The first two articles in this series highlighted key issues in the preparation for onsite testing of fresh concrete, selecting the appropriate location of sampling fresh concrete, air tests, and initial onsite storage of test cylinders. Assuming that valid tests have been performed on concrete that was properly sampled by certified lab and field technicians the question becomes: What do I do with these test results?
The obvious answer is that those results should be compared to the specified acceptance criteria, but this turns out to be more complicated than it might seem. The keyword here is “specified” meaning as written in the original contract documents, or as added to or modified by a formally accepted change. The acceptance criteria in question may be listed in the project specifications manual produced by the design team, listed in “general notes,” or otherwise shown on the drawings.
Within the specifications manual, it is likely that acceptance criteria are “included herein by reference” by citation of one or more industry standard documents or specifications such as ACI 301, “Specifications for Structural Concrete” or ASTM C94, “Standard Specification for Ready-Mixed Concrete.” But because such standards are updated frequently (often with substantive changes), it is critical that the contract documents indicate exactly which version of these documents apply.
The cleanest way for the specifier to do this is to cite the publication date of the governing standard. A cheaper and more obscure way to do it is to refer generically to “the version in force at the time the contract documents were issued.” A worse way to do it is to refer to the “version in force at the time of construction.” This option gives a lot of mid-contract spec-changing power to a bleary-eyed committee making changes to standard documents as the work progresses.
Rules of the game
Once the rules of the game (the acceptance criteria) are clear, further reading and connecting the dots to referenced documents will reveal the details of precisely how the concrete is to be sampled, which tests are to be performed, and how those tests are to be performed (much of this was discussed in the first two articles). Keep in mind, though, that nearly any provision of these “model” specifications can be overruled if otherwise stated in the contract documents.
These rules of the game (ACI 301 and ASTM C94) also include a lot of other important information. For example, both have clear requirements that test results are to be reported to the owner, architect/engineer, contractor, and concrete supplier. ACI 301-10 also makes it clear that “when it appears that material furnished or work performed by contractor fails to conform to contract documents, the testing agency will immediately report such deficiency to architect/engineer, owner, contractor, and concrete supplier.”
ACI 301-10 goes on to state that “the owner’s testing agency and its representatives are not authorized to revoke, alter, relax, enlarge, or release requirements of contract documents, nor to accept or reject any portion of work.” When encountering apparently unacceptable work or materials, therefore, the testing agency is to inform all the principal parties so that those parties can make decisions about how to proceed. If ACI 301 is specified, the testing agency cannot throw any concrete, concrete truck, or concrete work “out of the game.” But by quick action, the testing agency is expected to alert the folks who do have such umpire-like authority.
It is also important to know that both ACI 301 and ASTM C94 require that when test results, such as slump or air, are out of compliance with specified values, the first step is to perform a “check test” to verify the initial reading. A few years ago, I went in for a routine physical and when the nurse took my blood pressure she immediately hit a hot button on the wall that brought the crash cart and crew, “Stat!” My BP had pegged the meter and I began to fear that I was one bag of Doritos over the line. As folks were scrambling, a senior nurse suggested that we quickly try another pressure-meter before sending me to the ICU. Sure enough, the initial meter was out-of-calibration and my BP was only slightly higher than normal, due to all the excitement.
I recalled this event sometime later when I managed an elaborate field test of variability in air content, and knowing that I needed six properly functioning concrete pressure air meters, I asked the testing agency to bring seven, complete with certified operators. I started the day with a group hug and an air meter calibration session, and found only six meters and techs made it to the site, and two of those meters were out of calibration. Of the two “bad ones,” one could be fixed in the field, and the other was down for the count, leaving five trustworthy devices. Of course performing a checktest with the same bad meter, bad scale, or bad thermometer that gave the first unacceptable reading is not of much help, so once again, details make the difference—make sure the test equipment is correctly calibrated, and do the specified check test before calling for the crash cart.
Keeping the team informed
Although the test results of fresh concrete are known immediately, at least by the test technician, results for compressive, flexural strength or hardened air-void parameters, for example, are not normally available until formal test reports come out. In its mandate to keep the whole team informed, ACI 301-10 requires that “test and inspection results” be provided “within seven days after tests and inspections are performed.” I periodically run into owners or designers who prefer to keep the contractor in the dark regarding concrete test results, either out of some deep-seated control issues, or for fear that if the concrete turns out to be better than required (and as paid for), the concrete producer might modify the mix so that it is just exactly as required (and paid for).
