ABSTRACT This paper describes the development of an improved method for Noise and Vibration (N&V) chassis dynamometer testing using Road Load Simulation (RLS). Powertrain-induced noise and vibration testing on a chassis dynamometer has commonly been conducted using fixed loads or simplistic load versus speed approximations. Simple speed control and load control dyno test conditions are largely sufficient to provide representative noise and vibration performance assessment when the powertrain and its controls are insensitive to differences between the actual road load and the dyno load. With the recent growth of advanced engine control systems and hybrid powertrains, more representative road load simulation is required to ensure proper operation of the increasingly sophisticated and diverse powertrain and chassis control systems. Proper exercise of these control systems often determines the quality of the noise and vibration data. An example of this is provided for a full pedal acceleration test to measure interior sound pressure level. Noise data are acquired using a simple steady state speed control dyno test, and RLS real time run-up test. Sound pressure variation trends in the data from each test method are examined and the major source of variation is attributed to thermally-induced engine spark control changes caused by extended test duration under full engine load. Data from the RLS test method show greatly reduced run-to-run variation, when the number of consecutive test runs is limited to prevent thermally-induced spark retard. These results show that the improved RLS full pedal acceleration test more accurately represents the on-road performance, and results in a more robust powertrain noise assessment on the chassis dyno. Although the results presented include primarily sound pressure measurements, the advantages of the RLS dyno test method are also applicable to powertrain-induced vibration measurements.INTRODUCTION N&V CHASSIS DYNAMOMETER TESTING The use of specialized N&V chassis dynamometer testing to measure powertrain-induced noise and vibration has become commonplace for both vehicle manufacturers and major suppliers. Dynamometer testing has several inherent advantages over on-road testing. On a dynamometer, the environmental factors such as ambient temperature and humidity can be controlled much more closely than on the road. In addition, the load and speed conditions applied to the vehicle can also be precisely controlled via the dynamometer rolls. Due to the controlled test conditions afforded by chassis dynamometer testing, it is generally expected that test data acquired on the dyno are more precise than on-road data. This expected higher precision of chassis dyno data does not always exist in practice, as will be shown. Significant run-to- run variation is sometimes observed in dynamometer testing, and one of the contributors to the observed N&V data variation is thermally-induced powertrain performance variation. DYNAMOMETER CONTROL MODES To understand the background of current N&V dynamometer testing practices, some dynamometer control history is useful. Prior to the implementation of modern digital dynamometer control systems, two basic control modes were often used to manually adjust the dynamometer conditions to gather data across a vehicle operating performance range of interest. These two control modes were often referred to as Speed Control and Load Control. Under Speed Control, the dynamometer speed was manually controlled either at a fixed speed or at a manually ramped speed for run-up or rundowns. Under this condition, the load was controlled by manipulating Road Load Simulation Testing for Improved Assessment of Powertrain Noise and Vibration2011-01-0924 Published 04/12/2011 Mark A. Gehringer and Eric J. Defenderfer General Motors Company Copyright © 2011 SAE International doi:10.4271/2011-01-0924 SAE Int. J. Engines | Volume 4 | Issue 1 1210Downloaded from SAE International