Follow us on Twitter      

Forum Message

Heading: Test Reports
There is a natural tendency to point fingers whenever concrete strength test results are lower than expected. The concrete supplier and testing laboratory often point at each other. Sometimes the contractor is blamed for moving test samples or inadequate curing. The admixture company, cement manufacturer or aggregate supplier usually comes next. The situation is exacerbated when the test results arrive late and without notice that they fail to meet a specified strength level. Finally, everyone’s opinion of what is acceptable is asked, and usually answered incorrectly.

I believe one answer to this problem is to report concrete test results as efficiently, accurately and completely as possible, including cautionary notices at early ages and non-conforming notices at specified ages. Also, data should regularly be analyzed statistically for compliance to ACI-301 Sections 1.6.7 (Acceptance of concrete strength) and (Revisions to concrete mixtures), and standard deviation and within-test coefficient of variation should be readily available for troubleshooting problems. This information would be invaluable for early and measured response to individual low test results or downward trends in test results rather than an emotional reaction from a defensive posture.


david breitfeller

David points to a problem which I turned my attention to solving more than 20 years ago but which, unfortunately, remains to become official policy in USA.
There are at least 3 aspects to the answer:
1)Results should be analysed by multivariable cusum graphs, in addition to direct time line, multivariable graphs (multivariables should at least include density, temperature and slump). Such graphing will normally cause any testing procedural deficiencies to "stand out like a sore thumb" and will usually reveal the cause of any genuine downturn. I have recently been requested by TAC to review the latest draft ACI 214 proposals for the evaluation of test results and TAC have required a re-write as a result of my recommendations.
2)Certainly results should be promptly reported. They should appear on the control graphs within 24 hrs of any test taking place. Impractical? I was doing this as an independent service to consulting engineers in Melbourne, Australia in the early 1980s. Properly approached, testing laboratories are pleased to save time, effort, and expense by electronic reporting rather than preparing and posting paper reports. Generally a strength downturn should be detected from a very few 7day (or earlier) results long before any low 28day results are obtained. Confirmation that the downturn is genuine is usually obtained from seeing the cause of the downturn on the graphs of other variables long before it can be statistically confirmed by analysis of the strength results themselves.
3)As a result of the system I originated, Australian concrete producers have been carrying out their own testing and QC for more than 20 years, including amending mixes as necessary. The practice has spread to many other countries and hopefully even USA will eventually realise how much more effective this is.

The above is of course far too brief for full comprehension by the uninitiated but is explained in great detail in 3 editions (1995, 1999 and 2006)of my textbook. However the explanations are available at no cost (along with details of the book!)on my website
Ken Day


You are exactly right. It is critical that 28 day test data is shared as early as it becomes available. There are internet based tools that are available for this - CMATS originally developed Ontario ready mix association is one and there are others.

Another critical area is early age (1 to 7 day) testing and estimating whether 28 day breaks will be low and if so adjusting the mix design through a lower w/cm. Unfortunately this practice of adjusting mixes is not allowed in most US specs which require 28 day test data before making even small adjustments to mix proportions. I think the solution is either performance based specifications or at the very least bands around submitted mixture proportions expressed as +/- certain percent of cement or water weight for example.

Inappropriate initial curing is an important reason for low strength problems. So it should be made clear in appropriate ACI, ASTM standards who is responsible for maintaining the initial curing temperatures between 60-80 F at the jobsite for the first 24 to 48 hours (implication in most specs is that it is the responsibility of the testing lab). ACI 301 is addressed to contractor and if I understand it correctly it only states that contractor should provide space for initial curing specimens.
Karthik Obla

I think it is important for the ready mix industry to proactively educate engineers as to the significance of cylinder results. If you ask an engineer why cylinders are required, they sincerely believe they are testing to evaluate the individual building element. Most engineers I talk to in educational programs do not understand the purpose of cylinders. The few sets of cylinders made - sampled, cast, cured selectively and distinctly different from the building - are for the purpose of judging potential, not reality. It is the intent of the project concrete testing program to develop data for the overall evaluation of concrete. "If you want to know the quality of the actual concrete, take core samples."
This briefly describes the problem. The solution is education on engineers, one by one.
ACI 318 gives the statistical requirements the cylinders must meet. The testing agency must provide everyone involved with the appropriate data analysis early on for proper QC. Today, it should be available online to project principals for immediate review.
Roy Keck

