Using Parametric Simulation to Optimize Suspension Design
A project to design and implement suspension improvements to Amtrak’s F-40 Non-Powered Control Unit (NPCU) cars is described. The cars, built from former F-40 locomotives, had a history of poor ride quality. Rail Sciences Inc. (RSI) inspected one of the cab cars and measured its ride quality. Peter Klauser modeled the vehicle in NUCARS™ and validated the model against the test data. The vehicle response was primarily in pitch and bounce modes. To optimize the suspension, Klauser simulated vehicle response for a range of four suspension parameters: primary stiffness and damping, and secondary stiffness and damping. Nearly 2600 suspension combinations were considered. Simulation file setup and data analysis were performed automatically using parameter analysis software interacting with NUCARS™. The result was a five-dimensional response contour for each output variable, such as the engineer’s seat vertical and lateral accelerations, and car body acceleration. The most cost effective stiffness and damping parameters were selected from the response contour and translated into component specifications. RSI then provided Amtrak with new axle box springs and dampers, and re-tested the vehicle. The test result closely followed the predicted results from the simulation. Engineer’s seat vertical and cab lateral accelerations improved by 42% and 32% respectively for the worst-case conditions in the test territory.