Autogenous Healing of cracks in foundation 1

[Stl63]

An 80'Dia. concrete tank has been constructed on top of a 6'-0 thick conrete foundation, no piles. There was a keyway with 6" waterstop in place prior to the placement of the tank walls. A large amount of vertical rebar embedded in the foundation protrudes into the tank wall on both sides of the keyway.

When the tank was hydro tested with 50' of water, damp spots appeared on the exterior base of the tank wall, the full circumference of the tank. Also, moisture began to rise from cracks in the foundation slab surrounding the tank.

The contractor of the tank, has stated that the moisture will disappear in a month as the concrete hydrates. The tank and foundation were poured several months prior to the hydro-test. I assume the contractor is referring to autogenous healing. I have found little research to back his claim, any experienced tank builders seen this?

 

[oneintheeye]

autogenous healing can take place but I would question why the damp spots appear for full circumferance rather than at isolated areas. Suggests design or construction error?

 

[oneintheeye]

how is the base connected to wall? or is it sliding joint? Need more details really to supply full answer. Is the waterstop used suitable for the pressure head?

 

[Stl63]

Here is a rudimentary sketch of wall cross-section.

 

[JedClampett]

Did you provide enough circumferential reinforcing at the base slab to carry the shear force from the wall? See the PCA's "Circular Concrete Tanks without Prestressing" Table A-12. It's not well explained, but you need to account for that shear in the slab design.
If you were light on that reinforcing, you might get some horizontal cracks through the slab, centered at the wall, that could transmit leakage. The shear force would redistribute to the wall.
As far as autogenous healing, I'm not sure that's what happens, but I do know that a lot of leaks stop over time, maybe due to calcification.

 

[miecz]

JedClampett

Did you mean radial reinforcing, or do you carry the wall shear through hoop tension in circumferential steel? I don't see where PCA says how to handle the shear.

 

[fattdad]

I'm wondering whether you're looking at water from the tank escaping through cracks or whether you're looking at ground water that is worrying it's way through cracks under the pressure of the newly-applied tank-content loading. Where is the water table on this project?

I don't buy the "healing" via "hydration" concept.

f-d

¡papá gordo ain’t no madre flaca!

 

[JedClampett]

What you end up doing is integrating the shear from Table A-12 over the entire circumference. You'll get a force that is resisted by the base. You need to add circumferential reinforcing to carry the force. The force will be pretty large, especially in a tank with your depth and diamater.
The easiest way to explain the force is to consider the difference between a cylinder with only hoop stresses and one that is restrained at the base. See Figure 7 or 10 in the PCA book. That force needs to be accounted for.
Unfortunately, you're right and PCA doesn't do a good job of explaining this.

 

[Stl63]

My role is to accept or decline the sub-contracted tank hydro-test. Which means accepting their explanation of hyrdation closing cracks, and ending dampness (leaking).

I have not determined if the waterstop was sufficently sized, or know the details of the reinforcing in the slab.

My question remains, has anyone dealt with this hydration closure of cracks, or what I have determined they mean as autogenous healing? Is it typcial, or is there a more serious issue? I at first thought of fattdad's explanation of ground water pressure due to content load, but ruled this out due to depth of slab, and location of dampness at location of tank wall and slab connnection.

 

[oneintheeye]

Reasons could me numerous from design error to poor construction. At what time was formwork removed and what mix design was used? As there is continuity of reinforcement into the wall there will be vertical bending moments at this point. There will be high amounts of hoop tension which will only be increased by the fixity at base. Would suggest the obvious cause as it located at the base would be a) inadequate vertical reinforcement for applied moment B) Problems with waterstop C) possibly differntial settlement between slabs opening joints? D) Lack of thought with regards to shrinkage cracking and formwork struck before peak temperature rise in temperature had passed resulting in cracking at restrained points.
On your point about accepting this as a water test from my experience water tests are passed with maybe a small damp spot at corners or joints and these generally self seal fairly quickly not over months otherwise why water test immediatly? You seem to imply large areas of dampness which to me would look like something wrong rather than a local problem. Is your contractor just happy to shrug his shoulders and be happy with this without determining why? Doesn't seem to professional to me. What is the clients reaction to the damp spots?

 

[hokie66]

I think the dampness at the base of the wall would be due to restraint cracking of the wall. The slab had already done most of its shrinking when the wall was cast, so the slab restrained the wall, thus inevitably causing cracking. My guess, and it is only a guess, is that these cracks will heal.

As to the moisture coming up through the external slab, I can't help.

 

[JedClampett]

Do you have a leakage criteria? You need to be judging the loss of water vs. some percentage. We use .001 of the volume over 24 hours (I think), with allowances for gate leakage and evaporation. By my calculation that's .6 inches drop in level.
If you don't have a criteria, use the one in ACI 350.

 

[BarryEng]

To the best of my knowledge, I have only been able to find three articles on the healing of concrete by autogeneous healing. I'm talking about a definitive method of designing a concrete structure so that autogenous healing will take place & NOT "folk law" or "empirical (crystal ball)" methods of guessing if the concrete will heal.

In my day, it was always "if you have a cement content of 320 to 360 kg of cement per m3, with a W/C ratio of 0.5 or less, then cracks of up to 0.2 mm (& even on some occasions, of up to 0.5 mm) will heal." There was never a method given (or known about) to determine what was the limit of cement content, or how to calculate a crack size accurately. I know of a lot of research that has been done to try to calculate crack sizes & spacing (Gergley & Lutz, Base & Murray, Evans & Hughes etc etc) , & these are used in many codes, but there is still not a 'one only' formula that will guarantee the calculated crack size & spacing.

