Methods of Highly Repeatable Lubricant Testing for Electric Vehicles: Universal Motor Control and Statistical Test-Variation Analysis

This paper describes a systematic approach to evaluate lubricants for hybrid and electric vehicles (xEVs) that can detect impacts on efficiency as low as 0.1 percentage points. Two testing methods were developed to evaluate lubricants' efficiency effects: (1) on a complete vehicle (using the manufacturer's hardware and motor control) and (2) on a standalone drive unit (using custom power electronics and control). A Monte Carlo simulation was used to analyze the resulting data to determine the detection limits of the vehicle test method. To evaluate the effectiveness of the test stands and the data-analysis method, a Tesla Model 3 electric drive unit and a Chevrolet Bolt battery electric vehicle (BEV) were characterized for system efficiency. For the Bolt mounted on a hub driven chassis dynamometer, this method is capable of detecting a change in the drive unit's electromechanical efficiency between baseline and candidate fluids of <0.4 percentage point (pp) with 95% confidence at most of the operating points analyzed. This method can be applied to high-precision testing applications, such as lubricants or additives, cooling systems, gear trains, dynamic seals, bearings, or electric motor designs.