn't help either as my initial reaction is that there will be some kind of radial effect on the wall as well.

Other threads here have discussed the things to do to keep concrete shrinkage to a minimum.
1.  Low w/c ratio
2.  Avoid lots of cement in the mix.
3.  Use larger aggregates
4.  cure cure cure (water cure the best)

Other approaches include Type III cement or type K.

Helpful Member!  hokie66 (Structural)
8 Apr 11 0:27
Agree with JAE, but I wouldn't skimp on the cement.  You can't prevent cracks unless you totally remove the base restraint, which is probably not feasible, although due to the plan shape and depth of embedment, it could be possible.

I would use at least 0.6% Ag horizontal reinforcement.  The most likely place for objectionable cracking would be near the ends, and would slope from the end down to the footing as the top shrinks and trys to lift the end.
JAE (Structural)
8 Apr 11 9:37
My item 2 was based on the fact that high cement content mixes shrink more than low cement content mixes.  Aggregates don't shrink, cement does.

Helpful Member!  BAretired (Structural)
8 Apr 11 16:36
How about 5.  Post tensioning?


Brad805 (Structural)
8 Apr 11 17:05
How about adding internal waterproofing such as Kryton or Xpex?  It will protect you for cracks up to 0.5mm wide.  It has a huge impact on your set time, so some like the surface applied options better.

Bobber1 (Structural)
8 Apr 11 17:41
Also, you can pour your wall in segments, then after initial shrinkage, which is usually the most, you go back and pour the infill walls.  This is done regularly in reservoirs.  Construction is difficult, though.

hokie66 (Structural)
8 Apr 11 17:41

JAE...that is a popular theory, but I think it is misleading.  The paper which follows says that shrinkage of the paste in some cases is only 20% of the concrete shrinkage.

BA...with restraint from the footing and the curve of the wall, I wouldn't like to try post-tensioning.  Might do more harm than good.

I don't think sealing cracks is his problem.  As I understand it, the wall is exposed on both sides, and the products mentioned would just tend to make any cracks more unsightly.
Brad805 (Structural)
8 Apr 11 18:34
hokie, I see that now that I opened the picture.  Somedays using the virtual PC in Windows 7 is a nuisance.  The water comment might have sent me off on a tangent too.

JAE (Structural)
8 Apr 11 18:45
Interesting overview article.  The points I'd make in reply are that they are speaking entirely of drying shrinkage.  There is also autogenous shrinkage which occurs in higher cement content mixes.  

Also, they don't specifically say that cement content doesn't affect shrinkage.  What they are saying is that with higher strength mixes (i.e. high cement content) there can be low shrinkage results.  What is implied in their assertion is that for higher cement content mixes, there is the tendency to have lower w/c ratios, which indeed reduces drying shrinkage.

What they don't say is, for a given w/c ratio, higher cement doesn't affect shrinkage.  I think it does.  

See the attached table which represents cement content vs. drying shrinkage.  For a given w/c ratio, cement content does make a difference.
BAretired (Structural)
8 Apr 11 22:51

I am not sure about the benefit (if any) of post-tensioning, but I cannot see where the restraint of the footing or the curve of the wall would have any deleterious effect, recognizing that a curved member, prestressed at its centroid results in a uniform compression throughout its entire length and that footing restraint is similar to a prestress.

I believe that post-tensioning is a valid way of reducing cracks in a curved wall.


JAE (Structural)
8 Apr 11 23:17
Just a question on that - if you post-tension something, don't you wait until 75% f'c before tensioning and doesn't much of the shrinkage already happen before that?  Perhaps the full opening of the cracks occurs later and the tensioning does the job.  What do you think?

hokie66 (Structural)
8 Apr 11 23:37
My preferred interpretation of the graph you posted is "Within the range of practicable concrete mixes, for a given amount of water, higher cement content will reduce drying shrinkage".

