KEY CONCEPTS
• ASTM D2596 covers the determination of the load-carrying properties of lubricating greases. 
• Varied acceleration rates across machines performing ASTM D2596 have resulted in significant impacts on the repeatability and reproducibility of the test results. 
• In 2023, ASTM started two task forces to evaluate the machines in usage, determine what could be understood about acceleration rates across machines and identify next steps in addressing the challenge.
• Investigations by ASTM and end-users are underway to determine a path forward and are expected to have a significant impact on the grease industry. 

ASTM D2596 is the Standard Test Method for Measurement of Extreme-Pressure Properties of Lubricating Grease (Four-Ball Method). The four-ball extreme pressure test for lubricating greases has been a staple of standard laboratory testing in the industry since its original publication in 1967. This test is made up of segments where each segment lasts for 10 seconds, but there can be 10+ segments in a full test with assembly/disassembly/cleaning between each. This test serves as a measure of the load-carrying properties of lubricating greases and is used by many grease manufacturers and marketers across the globe as a key performance marker. 

Because of the significant usage of this test, many instrument manufacturers have developed machines to perform the evaluation. These machines have evolved significantly in recent years with transitions from devices using steel weights and analog dial readouts to fully automated, digital devices. However, as machines have evolved, challenges related to repeatability and reproducibility have started to emerge. A 2021 paper demonstrated that a critical factor in the consistent performance of greases in the test across instruments is the acceleration rate of the test spindle during the test.¹ By running a faster or slower acceleration rate, the results of the test can vary significantly. 

Acceleration rate is not a defined parameter in ASTM D2596. Because of this, machines have been built with many different motor drive designs and likely have significant variation in acceleration rate. In 2023, ASTM started a task force to evaluate the machines in usage, determine what could be understood about acceleration rates across machines and identify next steps in addressing the challenge. As a critical grease performance parameter, the results of this task force and other work within ASTM will have a significant impact on the grease industry. 

ASTM D2596
STLE member Mathias Woydt, managing partner at MATRILUB in Dahlem, Berlin, Germany, notes: “The four-ball test methodology was developed by G. D. Boerlage ca. 1931 at the laboratories of N.V. de Bataafsche Petroleum Maatschappij (Shell) in Delft, Netherlands.”² Woydt says, “ASTM D2596 (greases) and D2783 (oils) were initially approved in 1967.”^3,4

ASTM D2596 covers the determination of the load-carrying properties of lubricating greases. Testing involves three determinations:

• Load wear index (LWI) that was formerly called the Mean-Hertz Load
• Weld point
• Last non-seizure load (LNSL)

As is standard for ASTM methods, the test values are SI units and the test standard does not address all safety concerns associated with its use, leaving the user responsible for establishing safety, health and environmental practices.

However, Woydt notes that the test conditions according to DIN (Deutsches Institut für Normung – German Institute for Standardization) and ASTM differ. Woydt says, “As long as users mainly used the standard test devices for the respective test methods, there were practically no questions about the impact of equipment on test results, even though these have evolved over the decades, such as sensors, load control and drive technologies.”

Variability in ASTM D2596 results
STLE Life Member Mike Anderson, tribologist at Falex Corp. in Sugar Grove, Ill., notes that the four-ball test is widely used throughout the world “in specifications and the development of greases for having extreme pressure property requirements.” However, while noting that the D2596 test “is an easy test and provides relatively quick determinations” of LNSL, LWI and weld point, Anderson notes that concerns about the test acceleration rate have been recognized for years.

Woydt reports: “There is already widespread dissatisfaction with the four-ball apparatus technique in terms of reproducibility and comparability.”

STLE member Dirk Drees, CEO of Falex Tribology in Rotselaar, Belgium, has observed variability in results of the ASTM D2596 test across generations or across manufacturers of four-ball tribological machines. He has seen that “specifically, the weld load seems to be very sensitive to the acceleration, as well as to the continuous power to the motor.” 

Drees notes that in 2021 a presentation was planned for the ELGI Annual General Meeting describing a number of equipment-related factors that influence the weld result of a four-ball test.⁵ That research indicated that the major influence found on the ultimate weld load appeared to be the acceleration of the motor to nominal speed. The authors found that an acceleration shorter than 0.15 seconds would be sufficient to obtain consistent results.5 The COVID-19 pandemic led to a cancellation of the meeting, but meanwhile another 2021 paper confirmed those findings.¹ 

The paper planned for the ELGI meeting showed that a grease with a 400 kg weld load could reach more than 800 kg by reducing the acceleration to 1.5 seconds for reaching 1,770 rpm.⁵ The authors state that the power management to a variable drive motor is also very important.5 Contrary to a fixed speed motor, which will rapidly stall when the weld load creates a high torque, a variable drive motor may continuously inject more power when more torque is demanded by the frictional contact. This effect can also move or mask the weld load.⁵

Drees says, “For most older machines with a fixed motor speed (i.e., so no motor controller is necessary where the motor is directly wired into the 50 or 60 Hz grid), the weld loads are very consistent over different machine manufacturers and generations.” However, he has observed: “Where variable speed machines are introduced (e.g., to perform both ASTM type tests at 1,770 rpm, and DIN type tests at 1,500 rpm), the problems may appear.”