Back to my first para - articles on autogeneous healing. The first was done by C A Clear in 1985 by the CACA in UK land. This is the first article (that I have been able to find) that - for a cement quality & hydraulic gradient (5 m depth across a 500 mm wall is 10 m/m) with a crack size, calculates the time taken to 'heal' (autogeneously) cracks in the concrete.

That article was in 1985 & I consider that it never got the recognition it was due, & I have only in rare cases, see a reference to it.

I have only found two other articles (in 2002 & 2004 - I think). I'm not in my office & I do not have the references with me - I can find them if anyone expresses any interest in this thread.

The two most recent articles shed some light on the mechanism of the 'healing' process. It appears that popular thinking has always been "water leaking out thru' a reasonably cement rich concrete wall, will leach out lime from the cement. The water travels to the outer face of the concrete, evaporates, leaving the lime (after a chemical action with atmospheric contact) deposited as CaCO3 (very hard & white, calcium carbonate)."

The most recent articles indicate that this process is correct BUT IN ADDITION, the CaCO3 is also deposited within the depth of the crack (water path). A good proportion of the reduction of water flow thru' the crack, is by the action of "mechanical blocking, by CaCO3" within the crack itself.

At this stage, I have kept out of the structural concepts that may have led to the water leak (I'll see what other comments come up in this thread) I have confined myself to autogeneous healing only.

 

[BarryEng]

Testing of a water tank.
In the Australian Standard - Concrete Structures for Containing Liquids (AS 3735), in the testing clause is the following - fill the tank slowly, leave it for 7 days, then record a drop in water level (per day) for the next 7 days. During this second 7 day period, the total permissible drop in level (after allowing for evaporation & rainfall) should not exceed "average depth/500 or 10 mm."

 

[BarryEng]

I should have mentioned before - when I said that I had only seen 3 references to autogeneous healing of concrete, I should have said specifically "in relation to the definitive calculations for the autogeneous healing in a concrete tank".

There has been quite a few references in German (I don't speak the language) & also in other publications relating to the healing process, generally.

For a general view of the subject, have a look at: -

A M Neville, “Autogenous Healing – a Concrete Miracle?”, Concrete International, November 2002, pp 76-82.

I pinched this next reference from the Nov07 copy of "Pipeline" from the Conc Pipe Assoc of Oz"

Necessary conditions for autogenous healing
In a recent review, Adam Neville3 summarises literature having a bearing on conditions necessary for autogenous healing of cracks. At an early age, continuing cement hydration in which calcium silicates in the cement are converted to calcium silicate hydrate can play a direct role in the healing process.
In cracks healed at a later stage in the life of a structure, the main product which fills the crack is calcium carbonate, formed by combination of calcium from the hydrated cement with carbon dioxide dissolved in water from the atmosphere or other sources. Neither water hardness nor pH has been found to influence the process of autogenous healing. Various investigators have reported different maximum widths at which cracks have healed – “not surprising because the test conditions have varied widely. In some cases, the cracks were caused by shrinkage, in others by the application of tension, usually flexural but in some tests by direct tension. The age at the opening of cracks varied too. The healing took place in static water or flowing water. There was a head of water or not. The water was fresh or seawater.”
Among examples reported are reinforced concrete pipes that developed shrinkage cracks up to 1.5 mm wide, and were subsequently put into service. Five years later, the cracks were found to be completely closed by autogenous healing.
Hydrated cement is partly soluble in water4, whether or not there is carbon dioxide present. Carbon dioxide dissolved in low concentration from the atmosphere plays a role in the healing process, precipitating calcium carbonate, but is not necessary to initially dissolve calcium from the cement. Autogenous healing will take place unabated in concrete made with blended cement, even though such concrete may be found not to contain any “free lime”. Roberts5 confirms that the type of water and the cementitious material have minimal effect.
For reinforcing steel to corrode, moisture must be present both for its part in the chemical reaction at cathodic sites on the steel surface, and to provide a conducting path through the concrete for the electrolytic current. If the concrete is cracked, the same moisture provides a condition for autogenous healing. Initially, the environment at the root of a crack may be favourable to steel corrosion, particularly if the water contains chloride which breaks down the passive film on the steel. Some rust will be formed, occupying a larger volume than the original steel and perhaps tending to disrupt the surrounding concrete. However given an ability on the part of the concrete to resist such disruption, autogenous healing will progressively restore a situation in which the corrosion is unable to continue. Calcium hydroxide, silicates and carbonate will be deposited in the crack, on the surface of the steel and in the pores of he rust. Of the two competing tendencies – disruption by rust, or isolation of the steel surface, assisted by autogenous healing, circumstances will allow one or the other to predominate.

I have his book at home, on cracking in concrete. He has published many articles on concrete over the years.

 

[oneintheeye]

just looked at british code. States in summary :- tank should be filled at level maintained by adding water if neccessry for 7 days. Filled 2m every 24hrs. The stabilizing period for absorbtion and autogenous healing to take place is 7 days for 0.1mm crack width or 21 days for 0.2mm. Then level should be monitored for 7 days at 24hr intervals. drop in level allowing for evaporation and rainfall should not exceed 1/500 depth or 10mm or specified amount. Any evidence of leakage at outside face should be assessed against spec. Repairs carried out at the liquid face.