Passing concrete cylinders but failing cores

[aymanb]

I have high strength concrete cylinders that broke fine at 56 days yet cores taken of the same concrete came in below 85%, as low as 56% of specified strength. ASTM procedures were followed across the board during coring. Temperature has been mild since placement (within a range of 65 to 80 degrees F). Can anyone think of a reason for this?

 

[Ron]

Yes...there can be lots of reasons for this.

The concrete cylinders are intended to reflect the properties of the mix design as delivered to the jobsite. They are fabricated, cured and tested under standardized procedured and conditions.

The in place concrete can be significantly different in strength than the test cylinders.

You say that they came up to strength in 56 days. If you had to wait 56 days to achieve the design strength, you already have lower strength concrete than was desired.

Was there any water added to the mix after the cylinders were taken? Was the concrete pumped? What was the frequency of sampling/testing? You probably only took one sample for every 50 or 100 cubic yards of concrete. That's a lot of concrete that you didn't sample that has gone into the building. What were the weather conditions at time of placement? Hot? Cold? Humid? Dry? Was there any curing done for the in-place concrete? (not just spraying some curing compound on the surface, but actual moist curing)

Give us a little more info to work with here.

 

[aymanb]

The specified strength is 12,000 psi at 56 days. Specs require sampling every 100 CY. Weather conditions were perfect, dry with a temparature arround 70% F. No moist or other type of curing was done. The concrete was placed in a foundation wall 18 inches thick.

You point out a potential reason in questioning the frequency of the sampling. Any thing else you can think of relative characteristics o high strength concrete?

Thank you for your response Ron.

 

[Ron]

Does this concrete supplier have a record of being able to produce 12000 psi concrete with consistency? That's a high compressive strength that requires tight control over aggregate quality, aggregate consistency, cement consistency and properties, particularly fineness and chemistry. I would check with the supplier and see if they can produce statistical validation for the mix design.

You can do some correlation testing using cores and Windsor probe to check the in-place strength. You might also consider using low-frequency ultrasonics (pulse-echo) to check the consistency of the strength, since I assume you have access to both sides of the wall.

 

[aymanb]

Thanks Ron will do.

 

[BigH]

Did your mix contain flyash?? Sometimes, strengths are chosen at 28 days based on 100% portland cement. With the advent of substituting fly ash for the cement, 28 day strengths will be lower. As Ron indicated that the moist cured cylinders always had sufficient moisture to keep on hydration - perhaps your structure didn't and since fly ash gain strength is, in most cases, dependent on the cement hydrating first, . . .

 

[aymanb]

The mix is as follows:

Mix proportions per cubic yard

Cement 375 lbs
Water (batched) 30.1 gal
Fine Aggregate 1160 lbs
Coarse Aggregate #57 1210 lbs
Coarse Aggregate #8 645 lbs
Admixture (BASF/Glenium 7500 - High Range) 85 fl oz
Slag 525 ibs
Silica Fume 50 lbs
Chloride Ion 0.11% by wt. of cement

 

[bb29510]

make sure your cores are broken right, they have to air dry for 48 hours

 

[aymanb]

Did that...followed ASTM to a T

 

[Ron]

Are you sure about the mix proportions? You have an extremely low w/c ratio...barely enough for hydration to occur. If you have any aggregate that is not saturated, surface dry at the time of mixing, you'll pull even more water away that would cause hydration.

I would not use this little amount of water in a mix, since EVERYTHING would have to be perfect for proper field hydration.

Have a petrographic examination done on a core and see what the ratio of unhydrated cement is. I think you'll find an issue with that.

Further, your slag cement to portland cement ratio is very high. We generally do not like to see more than 50 percent of the cementitious material as slag cement. When you have that much slag cement, it will retard the strength gain significantly, particularly in the field as compared to the lab.

 

[bb29510]

note: test cylinder will always break higher than cores because cylinders are taken in the perfect world. perfect mix, perfect cure, perfect transportation. core are the real world. finishers method. curing, even how they take the core.