Drees provides examples of anecdotal evidence of major challenges that were communicated to him, all related to variable drives testing equipment:
• He reports: “One major grease producer had done a round robin over their own labs and found weld loads ranging from as low as 300 kg, up to 800 kg, for the same grease.” Further investigation by this producer “found that ‘self-altered’ machines, where the company had replaced variable speed drives themselves without knowing the background of this, were giving the too high weld loads.”
• A sudden deviation of weld loads was reported by a research institute. “They found 800+ kg, whereas the grease was supposed to weld at 350 kg.” At this facility, “investigation showed that they had internally replaced the speed drive and programmed it to a very slow acceleration.” As the ASTM test segments only take 10 seconds, their “internal electronics engineer had thought it safe to program a slow ramp of two seconds to reach the nominal speed of 1,770.” This slow ramp resulted “in a slow run-in of the test balls, giving them a much larger contact area, and thus less contact pressure when the load was applied,” postponing the weld event from 350 to 800 kg. Once the acceleration rate was speeded up as fast as possible (i.e., in less than 0.3 seconds), the machine returned to behaving correctly, with weld loads of 350 kg.

STLE member Jacob Bonta, lead formulation scientist for industrial fluids and greases at Valvoline Global Operations in Lexington, Ky., has seen differences in performance of greases in ASTM D2596 in both older and newer four-ball test machines. He has seen “significant differences observed in both the LNSL and weld point across the machines.” He has found this to be “particularly noticeable with high load-carrying greases.” He has found: “The major factors observed have been differences in the acceleration rate of the spindle during the test and skew in the applied loads at high levels on the older style lever-arm machines.”

Woydt also notes that cleaning solvents used in the test apparatus can impact results, stating: “It is well known that the type of cleaning solvent ranks the tribological results.”^6,7 Cleaning test apparatus and type of cleaning agents are not specified in the test methods.

Impacts of ASTM D2596 variability
Bonta sees several different impacts of the test variability in his organization. “The inconsistency in results between older and newer instruments, as well as unknown configurations of machines in third parties’ labs, results in a significant number of repeated tests to confirm results, which increases time and cost burdens in development work.” He finds: “Another effect is the impact on formulations, requiring products to be ‘over-engineered’ to exhibit consistent results on ASTM D2596.” These impacts can create negative effects in other areas of a product’s performance, which can translate to companies being forced to reassess formulation strategies.

Bonta finds these impacts have caused differences in how companies design and formulate grease products. “Companies hold themselves to strict quality and performance standards for their products, so if we claim a 500 kg weld load for a product, we want that result to be consistent wherever it is tested,” he says. This translates into having “to spend extra time and effort on ensuring formulations are robust enough to perform consistently across an inconsistent ASTM D2596 testing environment.” He notes that while forcing companies to spend time and money ensuring their products meet standards across an inconsistent testing environment may seem like a consumer benefit, however, “using extra extreme pressure additives to achieve these results can have negative effects in other areas of product performance, throwing off the critical balance of a lubricating grease.”

Bonta notes that these ongoing testing challenges with the ASTM D2596 method “have consumed a lot of time and energy during the ASTM meetings.” He finds the biggest hurdle in overcoming these problems and identifying the best path forward “is the long history and significant amount of historical data generated using ASTM D2596 that seems to be becoming more and more skewed.”

Improving ASTM D2596
Drees notes that steps can be taken to improve repeatability and reproducibility across four-ball test machines in ASTM D2596. He observes: “Companies and labs spend a lot of time on fine tuning the variable drive systems for the four-ball machines so that the correct weld results are obtained.” These correct results are “not only a matter of acceleration, but also of the sustained torque at higher loads and the torque cutoff so that a weld can develop instead of the machine continuing to rotate when severe adhesive wear takes place, thereby masking the weld event.” 

Drees believes that ASTM D2596 can be improved and that a new technique can be developed for four-ball extreme pressure testing. He suggests: “First of all, a new round robin should be conducted to show which machines, from which manufacturers, are still displaying the correct weld loads.” He believes: “The ASTM Proficiency Testing Program (PTP) results suggest that the majority of participants in this program have correctly working test machines, whether they are old or new machines.” The PTP includes statistical quality assurance programs that enable laboratories to evaluate and improve performance and maintain and fulfill mandatory accreditation requirements.

Drees also suggests: “Some outliers in this PTP should be actively investigated.” For example, where the outliers occur, Drees would encourage asking, “Do they use self-modified speed drives? Do they use non-conforming test machines?” Also, he notes: “Modern machines allow for the measurement of friction during the test intervals, and this information should be added to the reporting.” More emphasis is now placed on friction reduction by lubrication; however, the current method has no information on friction.

Anderson agrees with the impact of acceleration rate on tribological performance in ASTM D2596, stating: “This has been known for many years.” He notes: “The problem is that there are test instruments that span 75 years of production within the industry.” Many of those instruments have not been inspected to ensure that they are performing as designed, and “replacement motors are no longer identical in performance to those originally used.”