The news flash here is that both the ACI 318 Building Code and the ACI 301 Specification provide for controlled modifications of a mixture that is over-performing. Besides, a mix that is stronger than required may be producing more heat of hydration at a larger carbon footprint, or exhibiting more shrinkage than is necessary. The same Building Code and Specification also require immediate modifications to a mixture that is under-performing, but this correction process cannot be initiated if the contractor and concrete producer are not in the loop of timely distribution of test results.
Tracking test results
Let’s say the work is progressing and test reports are being provided routinely. What do the contractor and concrete producer do with them?
Step 1, normally, is to give the report a quick glance to look for a red stamp that signals an unsatisfactory result.
Seeing none, step 2 often entails use of a three-hole punch followed by insertion into a dusty binder and relegation to a shelf in the office trailer. (A favorite task of summer interns and junior engineers.) But limiting data-processing to merely sorting test results into the categories of “go” and “no go,” causes the contractor and supplier to miss a key opportunity to control the product, and can lead to serious problems later on. Can you imagine playing any competitive sport and not keeping track of trends over the season? Can you imagine a golfer not knowing whether her game was improving? Could you manage a baseball team without knowing which pitchers were getting stronger and which are getting weaker?
In all sports, the trends are tracked routinely in addition to wins and losses, and the same needs to be done with concrete test results. Diehard fans know how the team stats change as the season progresses. Do you know how your concrete strengths change as fall turns to winter? Coaches know the reliable players. Does your workhorse mix have a reliable strength and air content, or does it change with every pour?
And just as in sports there are all sorts of statistics that you can track: strength versus air content, strength versus concrete or air temperature, or the chronological record of strength over the duration of your project. This latter option actually is required by ACI 318, ACI 301, and ASTM C94 in the two fundamental criteria for acceptance of concrete on the basis of compressive strength. To quote ACI 301-10,
“The strength of concrete is satisfactory provided that the criteria of 126.96.36.199.a and 188.8.131.52.b are met. 184.108.40.206.a: Every average of three consecutive strength tests equals or exceeds the specified compressive strength f´c. 220.127.116.11.b: No strength test result falls below f´c by more than 500 psi when f´c is 5000 psi or less, or by more than 0.10 f´c when f´c is more than 5000 psi.”
To evaluate these criteria, you need to first recognize that a single strength-test result is not the result of breaking a single cylinder, rather it is defined by averaging the strengths of at least two 6x12-inch or at least three 4x8-inch cylinders. Next, criteria 18.104.22.168.b limits the value by which such single test results can fall below the specified strength. At the same time, criteria 22.214.171.124.a looks at the bigger picture by requiring the running average of three consecutive strength tests always equal or exceed the specified strength.
Let’s consider a concrete mixture for which there are test results for 37 6x12-in. cylinders, cast over a number of placements and broken at an age of 28 days. The specified 28-day strength is 4500 psi. The field test technician was careful to make a pair of cylinders at each desired age at time of testing, so that for each batch (truckload) of concrete that was sampled, two cylinders were subsequently tested in the lab at 28 days.
In each case, the average of those two cylinder strengths is reported as the single 28-day compressive strength of that batch. (If more than two 28-day cylinders are available for that single batch, the average of all of them can be used.) The strength of that batch and of subsequent batches would be considered acceptable if each two-cylinder average from the batch is not lower than the specified strength by more than the specified strength minus 500 psi (in this example 4500-500=4000 psi).
But to guard against a systematic run of low strength values, requirement 126.96.36.199.a demands that the average of the strength values for the first, second, and third batches equals or exceeds 4500 psi. For the concrete mixture to remain in compliance, the same requirement applies to the average of the second, third, and fourth batches; the third, fourth, and fifth; and the fourth, fifth, and sixth; and so on. Among other things, this means that unless you track the trends, the only way you can be sure your concrete meets the specs for compressive strength is if every single cylinder break, without exception, exceeds f´c. (See graph.)
Of course tracking your results requires that you “do the math,” but you will probably find that your interns and field engineers will actually enjoy and benefit from the task. In many cases, the testing agency can make the data available to you in the form of control charts (such as the example shown here). You will not only be able to know exactly how you are doing in terms of compliance, but you also will know if you are trending higher than needed or frighteningly too close to the line. If you are troubled by being out of the loop of timely receipt of test data, point out the standard specification language that makes that information available to you. Remember, you have the right to acquire and manage your concrete test data. If you give up that right, anything you don’t know, anything you did not find out, and anything you didn’t do to correct the situation can and will be used against you.
Kenneth C. Hover is is professor of civil and environmental engineering and Stephen Weiss Presidential Fellow at Cornell University, Ithaca, N.Y. He currently is serving as president of the American Concrete Institute.