 

[concretemasonry]

Just a note -

Cylinders are not always perfect, but they do represent a measure of the properties of a mix under more controlled conditions.

The technician does have some control despite the procedures required (loop-holes). When rodding the the rod can be slightly extended. It can be drawn out when not truely vertical. Timing is also a factor. These are not tricks, but common errors. For a true comparison, the same technician should do all cylinders.

Job site is not always perfect, but it is reasonably so.

Transportation is not always perfect because not everyone straps cylinder to a pallet so they cannot tip. Fortunately, the handing mistakes are usually obvious by the results to exclude them.

When we had a concrete technology class we made cylinders and unknown to us, the professor allowed the usual tricks. Some were inserting a 10" long rebar into a cylinder (a 10" long straighten coat hanger makes a big difference). Sugar and salt were the common forms od pranks and sabatoge. Fortunately, the professor knew them all and used his background to show what could occur in the real world.

Cylinders are certainly a good measure of the mix properties if you have a good technician and lab.

Cores only measure the conditions the mix was subjected to after batching.

Dick

 

[conceng]

I would certainly look closely at the handling and testing of the cores. The current ASTM C 42 requires the cores to be immediately dried of free water, dried to a visible dry condition, then sealed to prevent moisture loss and set for 5-days to allow the moisture gradient to dissipate. We have found that testing immediately after sampling, the moisture gradient may cause a strength reduction of up to 25%. I would think the reduction may even be greater with your very very low w/c ratio.

I'd also get the slag supplier involved. I think the % slag used is really pushing the limits, but that also makes it very fun and interesting. Please let us know what you find out.

 

[Jabson]

One other issue to look at is the coring process in itself. if the cores show signs of poor core work, that could also affect your results. Check to make sure the cores are smooth, with no 'ribs' caused by a wobbling core machine. I know I should say this, but make sure they didn't get any reinforcing steel within that core, that will lower strength also.

That mix in itself is also a bit odd. Seattle has mixes that achieves that strength on a consistant basis, but we don't use Silica Fume, way too variable.

Regarding the sampling, have you thought of batch plant inspection? Also, make sure that the cores taken are also from the aproximate location the cylinders were taken.

Other than that, do a petro on both cores and cylinders. Also, with that much slag, was the column contained a few extra days within its form, or did they 'pour and pull', stripping the form the next day? Remember that slag will reduce the initial hydration, so leaving on a few extra days will give it that bump needed.

And one last thing, gotta ask, did the tech break the cores according to ASTM regardin rate of load?

 

[bb29510]

thats only a 4 sack mix, sure is low for 12000 psi

 

[BigH]

Further to Jabson's post - Neville describes how to "accommodate" in your core assessment if there is steel rebar present in one of his books.

 

[msucog]

something looks off on the mix to me

typically (95% of time from my experience) the cores break higher than the cylinders (keeping in mind you're looking at two different strengths and are not necessarily appropriate to compare directly to each other--only compare "apples to apples" thought process)

also, what are you coring? where did you core? how many cores versus how much concrete placed versus how many sets of cylinders?

and i don't want to sound like i have all the answers, but cylinders and cores are rather "no brainers"...given things can happen but for my company and from all my experience, it usually boils down to the contractors actions in the field. sometimes it's the mix and sometimes the difference is somewhat attributed to simply the randomness of sampling and/or difference in curing. very very seldom do i ever see that the cylinders and/or cores were or even could be the problem.

i'd focus on what the contractor may have done after the concrete left the truck...but still take at look at all aspects of the mix and testing results as well.

 

[theCorkster]

With compressive strength this high, did the supplier or anyone else prepare a trail batch? I've been asked to design 10,000 to 15,000 psi concrete for different projects; I just won't do it without a laboratory and plant mix trial batches. There are so many things that can go haywire. I'd want to find out if the mix was designed (and stamped) by an engineer, and I'd want to see some data on the results of the design.

Petrographic analysis will let you know the specifics of what is is you've really tested, and a worthwhile investment.

Then you can start to figure out why the apples are oranges!