You may have read that the Sequence VIII recently become unavailable. This engine test, officially called ASTM D6709, measures weight loss of copper bearings and oil viscosity loss due to mechanical shear. Its off-line now because engine test laboratories ran out of bearings to run it. As a result, the American Petroleum Institute on April 13 declared that any test program for oils claiming to meet categories ILSAC GF-5 and API SN, SM, SL and SJ will be provisional.
That means that an engine oil marketer can run a test program without the Sequence VIII and still receive a temporary API license on the product. However, once the test becomes available again, the programs in question have six months to run and pass the Sequence VIII – or API will cancel the license. It can also demand that the marketer recall the offending oil.
I think that a provisional license is very disconcerting. In the case of the Sequence VIII, its not quite so bad since the test is an older one and is considered by the industry to be relatively easy to pass. Basically, the candidate oil just needs some corrosion inhibitor, and enough shear stability to meet the test requirements for viscosity loss due to mechanical shear.
The greater question is how often things like this happen. I thought it might be interesting to look at other cases when tests have gone rogue – why it happens, how problems are detected and who solves them.
First thing to understand is how engine tests are maintained. Once a test has been proposed and a procedure developed, how do you determine whether or not it is measuring what it is supposed to measure? Establishing the tests time, temperature and other operational characteristics is just the start. Obviously, the test development task force must find a precise yardstick to gauge whether an oil performs satisfactorily. The answer is to have reference oils that can be used for comparison. Reference oils can be used as a baseline for parameters such as wear control, deposits or friction reduction, to name a few, as well as to check whether the engine test itself is straying from the norm.
Every ASTM sequence test method has reference oils. These oils are used to calibrate test stands and to confirm that the test is running as it is supposed to run. Procedures set by ASTM require that every laboratory running an engine test must run reference oils on each test stand at set intervals.
I got in touch with Frank Farber, director of ASTMs Test Monitoring Center (TMC) at Carnegie Mellon University in Pittsburgh, Pennsylvania. Since the TMC is charged with maintaining test fuel and reference oil supplies as well as evaluating reference oil test results, I knew that he could explain the process.
Farber filled me in on the strict procedures for managing reference oil tests. He explained that once an engine test has been created, the tests developers select the reference oils to be used. Typically, they try to select one poor-performing oil to ensure the test can maintain discrimination, and then usually one or two borderline oils of different chemistries that can be used to assess performance. These oils can also be used to make severity adjustments near the pass limit. Thats a way to correct the limits with a bit of a fudge factor, adjusting for minor variations in results.
Farber also noted that the lubricants industry has been trying to include a category reference oil – for example, one oil that passes all the ILSAC GF-5 tests – for each engine test type, to further bolster the monitoring of test severity. SAE viscosity may also weigh into the selection of reference oils. An example could be the new Sequence VIF engine test, which is specifically intended to measure the fuel economy improvement of the new SAE 0W-16 grade oils. Conventional SAE grades will use the Sequence VIE, employing a different slate of reference oils.
How often are reference oils run, and which one should be selected? Is it the choice of the lab or is there some systematic assignment of references? Farber replied that each test has a Lubricant Test Monitoring System with a predetermined assignment rate established by the development task force. For example, 30 percent of the assignments might be on oil A (a poor performer) and 70 percent on the two remaining oils (B and C).
When test stands need calibration, it is the TMCs responsibility to assign reference oils. Farber pointed out that the TMC strives to maintain the specified oil assignment ratio on an average basis. However, they may assign repeats on a given oil if a test fails, or if they note a particular stand is showing high fail rates on a particular oil. Since each test stand is registered, it is important that all of the results be consistent with each other across the industry.
Some oils tend to be robust on deposit performance but may be better indicators of viscosity increase or wear problems, Farber commented. So the intent, he explained, is to assign reference oils that represent different chemistries, viscosity grades and performance (to the extent that the slate of reference oils permits) in order to assess an engine test stand or labs severity and precision.
My curiosity kept me asking for more. First, how often should references be run on a specific engine test? It depends on the test type, Farber responded. Some are determined by the number of tests that have been run. Others are run based on how many hours the engine stand has run since the previous reference, and some depend on both criteria.
One use of reference testing is to verify that new batches of parts are equivalent to earlier parts batches. Each engine type has parts that are critical; these are measured to determine satisfactory performance is maintained. If a sequence tests performance strays when a new batch of critical parts is introduced, the reference oils should flag it – which is exactly what caught the Sequence VIII problem. In this case, it was a faulty batch of connecting rod bearing sets for the Sequence VIIIs engine, the part which is weighed to measure for corrosion.
Farber commented that ASTM Surveillance Panels (the experts who monitor each test method) have good foresight and judgment, and they try to run a small matrix to evaluate the impact of any new parts batch before its introduced into the system. However, he noted that sometimes the economics win out and the new parts batch is brought in along with the next reference. Ideally, the calibration system is not used for parts screening.
As I mentioned, the Sequence VIII is a relatively easy test. That is, if any test is easy. Consider that you are dealing with a combustion engine operating under specific conditions, and that many engines have several hundred parts. What could go wrong?
The Sequence VIII uses a single-cylinder engine operating on unleaded gasoline, which is brewed to specific properties such as octane, aromatics content and boiling range to name a few. The connecting rod bearing set is the object of the corrosion test limits. It must be manufactured to specific parameters so test results can be repeated.
How did we know it has a problem? The results of tests run on reference oils were not in line with the history of reference tests run on prior batches of bearings.
In the case of the Sequence VIII (actually its predecessor, the L-38 engine test), this happened one other time that Im aware of and that was many years ago. The monitoring procedures were different then and the reference runs didnt catch the problem. Needless to say, there were a lot of frustrated formulators who thought the oil was at fault.
Its surprising that there havent been more problems, but Farber could only remember one significant event. That was when the Cummins NTC-400 diesel engine test (which measures crownland and piston deposits) was sidelined because of parts no longer being available. Readers may recall another case back around 2000, when a shortage of parts for the Sequence IIIE engine test also forced API to implement provisional licensing.
In the more recent past, the Sequence VG test for sludge and varnish formation was unavailable for a period in 2011, when the fuel batch in use ran out and the new fuel batch would not give the same results. (Fuel is also a specified part for sequence tests). Ultimately, the batch had to be re-blended to achieve the proper test results with the reference oils.
When you think of engine oils and their performance, it is only through the specific, detailed and costly use of test engines that marketers can demonstrate their products meet the needs of modern engines. ASTMs test developers, the TMC and the surveillance panels who safeguard the precision of these tests deserve everyones thanks.
Industry analyst Jim Lang has noted that where engines once lasted less than 100,000 miles, they now perform satisfactorily for over 200,000 miles, thanks to improvements in engine designs, monitoring systems (think oil life systems here), and engine oils.
Ill add that modern oils backed by rigorous performance tests are key to long engine life. They better be good since most folks are keeping their treasured vehicles more than 11 years. Happy motoring!
Industry consultant Steve Swedberg has over 40 years experience in lubricants, most notably with Pennzoil and Chevron Oronite. He is a longtime member of the American Chemical Society and SAE International, where he was chairman of Technical Committee 1 on automotive engine oils. He can be reached at steveswedberg@cox.net.