Leaving aside the curved member argument for a while and just assuming that the wall is straight, footing restaint is the opposite of post-tensioning.  Footing restraint tries to keep the wall long, while post-tensioning wants to make it shorter.  Footing restraint introduces tension, post-tensioning introduces compression.  The further the prestress is from the restraint, the more effective it is.  But a large amount of mild steel reinforcement in the wall just above the footing restraint will keep the cracks together.
BAretired (Structural)
9 Apr 11 0:01
JAE, Some shrinkage takes place prior to post-tensioning, but the subsequent compressive stress tends to close the gaps (perhaps not completely).

hokie, you are correct, the soil would tend to resist the concrete from shrinking, so perhaps this would suggest a greater prestress at the bottom to overcome soil friction.  I have not used this solution, so I suggest it only as a possibility.


hokie66 (Structural)
9 Apr 11 0:31
According to my references, a 12" slab would have only reached about 20% of its ultimate shrinkage after 12 months.  A wall exposed on both sides might be a bit different, but not too much.  The thicker the element, the slower the shrinkage.

In typical post-tenioned structures, elastic shortening due to prestress only accounts for about 7 to 10% of the total shortening.  The rest is shrinkage, creep, and temperature.
hokie66 (Structural)
9 Apr 11 5:37

Back to my first post...removing the base restraint might not be so difficult.  I would check the monumental wall for overturning stability with the bottom of the wall at ground level.  Due to its shape, I doubt it will turn over.  If that is OK, I suggest just digging a trench to the required depth for bearing and frost, fill the trench with lean concrete, and trowel the top level.  Then provide some means of making the surface as frictionless as possible, maybe just plastic sheeting, but others may have better ideas.  Build your circular wall, and if it is not restrained, it won't crack.  
kieran1 (Structural)
9 Apr 11 6:08
Wouldn't cement replacement with ground granulated blast furnace slag help. It certainly helps against early age thermal cracking


hokie66 (Structural)
9 Apr 11 16:48
That might assist in limiting shrinkage, but as the monumental wall will be polished, it could affect the desired appearance.

Your terminology, "early age thermal cracking", seems to be typical in Europe, and in most other places we call it "drying shrinkage cracking".  It is essentially the same thing, cracking caused by restraint of volume change.  
JAE (Structural)
9 Apr 11 22:12
hokie, "early age thermal cracking" I think is the autogenous cracking I spoke of earlier.  This is different than drying shrinkage cracking.

dik (Structural)
9 Apr 11 22:32
and a very low slump... if it's going to be polished, then you have to be careful with the concrete batches, they will be a slightly different colour that will be amplified by the polishing.

You may want to use HDG rebar to prevent staining and still have the bond...  You might be looking at .5% or greater, even 1% reinforcing steel ratio and 2" cover.  Rebar spacers to achieve cover will be a problem... in particular with the polishing.

Curing will be critical and as JAE suggested, the curved surface will likely increase flexural tension cracking.  Any chance you can creatively introduce a couple of control (pour) joints.

As noted, post tensioning may help, but I'm not sure how this will shake out with a curved surface.  Maybe some type of shrinkage compensating concrete.

Might want to do a couple of test panels...

dik (Structural)
9 Apr 11 22:39
Also silica fume may reduce shrinkage... but don't know what the overall impact is... again a couple of test panels...

hokie66 (Structural)
9 Apr 11 22:47
Here is another article about early age cracking.  The terminology (chemical shrinkage, autogenous shrinkage, drying shrinkage, early age thermal shrinkage) is all double Dutch to me.  Seems overly complicated for engineers, maybe not for concrete technicians.  We know it shrinks, and if restrained, it cracks.  This paper says that using higher w/c ratio will reduce autogenous shrinkage, and I find that hard to accept.
hokie66 (Structural)
9 Apr 11 22:50
Anybody have any comments about the idea in my post of 9 April, 0537?
BAretired (Structural)
9 Apr 11 23:19

If tendons are located at the middle of the wall from end to end, post-tensioning will produce pure compression throughout the entire arc length (i.e. no bending).  What better solution could there be to prevent cracking?    


hokie66 (Structural)
9 Apr 11 23:32
The horizontal curvature in the tendons has to be resisted.  Just as vertical draping in a slab or beam balances gravity loading, horizontal draping has to be balanced by loading in the opposite direction, and that would be by bending or arch action.  Without rigid supports at the ends, you don't have an arch.
dik (Structural)
9 Apr 11 23:49
Thank, Hokie... I was thinking that they would want to straighten out the wall... different situation if they were bonded... like a sag and an equivalent UDL uplift...

Regarding your Apr 9 posting, the more cement powder you have the greater the shrinkage, in general.