Roy - Actually, CTL Engineering has a proprietary in-house program that has been very well received. The progam is web-based and follows requirements of ACI 301 in Sections 1.6.7 and for reporting and statistical analysis of test results. It also has automatic e-mail notification of cautionary or non-conforming test results to a password-protected distribution list. It is really set up for testing laboratories like ours to inventory compressive strength samples, test them and report the test results. Our project managers review and approve reports online. Everything but the field report is paperless, including the final report. We don't pay postage and our clients don't wait for the mail. All reports can be indexed and downloaded for saving rather than filing paper reports. We are trying to decide whether to market our program for others to use.
david breitfeller

Mr. Keck brings up a very good point that the ready mix industry needs to educate not only the engineers but also project managers and owners as to the importance of test reporting. Over the years we have encountered testing firms that under the direction of one of the aforementioned do not send test reports to the producer. In my opinion as a producer we will be more proactive in helping correct possible problems if we know about them upfront instead of 28 days or later. I'd be interested in knowing if other producers have experienced this and how they handled it.
martin sublett

I think I replied to david's second post previously, pity the comments don't all appear together. ACI methods of analysing test results are lagging behind other countries. I am currently trying to convince ACI 214 (analysis of test data) that they should advocate MMCQC (multigrade, multivariable, cusum QC - read about it on my website Using this any downturn in strength can usually be positively detected on the first few 7 day results (of mixed grades).
In Australia the testing is done by the concrete producer (inspected by NATA (National Assn of Testing Authorities). Check testing by the purchaser is allowed under their Standards and used to take place, but has been largely discontinued as unnecessary and inffective. If the purchaser or his representative engineer detects a downturn early, he does not have enough evidence to insist on a mix change but the producer will react if he knows low 28d results will follow (cash penalties should be specified to make sure he does, but I can't convince anyone of this yet).
As Roy says, cylinder testing is to monitor the quality of concrete being produced and detect change ASAP - NOT to show whether an individual truckload is OK.
P2P should mean Purchaser to Producer as to who controls the concrete, not only Prescription to Performance as to how it should be specified.
Ken Day

Ken makes some great points with all of his posts. It seems that we are not progressing in our ability to measure real time results. Our affection for the slump cone and the test cylinder is really the anchor that holds us back to the Delmar Bloom era.
As I mentioned in a post re iCrete elsewhere, we fear artificial means of analysis, and we fear leaving the cylinder. If we designed structures as we test and evaluate concrete, we would still be using our slide rule.
Roy Keck

Ken - is right on several accounts. This forum has an unsual way of handling postings. One initial post, can have six different threads going at the same time. Your never quite sure whether your reply is going to end up.

Ken - in Australia does each ready mixed producer have their own lab? If they're suppling 20 jobs on a particular morning from 3 different plants, do they send techs to each job? Do small producers hire independent labs to do the testing? I would prefer this type of testing since I believe our in-house lab produces fairly consistent reasults.

Thomas Adams on another thread, brings up an interesting point on test results. How much do we trust the testing labs?

I've seen test results where there is 1000 psi difference between two 28 day cylinders from the same set. Or reports where the 28 day results were lower then the 7 day results for the same set.

I've seen lab technicians not bat-an-eye when they broke core samples that looked like the Leaning Tower of Pisa. I've been to pre-job meetings where I asked how the cylinders were going to be cured and the response is always "We have an insulated curing box." Great for keeping cylinders from getting to cold, but when I ask how are you going to keep them from getting to hot, I get blank stares.

I've opened curing boxes the day after a 1500 yard pour, finding them stuffed with cylinders, that are hot enough to fry an egg.

I've seen test cylinders made on the Friday before a three day weekend still sitting on the job on Tuesday morning.

I've seen test labs that turn off the fog to their moist closets and leaving the doors open during the day because technicians are constantly going in and out to get cylinders or to add new cylinders.

These were not isolated incidents nor where they small independant testing labs. And all of them have NAVLAP accreditation. I sure everyone on this forum has seen similar types of things.