On top of older equipment, which can be 50 to 70 years old, Anderson notes: “There is new technology available that allows users and manufacturers to measure the acceleration rate that the units provide.” This new technology shows that “acceleration rates have been found to decrease as the equipment ages.” While this may be “a testament to the durability of the test unit, performance never gets better as the equipment ages out.” Anderson notes: “The newer units, however, have the ability to control and program the acceleration rate.” However, he says, “The problem is, the faster the acceleration, the more severe the test because the slower acceleration rate allows for the development of a lubricant film bearing surface between the test balls allowing for higher test loads.”

New method development
Anderson notes that weld point is the most widely used ASTM D2596 test result, but questions whether it is the best result to use for describing grease extreme pressure performance. Anderson suspects weld point is the most widely used test result because it is the biggest number generated. Weld point “is the test load where the film strength and chemical surface interaction completely breaks down.” Anderson considers: “The LNSL is the better result because it is the highest test load where the film strength, and chemical surface interaction is still intact and providing the necessary protection.” 

Drees concurs with Anderson, noting: “More emphasis should go to the other values that are being measured in the ASTM D2596 method, namely LWI and LNSL.” By knowing these values, Drees believes: “A better description of lubricants in ‘normal working conditions’ can be given.” He also wants “the graph of wear scar versus load to be mandatory in the reporting as the weld load is only a moderately useful value” that indicates the content and efficiency of extreme pressure additives but does not have any value regarding antiwear properties below catastrophic failure.

Anderson mentions there are other tribological techniques that could substitute for the ASTM D2596 method. There are other test instruments that measure the extreme pressure properties of greases; however, he notes that the four-ball test “is probably the most recognized and available throughout the world.” In addition, complexity and cost can be challenges with the other tests.

Conclusions
When asked if changes to the ASTM D2596 method would impact his work or industry, Drees noted: “This would depend on what changes are going to be implemented, but the impact would ultimately be positive when better consistency and more information in a standard test are reached.”

Anderson notes that the ASTM Subcommittee D02.G on Lubricating Grease has two task forces assigned to investigate the ASTM D2596 method challenges. The first task force “is to identify the acceleration rates of existing instruments in the industry and to determine how to correlate the acceleration rate with test results and then to either require the acceleration rate to be disclosed with the test result or mandate the acceptable acceleration rate when reporting test results according to ASTM D2596.” He notes that measuring an instrument’s acceleration rate is not easy, requiring instrumentation specific to that purpose. 

Anderson says that the purpose of the second task force is “to look into the possibility of modifying ASTM D2596 using different load stages, while addressing the acceleration issue.” This second task force “will also look at alternate ways of defining a failure point, which right now, is the weld point.” Anderson notes: “If weld point is going to remain as the desired reporting result, perhaps either ASTM D2596 needs to define an acceptable acceleration range, how to measure it, or at the very least, require reporting the acceleration rate along with the weld point results.”

Additional resources
1. NLGI Lubricating Grease Guide, 7th Edition, NLGI International Headquarters, 4635 Wyandotte St., Suite 202, Kansas City, Mo. Available at www.nlgi.org/grease-guide/.
2. ASTM D2596 (2020), “Standard Test Method for Measurement of Extreme-Pressure Properties of Lubricating Grease (Four-Ball Method),” West Conshohocken, Pa., ASTM International. Available at www.astm.org/d2596-20.html.

REFERENCES
1. Joysula, S. K., Dube, A., Patro, D. and Veeregowda, D. H. (2021), “On the fictitious grease lubrication performance in a four-ball tester,” Lubricants, 9 (3): 33. https://doi.org/10.3390/lubricants9030033.
2. Boerlage, G. D. (1933), “Four ball testing apparatus for extreme-pressure-lubricants,” Engineering, 136, pp. 46-47.
3. Stallings, L. (1968), “The four-ball wear test – ASTM method no. D-2266,” NLGI Spokesman, pp. 396-401.
4. Brown, E. D. (1971), “Friction and wear testing with the modern four-ball apparatus,” Wear, 17, pp. 381-388.
5. Espinoux, F., Bourdin, F. and Genet, N. (2020) “When a lively four-ball crescendo takes on a weld!” CRES-TOTAL Research Center, Chemin du Canal, F-69360 Solaize, France, in ELGI Eurogrease publication, 3, pp. 21-28.
6. Woydt, M. (2001), “Influence of test parameters on tribological results – Synthesis from round robin tests.” Taken from Totten, G. E., Wedeven, L. D., Dickey, J R. and Anderson, M. (eds.), (June 2000), ASTM STP1404 “Bench testing of industrial fluid lubrication and wear properties used in machinery applications,” Seattle, ISBN 0-8031-2867-3, p. 199.
7. Woydt, M. and Ebrecht, J. (2003), “Influence of test parameters on tribological results – Synthesis from international round robin tests,” Tribotest Journal, 10 (1), pp. 59-76.
Andrea R. Aikin is a freelance science writer and editor based in the Denver area. You can contact her at pivoaiki@sprynet.com.