Cylinder break values out of whack.Whats going on?

[03LabGrunt]

Ive been breaking 4x8 cylinders for a very large project that requires a design strength of 8000 psi. We are getting passing #s. However, every once in a while (maybe once or twice a day) I'll get 2 samples from the same mix design that have psi breaks that are 16 to 25% difference between the 2 samples. Ive already made certain that there is no mixing of samples from different pours.The curing process is the same for all the samples.So that rules out that possibility. Anyone have any clues as to why this would happen? I will consider any well meaning answer.
Thanks in advance.

 

[civilperson]

Banged during transport, i.e. dropped or kicked, normal variability from segregated mix, (not remixed after sampled).

 

[henri2]

In the context used, are "samples" to be construed to mean "concrete cylinder specimens" as opposed to samples of fresh concrete obtained from say, a ready mix truck, in accordance with ASTM C 172?

 

[boffintech]

When you write "the curing process is the same for all the samples" you mean:
1) initial curing at the job site is the same for all the samples
2) standard curing performed at the lab is the same for all the samples
3) both

If initial curing was performed at the job site, what exactly did this initial curing consist of?

Also, how are the 4x8 samples prepared for testing: sulfur caps or rubber pad caps?

 

[03LabGrunt]

Henri2,

Samples meaning "cylinder specimens" ASTM C-39.


Civilperson,

I did consider your conclusion. It seems the most likely, however, you'll never get a tech to admit he drop kicked a sample...

Thank you both very much.

 

[03LabGrunt]

boffintech,

Usually 24 hours at the most curing on site. Then transported to the lab the next day and stripped and tanked in the "curing room".

So, the answer I guess is 3. The initial curing process on site actually not known to me. I assume they are in a safe location nearby the initial pour. I believe thats how it is done.

We use rubber only for all breaks.

Thanks for the response.

 

[boffintech]

1) Instead of trying for "passing #s' why not try for accurate meaningful results.

2) Regarding job-site initial curing, you believe incorrectly.

ASTM C-31
10.1.2 Initial Curing— Immediately after molding and
finishing, the specimens shall be stored for a period up to 48 h in a temperature range from 60 and 80°F [16 and 27°C] and in an environment preventing moisture loss from the specimens. For concrete mixtures with a specified strength of 6000 psi [40 MPa] or greater, the initial curing temperature shall be between 68 and 78°F [20 and 26°C].

Your breaks are all over the board because undoubtedly initial curing is non-existent.

Why don't you know what the initial curing conditions are?

Who was at the pre-con meeting and why hasn't your inspector documented the initial curing conditions so that the project team will know what is going on?

Why hasn't the EOR dropped the hammer on:
1) the contractor for failing to provide job-site conditions suitable for gathering accurate meaningful test results, and
2) the testing lab for failing to report the failure of the contractor to provide initial curing conditions suitable for gathering accurate meaningful test results.

Also in ASTM C-31:
The storage temperature shall be controlled by use of heating and cooling devices, as necessary. Record the temperature using a maximum-minimum thermometer.

Anybody doing this at the job? If not why not? And if not why not reported.

 

[DarthSoilsGuy]

you should be documenting init curing like BT is saying. you don't want to be stuck with a "hot potato" when push comes to shove on any project. you should expect a response similar to BT's by the concrete company if this were to happen, even if the field curing is adequate as it very well may be (summertime and shaded).

you didn't tell us if these are 7-day breaks or 28-days. Variance at 7 is not as unusual.

shooting in the dark, i'll throw a few things out that haven't been mentioned

1)is the aggregate for the mix design too big to test 4x8. there is another similar post to find the max agg size. this would make variance between cylinders occur as the cement paste/agg matrix could be less homogenous along the failure planes.

2)is it loaded too fast sometimes? this would be an operator error at the break machine, but would have a "hammer" effect on the cylinder. be tactful looking into this. you obviously have honest staff reporting to you and wouldn't want to inspire any "editing" in the lab.

3)intermediate break, with some of the high-strength stuff. you can get a fracture in a piece of the aggregate. (this makes a loud noise) the load pressure goes down b/c of some displacement of the piston but.. many times the load returns higher as the paste has not fractured. if you switch off the machine after the 1st drop you won't know it. you don't have to ever worry about this if you report the failure type as a,b,c,d (crush it until it's a mess always)

this may be old news already, i've had the browser in the background for a while today.

cheers,
DSG

 

[03LabGrunt]

excellent info thanks.

 

[henri2]

As Boffintech correctly stated, the initial curing temperature has to be between 68 and 78°F [20 and 26°C]. And since the heat generated during early stages may exceed this requirement, were specimens immersed in lime saturated water or other precautions taken as recommended by C 31? Refer to the note below the sub-section dealing with Initial Curing. It is either Note 5 or Note 6 depending on the edition of C 31 you have.

 

[dirtsqueezer]

Could be that the concrete isn't mixed well in the wheelbarrow that your tech samples from. It gets worse when it's very fluid, as it tends to segregate, but can also happen to drier mud. If they're by the book, they'll be taking two composite samples from the truck, one before and one after the truck is halfway discharged. If these samples aren't mixed well prior to molding, you will get weird numbers in the lab. If they're using very fluid mud, it's hard to keep it homogenous, so that could also be a clue.

 

[boffintech]

If you are making 4 cylinders per set, 1@7, 2@28, 1@reserve, break the reserve cylinder with the 2@28 and see if it improves the percent difference. One 4x8 cylinder at 56-days is worthless anyhow.

 

[dirtsqueezer]

I'm sure you already do this, but you could also save one 28-day cylinder for a 56 if the first one doesn't make strength. Pretty standard, I just didn't see it covered yet.

 

[msucog]

what were the results like in the mix design submittals? how many cylinders are you talking about (1 out of 100, 1 out of 20, 1 out of 3, etc)?
for a mix like that, i would probably guess variability in the sampling. stay on your guys to take care in handling the specimens and to follow astm as closely as possible.
also, if i'm not mistaken, aci 214 has different guidelines for statistical analysis of the results. table 3.2 is for <5000psi and 3.3 for >5000psi...the differences are fairly significant in my opinion and (to me) pretty much says that the variability will be greater for the higher strength mixes than for you're normal mixes. (it appears by almost a factor of 2 to 2.5). run the statistical numbers on all the breaks and see how it compares on the chart.