Getting back to Mr. Breitfeller's original comment. It would be helpful if building codes required that test data be given to producers immediately. ACI 318 and ASTM C 94 should require the same thing. ASTM C 94, reads that concrete test data "shall" be given to the manufacturer and then weasle out by adding "in a timely manner." Is 28 days timely enough?

Mark Williams on another thread said he been told that labs can not give out the data to just anyone for legal reasons. And on this point, I think the test labs are correct. The data is confidential and belongs to the person paying for it. Therefore, only if the building codes require that data be sent to producers will things start to change.

One final thing, how many of you have submitted mixes to projects after being awarded a contract, and had then rejected because you "need to have 30 sets of data on the same mix as per ACI 318"? Even though you submitted three-point curves to establish standard deviations. I sometimes wonder if the engineering community understands that ACI 318 requirements are for determining the standard deviation for similar mixes produced at a given batch?plant.

ASTM C-94, which is reference by the Building Code and so is mandatory, states:

6.6 The purchaser shall ensure that the manufacturer is provided copies of all reports of tests performed on concrete
samples taken to determine compliance with specification requirements. Reports shall be provided on a timely basis.

The problem is that the Purchaser is not usually the same person that purchases the testing services. In fact, the purchaser may be a sub-contractor and not have a direct relationship to the Owner. Until ASTM C-94 is revised to place responsbility directly on the Laboratory, there will continue to be a disconnect in this requirement.

Jay Shilstone

Jay - you are absolutely right that since it is referenced in the code, c-94 is part of the code. I've used that very arguement when trying to get on the mailing list for strength results.

What I'm saying is that few engineers or building officials or owners are as familiar with C-94 as they might be with UBC or IBC. So instead of having to argue that C-94 is part of code and therefore must be follwed the same as the code. Why not simily add a line to the codes that specifys "all concrete data will be given to the producer as soon as it becomes available."

My experience is that the labs are willing to turn over the data, if someone relieves them of the legal reponsibility of turning over something that doesn't legally belong to them.

I have been reading the notes on concrete test cylinders and the reports. In my opinion we need to have the testing labs cure the test cylinders by C 31/C31 M-08a and in the moist environment by the immersion method as note 5 of C31.Parg. 10.1.2 Page 9 of Volume 04.02 and remove the other 5 initial curing procedures (2), (3), (4), (5), and (6). They are out dated.

The testing labs should be held liable to properly cure the concrete and testing the specimens. The general contractor should not be responsible for any of this. He should only be responsible to give the lab a place to set up and pay for the power for a curing box.

I have used a Thermal II Curing box @ $ 3500 each and have Easy Cure by (Engius) boxes @ $1700 each. These keeps the cylinders in water at 73 +/- . This helps eliminated compressive strength problems. The box was placed by the producer. Where the box was used the cement efficiency was over 10 psi per pound of cementitious material with a coefficient of variation of less 8%.showing good Quality Control. On another project the laboratory did not use the curing box and the efficiency was down to less the 8 and coefficient of variation of about 16%. Many of the test cylinders did not making the 28 day strength, resulting in very bad Quality Control.

If we go to P2P or the icrete method we need to hold the testing laboratory liable for the testing. There are many labs that do it right and many more that do it half way. The labs need to keep the same personal on site and set up the curing boxes and do it properly. Another option would be to have the testing laboratories post a bond like the sub contractors. When it comes to testing concrete we let the labs do what they want to. It is time we make the laboratories responsible in ACI 301 and / or ACI 318. We are at 2009 and not 1959. We have come a long way in 60 years.

Kozeliski Kozeliski

Of course labs should be properly regulated, but you will find it much easier when they are are owned and operated by, or in partnership with, concrete producers.
Perhaps America should consider a reciprocal personnel exchange with NATA in Australia (which is where effective laboratory regulation began in 1947 - a world first). My website contains a paper describing how I compared two labs testing the same concrete. One relatively cheap comparison exercise is to have a double set of specimens cast by one lab but cured and tested by two labs.
If a lab is still not performing well on a second assessment, after failing a first, it should be de-registered, at least for a period, or alternatively "inspected to death" at its own expense, until it reforms.
Ken Day
   Click Here to Reply    Areas
Advertise Here