BAretired (Structural)
10 Apr 11 0:17

A concrete cylinder prestressed with wires wrapped around the periphery is stressed in pure compression.  There is no bending.  

I am simply extending this to a sector of a cylinder stressed by tendons located at the centroid of the section.  In such a case, I believe the stress in the concrete is pure bending.

Thus, it seems to me that post-tensioning is an ideal way to prevent cracking of the wall.  I do not understand your objection.


hokie66 (Structural)
10 Apr 11 0:28
1) Not different, whether bonded or unbonded.  The force from the tendons wants to straighten the wall either way.  Bonded tendons are unbonded before grouting, so impose the same forces on the wall, radially in this case.

2) I think you may be referring to my post of April 8, 2337.  The graph which JAE posted April 8, 1845 supports my statement.

3) But enough of all this speculation about how to limit shrinkage.  I would still like opinions on my proposed solution to the OP's problem, as stated in my post of April 9, 0537.   
BAretired (Structural)
10 Apr 11 0:46

Commenting on your April 9, 0537 post...I would agree that if you are able to remove all friction between the wall and the lean concrete, there will be no resistance to shrinkage and the wall will not crack.  However, in climates such as we have in central and northern Canada, it is doubtful that plastic sheeting would do the job.  Perhaps there are other materials available which would work, I do not know.


hokie66 (Structural)
10 Apr 11 0:51
Maybe you could build it on intermittent teflon slide bearings, with something like polystyrene between the bearings.
jtx (Structural)
10 Apr 11 0:58
To be honest, I don't think you will avoid not have some degree of cracking (microcracking at the least).  However, ACI 350 section 7.12 does require a minimum horizontal reinforcement ratio of 0.005 if movement joints are spaced 50' or greater (or in this case none at all).  Also, keep an eye on the mix design and curing just as others have mentioned.  I think ACI may also have recomendations on that somewhere in the Manual of Concrete Practice. Having the horizontal bars closest to the faces of the walls, as shown on your sketch, also helps.

Just a thought, maybe you can treat the concrete with Xypex to fill any cracking that may occur.
BAretired (Structural)
10 Apr 11 1:16
It seems to me that the best way to prevent cracking is to prestress the wall prior to backfilling.  A prestress of about 100 psi should be sufficient to prevent cracking.  This can be achieved with concentric tendons placed along the center of the curved wall, anchored at each end.


hokie66 (Structural)
10 Apr 11 2:02
I missed your post at 0017.  A cylinder is a lot different than a part of a cylinder.  As soon as you cut it, the hoop stress is not pure compression.
Zambo (Civil/Environmental)
10 Apr 11 6:00
Hope I'm not too off topic because I haven't read all of the above posts, but.......

I doubt post-tensioning will work here. As hokie mentions the wall is restrained at the base. If the base is restrained there will be no shortening while stressing, without shortening there is no prestress in the member.

The post-tensioning stressing force applied to the curved wall does not tend to cause a straightening. It acts more like the brake cable of a bicycle. Pull your brake handle and the brakes are applied, the cable (a wire in a sleeve) itself doesn't move from whatever curved shape it is in.  
hokie66 (Structural)
10 Apr 11 6:49
I don't get your analogy with a brake cable.  Surely a draped tendon causes deflection, therefore bending.  Otherwise you wouldn't be able to balance deflections in a slab, and you wouldn't sometimes have upward deflection in bridge deck units.  The only case where there is no bending is if the post-tensioning is completely around the perimeter, thus applying a uniform radial force.
JAE (Structural)
10 Apr 11 10:16
If the cable is prestressed prior to concrete placement and cast fully bonded to the concrete there would be NO bending.  Just axial compression.  

Post-tensioned cables, where the bearing is at the ends, would cause bending.  

However, it would be difficult, if not impossible, to harp the cables in place and tension on that kind of curve anyway - so a post-tensioned method would be required and there would be bending.

hokie - On your idea - I would just wonder on the practicality of doing that and ensuring that the friction would be minimial - there's a lot of weight on the base of the wall after all.

Zambo (Civil/Environmental)
10 Apr 11 10:32

correct a draped cable causes an upward deflection i.e. a precamber. This is due to the tendon profile which is designed to provide the required precamber. But in this case it is as you state, the post tensioning is completely around the perimeter thus applying a uniform radial force.

