Continuous Ramp Load Four-Ball Testing with Controlled Force Application and Objective Detection of Extreme-Pressure Transitions—A Systematic Study

This is the authors' abstract. We don't add key points for this paper.

Conventional Four-Ball extreme-pressure (EP) testing determines lubricant transitions through short tests at discrete force levels, combined with post-test wear scar evaluation. This study presents a continuous ramp load procedure in which the contact is run-in at 392 N and 1760 rpm for 60 s before normal force is increased linearly while measured force, friction and speed are recorded. Validation of the method uses a broad set of ASTM proficiency-testing and independently established reference fluids. Interrupted tests link the first significant friction transition to the onset of seizure-related surface damage. A central requirement is preservation of the measured normal-force trajectory during the event; on the pneumatically controlled system used here, the ramp remains near-linear through major friction and displacement changes without a corresponding seizure-induced drop in load. The precision of this ramp linearity has been found to be less than 10 N throughout regime transitions. To reduce operator dependence, an algorithm is used to identify the onset of frictional transitions. The basis for the algorithm is Savitzky–Golay polynomial filtering, which suppresses high-frequency noise while retaining local transition shape, followed by derivative-based event detection and an operator qualification step, to identify the onset of physically meaningful transitions. The continuous-test results remain closely anchored to the conventional Four-Ball load framework, such as the Last Non-Seizure Load (LNSL) and weld load (WL), while resolving these transition loads on a continuous scale instead of the discrete steps. A close correlation is found between the friction transitions and LNSL/WL for a set of nine oils with known ASTM results, either from ASTM PTP oils and greases investigated or other commercially available reference fluids, with a Pearson coefficient, r, of 0.955.