ArvinMeritor and LMS Engineering Services develop accelerated proving ground and rig test scenarios

To consistently deliver durable and safe vehicle subsystems, ArvinMeritor regularly benchmarks its durability engineering processes, methodologies and technologies against industry best practices. To support this continuous improvement strategy, ArvinMeritor recently partnered with LMS Engineering Services to define a proving ground test schedule and an accelerated front axle rig test scenario. The project resulted in an 8-day accelerated rig test to validate the durability performance of an axle system against a one-million-mile customer usage target. The team’s extensive project experience and use of LMS TecWare load data processing software helped a great deal in achieving accurate results and fast execution.

ArvinMeritor is a leading supplier to the global automotive industry, with headquarters in Troy, Michigan, and approximately 31,000 employees distributed over 25 countries. The products ArvinMeritor delivers to the world’s leading passenger car and light truck, as well as medium and heavy duty truck OEMs include axle, brake, suspension and exhaust systems and modules. In addition, ArvinMeritor strifes towards becoming a one-stop source for technologically advanced drivetrain systems for medium and heavy-duty trucks, trailers, buses and off-highway equipment. To shape its role as a total solution provider, ArvinMeritor systematically searches for new ways to further optimize its complete chain of design, engineering, validation, manufacturing and integration.

Developing compact and light-weight designs with sufficient fatigue strength in record development lead-time leaves no room for lengthy field testing to validate vehicle system durability. To meet a 500,000-miles durability quality target, for example, vehicle field tests would take 3 years to complete at an average of 500 miles a day! ArvinMeritor selected LMS Engineering Services to perform a durability engineering project on a recently developed ArvinMeritor front axle and suspension design. Phil C. Kittredge, Director – Experimental Mechanics and Test Operation at ArvinMeritor in Troy, Michigan commented: “The scope of the project consisted of the definition of a proving ground schedule representing a specific one-million-mile user profile followed by the development of a corresponding accelerated durability test scenario for a front axle rig test. We selected LMS for its strong automotive engineering expertise, and more specifically, its experience in the area of advanced load data analysis and test schedule development for durability track and rig tests.”

Meeting a one-million-mile customer usage profile

The available road load data for the ArvinMeritor front axle and suspension durability validation project was acquired using a test vehicle that was equipped to measure the brake pressure and 3-dimensional spindle forces at the front wheels. The tests involved the completion of 10 tracks of public roads located in Michigan, Indiana, Iowa, Nebraska, Wyoming, Utah and Nevada, altogether adding up to approximately 2,300 miles. The tracks represented a mix of highway as well as city traffic and rural roads with steep grades and unpaved sections. In addition, the vehicle was subjected to 19 different events on the NATC proving ground, driven twice in clockwise and twice in counterclockwise direction. LMS engineers consolidated the acquired data using LMS TecWare software, by performing visual data inspections as well as data consistency checks to verify signal unit, sign and length. LMS TecWare was also utilized to efficiently trace and correct data failures, such as signal drifts, spikes and offsets.

In a first phase, the project concentrated on developing load data that accumulates the same damaging effect as a targeted customer usage profile of one million miles, starting from the 2,300 miles of acquired road load data. An initial effort in this regard consisted of the decomposition of the vertical dynamic spindle forces into low-frequency forces, representing vehicle maneuvers, as well as high-frequency forces, indicating low road quality. Engineers also decomposed longitudinal forces, depending on whether the vehicle was actually braking or not. The brake pressure was measured because braking maneuvers generate different load flows in the axle suspension compared to cruising conditions. In its capacity as a brake system manufacturer, ArvinMeritor also has a special interest in tracking how many times and under which circumstances brakes are applied.

“To actually match the targeted one-million-miles customer usage profile, engineers faced the challenge of extrapolating the damage content of each track and event as well as superposing all resulting rainflow matrices,” Phil explained. “In the extrapolation step, engineers computed the damage potential at one million miles using a sophisticated probability relationship (developed by LMS and field-tested with OEMs) that is applied to the damage potential for 2300 miles. The probability relationship takes into consideration that variability in loads exists and produces results consistent with data at a full one-million miles with statistical confidence. For the characterization of individual proving ground events, they applied ranking matrices for single dynamics, following the common ranking method. In addition to achieving an accurate approximation of the target in terms of total damage, engineers also succeeded in establishing an acceptable match with the target matrix on matrix subsector level.”

Further optimization with LMS TecWare allowed ArvinMeritor engineers to select only 7 out of the 76 proving ground measurements that were actually needed to accomplish an acceptable match with the targeted customer usage profile. This simplification yielded an optimized test schedule at the proving ground, which drastically reduced the overall runtime from about a year to just 450 hours. The seven proving ground events – which serve as the basis for the axle rig test scenario definition – included Belgian blocks, gravel-to-bumps, alternating bumps, gravel, and bumps-to-embedded-rocks.

The second project stage focused on the development of an accelerated axle rig test schedule that further reduced the runtime from 450 hours to less than 200 hours. Besides additional flexibility and productivity, rig testing offers the possibility to test an axle system on its own, instead of having to build in the axle into a running vehicle. Besides limited load amplitude upscaling, engineers compressed time series by applying Range Pair (RP) filtering for multiaxial loading. As a consequence, the frequency content of signals above 30 Hertz slightly reduced when compared to the original load data signals. This is of minor relevance to axle systems, for which the most important frequencies typically range between 0 and 25 Hertz. Ultimately, the duration of the rig test sequence was reduced to just 185 hours, while maintaining 97% of the total damage content.

“The know-how to develop compressed testing cycles and reproduce an equivalent laboratory test is essential in enhancing our current durability development processes,” Garrick T. Hu, Vice President – CVS Advanced Engineering, ArvinMeritor concluded. “Through accelerated rig testing, it becomes possible to subject assemblies to more accurate representations of realistic road load data than was feasible before. The project confirms that one single week of continuous rig testing is sufficient to validate an axle prototype system on the basis of a realistic one-million-miles endurance test schedule. The cooperation with LMS project engineers enabled us to gain additional confidence in identifying customer usage profiles and to validate the accuracy and robustness of fatigue prediction and customer correlation based technologies. Overall, the project enhanced our ability to implement the targeted durability performance in context of ever-demanding requirements for compact and light-weight designs with top-class durability performance.”