Having said that you have made me think again - in fact prestressing often has that effect on me as it is not a straightforward concept.

So if you look at a "straight" post-tensioned beam the trumplates (end anchorages) will be angled upwards and then the tendon profile will result in a "drape" which will provide an upward precamber. If the end anchorages were angled downwards so that the tendon profile was a downward parabola I do not believe there would be an upward deflection - just compression.

Could be wrong and ready to be corrected.

dik (Structural)
10 Apr 11 11:02
Hokie... I was thinking pre-tensioned for the bonded system, and not post-tensioned with conduits. Tensioning with conduits would be no different as you note. JAE is correct regarding the harping of the strand.  You would require substantial and costly external buttresses.

BAretired (Structural)
10 Apr 11 12:45
I think that the best argument against post-tensioning is the restraint at the bottom.  Maybe using hokie's plastic sheeting or teflon slide bearings, the friction could be reduced to the point where post-tensioning could work.

Contrary to some views expressed above, a curved member subject only to internal prestress will have only axial force if the prestress is concentric at all sections.  See the following link:

I believe that elimination of all cracking is not possible without using control joints spaced at reasonable intervals.   


SkiisAndBikes (Structural) (OP)
10 Apr 11 13:14
Thank you all for the interesting discussion and your suggestions. Below are some of my thoughts on moving forward with the project:
- minimizing cracks and crack width's in the exposed portion of the wall is primarily for asthetic reasons due to the proposed polished concrete finish.
- using supplementary cementing materials is frowned upon by the polished concrete industry for asthetic reasons.
- using a 0.6% reinforcement ratio with 2" concrete cover is possible (15M at 8" o/c each face).
- due to the project size, budget and local expertise in the area, I highly doubt we would be able to implement any form of post tensioning.
- I am in discussion with the landscape architect. There is a centrally located engraved granite stone being placed right in front of the wall, which if extended a few feet higher, could conceivably cover a vertical control joint at the centreline of the wall. Much easier with two 37'-6" wall sections.
- I am a huge fan of curing, curing and more curing.
- I am still pondering Hokie's comments re removing the base restraint. Overturning stability is not a significant problem with this wall.

I do not think hairline cracks will be a significant problem. Yes, me and the concrete polishers will know they are there, but to the casual observer walking by or spending time at the memorial, they likely won't even notice. Larger 1/8" or 1/4" cracks would be visually problematic as you would see them after repair. While I'm not a fan of external coatings in these types of situations, there has been some discussion by others that an external 'faux stone' coating could be applied in the event of an unsuitable result. I am recommending a test panel be created and approved, prior to construction of the wall so that the desired polished look is agreed upon by all involved.

Thank you all for your input, discussion and ideas.

hokie66 (Structural)
10 Apr 11 17:55
Thanks for that reference, BA.  You have made me scratch head and reconsider.  Although in some of the shapes shown, it would be impossible to keep the strands exactly centred all the way around, in a circular arc it could work.  As well, when you think about it, this thing when post-tensioned would work as a classic circular arch, so bending would be negligible.

Removing base restraint is the way prestressed concrete water tanks are built.  I think they generally use rubber bearings, and there is no structural connection of the walls to the base slab.
Zambo (Civil/Environmental)
10 Apr 11 20:23
I've been using this topic to try to understand the theory of deflection due to post-tensioning. What I know is that if you have a series of box beams in a multi-span structure when you come to a curved section the beams will still have an upward deflection when stressed but there will be negligible deflection in plan.

So I withdraw the brake cable and end anchorage angle theories. But I'm pleased to see the link provided by BA. In the curved beam scenario the upward deflection is caused by the design of the tendon profile to prestress the beam to resist the moment caused by dead and live load. However in plan the tendon layout causes an axial compression and no horizontal deflection.

Is that right? Anyway it looks as if prestressing isn't under consideration.

BAretired (Structural)
10 Apr 11 22:08
You stated


However in plan the tendon layout causes an axial compression and no horizontal deflection.

I think it would be more correct to say that the tendons cause an axial compression and no bending.  They do cause a slight horizontal deflection because the radius shrinks slightly under the compression.


Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Back To Forum

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close