Picasso would have been a terrible technician
By Mike Beavin
Why is it the exception, rather than the rule, that two asphalt binder test results from a seemingly identical procedure are identical? Same material, same standard, same software, same techniques often produce different results. Results that are different enough sometimes result in disputes, penalties and general frustration.
Despite our best efforts, along the way something will almost certainly differ between two labs. Because it’s not always obvious where that inconsistency occurs, a clear, shared interpretation of the published standards and a common best practice is vital to minimize the impact of testing variability in the binder lab. There may be a day when the PAV DSR (Pressure Aging Vessel, Dynamic Shear Rheometer) test results of your lab and those of the state always match. Until then, Asphalt Institute (AI) has assembled the six most likely ways to reduce variability in the binder lab (or any lab for that matter).
1. Know your way around the standards
The American Association of State Highway and Transportation Officials (AASHTO) and American Society for Testing and Materials (ASTM) standards are written in a consistent format and use language and key words intended to be interpreted very specifically. Unfamiliarity with the intended meanings of these key words and phrases can lead to improper and inconsistent laboratory practices. There are many online resources that clarify terminology and format. Care should be taken to make sure any online resource is factual and represents the industry as a whole. A technician should always have ample time to read, comprehend and practice the standards before generated data can be considered reliable.
2. Standardize
Let’s quickly review some important terms.
Standardization (formerly known as verification): A process that determines the correction to a measuring system or instrument or the adjustment that must be made when compared to a standard reference material. (AASHTO PP 57, section 3.7)
What does this mean? For instance, if you have determined that a thermometer is reading two tenths of a degree higher than a reference thermometer – you have standardized that thermometer, but have not yet calibrated it.
Calibration: A set of operations that establish, under specified conditions, the relationship between values of quantities indicated by a measuring instrument or measuring system, or between values represented by a material measure or a reference material and the corresponding values realized by standards. (AASHTO PP 57, section 3.2)
What does this mean? You must ensure that measurements are traceable to the international System of Units (SI). For the purpose of this article, please see AASHTO PP 57 for terms associated with calibration, when to have calibrations performed, and who should perform the calibrations.
The importance of frequent standardizations and subsequent calibrations cannot be overstated when discussing a well-run binder lab. In order for questionable data to be isolated and prevented from becoming part of a larger data set (a less threatening way to say “cost you money”), each binder technician should be familiar with the standardization procedures and how they impact the quality of test results. While acknowledging that it may be necessary for standardizations to be handed off to a designated individual, each technician should be aware of all of the potential causes of variability. Improper standardization is perhaps the leading cause of variability. Armed with this knowledge, the appropriate questions can be asked when a dispute arises.
3. Standardize even more
Once the mysteries of the calibration and standardization procedures have been cleared up in your lab, you would expect accurate, precise and quality data to flow from that point forward, right? Maybe, but as we’ve discussed, sources of variability are always present.
Take this scenario:
According to AASHTO T315-09, section 9.1, you must “Verify (standardize) the DSR and its components at least every six months.” Given this directive, a technician in the state of Kentucky could perform the required standardization procedures on a DSR at the beginning of the paving season and generate data until operations are winding down in the fall. Six months pass and it is time for the next round of standardizations. The technician discovers that the DSR is out of range in temperature by 1.5°C. When did this happen? Who knows?
Given that for every +/- 0.1°C variance from the required test temperature, there can be a 1.2% resulting error in the measured G*/sinδ, a potentially significant problem reveals itself. The potential error in reported data for an unknown period of time may be as high as 18% in addition to the allowable tolerance of 17.0% for multi-laboratory precision estimates. The potential variance from reported agency lab data now stands at 35%. That number speaks for itself. The ramifications can be significant.
If you find yourself needing a refresher course every time you standardize your equipment, you are probably waiting too long between standardizations. AI recommends waiting no longer than one month between DSR temperature and standard fluid checks. Examine all of your in-house standardizations and increase their frequencies to as often as practical.
4. Don’t get creative
Creativity does not always translate to excellence. The genius that drove Picasso’s paintings to sell for millions at auction would have made him disastrous in the binder lab. An urge to think outside of the lines in routine binder acceptance testing is a liability to put it mildly. Do you think that the thirty minute hold in the conditioning bath prior to trimming during the elastic recovery test is just ridiculous? Resist the urge to tweak. Tweaking the standards to best fit your sensibilities will certainly not increase your testing variability. The laboratory setting is no place to get stubborn. If you feel so strongly that a portion of a standard needs fixing, get involved in ASTM or AASHTO and champion that cause. It is better for everybody to perform a flawed test the same way than to have multiple fixes for that flaw.
5. Grade yourself
Precision and bias estimates are included in all of the Performance Grade (PG) standards and are located just after the report section. They are based on test data from hundreds of samples tested in several laboratories and provide a way to “grade yourself” against the inherent ruggedness of the test. With some very simple calculations, you can determine if your data falls inside or outside allowable tolerances.
But, don’t stop there. For instance, the precision and bias estimate for the PAV DSR allows for 40.2% variability between two results submitted by two different operators testing the same material in different labs. As written, 40.1% variability is within tolerance but should we be content with this result? Of course not. Would your parents have been content with a “D” grade in your algebra class? We should always strive to reduce that number and increase our grade by sticking strictly to the test methods and by adopting best practices that are consistent with other labs that perform PG testing.
6. Use control charts
So, there’s a problem. How did we get here? To an overworked technician or lab manager, a control chart can look a whole lot like just another administrative layer. But if they are maintained properly, control charts can be lifesavers.
Referenced in AASHTO PP 57, control charts are designed to recognize assignable and random causes of variability and can be used to identify and control those cases that are assignable. In other words, they can locate the weak spot in the fence before the cattle escape, but they cannot predict when a tree might fall on that fence. If you haven’t already done so, a review of AASHTO PP 57 is highly recommended.
The best way to address questionable data is to prevent the creation of questionable data. We can drastically reduce penalties, deducts and disputes by relying on a set regimen in the lab that reads something like the policies of a hard line dictatorship. Creativity and free thinking = bad. Conformity = good.
Mike Beavin is the Asphalt Institute Technical Training Coordinator. He teaches the National Binder Technician Certification (NBTC) courses.
Contact: Mike Beavin
Read more in MS-25