The Association of American Railroads (AAR) standard M-1001, “Design, Fabrication, and Construction of Freight Cars” describes the tests to be performed and the required test conditions. This standard is in Section C, Part II of the AAR Manual of Standards and Recommended Practices.

Chapter XI, “Service-Worthiness Tests and Analyses for New Freight Cars,” of M-1001 specifies a number of certification tests including:

• Hunting (vehicle lateral instability)
• Constant curving
• Spiral
• Twist, roll
• Pitch
• Yaw, sway
• Dynamic curving

Because this trailer was designed to not only be used on roads, but also to be transported on railroad tracks, it had to also meet the requirements for rail-compatible vehicles spelled out in Chapter IX of M-1001.

At the heart of the test system SPE devised for these tests are two SoMat eDAQ data-acquisition systems. Other components in the system included:

• Two Simultaneous High Level Layers (EHLS), which allow the eDAQ to simultaneously sample 32 differential analog inputs (16 channels per unit);
• 14 SoMat Smart Modules, which provide strain gage signal conditioning for 1/4-, 1/2- and full-bridges;
• Six SoMat ICP Signal Conditioning Modules.

The eDAQs were mounted in weatherproof NEMA boxes attached to the transition flat car, as shown in the image. Ahead of the flat car is the engine which powers the system on rail. Attached to the flat car is the trailer and rail bogie being tested. In order for the trailer to be transported by rail, the trailer is attached to a bogie, which follows the trailer. One or more trailers with bogies may be attached to the train for testing purposes to simulate actual operation.

SPE used a number of different sensors to measure important parameters. Three vertical accelerometers were used to measure vertical excitation due to track perturbations, impacts or other phenomena. Three lateral accelerometers were used to determine if lateral accelerations met the requirements specified in Chapter XI. 

To measure deflections, and ultimately determine the roll of the units and flat car, SPE used six potentiometers. The longitudinal load at the bogie hitch plate to trailer connection was determined using a specialized load pin. This data was used to determine the longitudinal loads through the system.

Finally, strain gages were applied to both the bogie and trailer. The locations for these strain gages were determined by doing a finite-element analysis of the design. SPE engineers identified fourteen particularly critical areas during this analysis, four on the trailer and ten on the bogie.

To power the system, SPE installed two SunWize SW90C solar panels on the flat car. The solar panels provide up to 90 W each. The solar panels charge a bank of three batteries, which is then connected to the data-acquisition system and ultimately powers it.

SPE decided to use solar panels instead of larger batteries for this test system because it was more cost-effective to use smaller batteries and it eliminated the need to replace batteries in adverse weather conditions. Some of the tests were run during the winter, and battery life is much shorter in colder weather. 

The solar panels also eliminated the need to check battery capacity while running the test. They provided more than enough power to keep the batteries charged throughout the duration of the test. Using solar panels meant that having enough power was one less thing that SPE had to worry about.

Once the test system had been assembled, and technicians had put the sensors in place, SPE engineers used SoMat TCE software running on a laptop computer to program the system, as shown in the image. This software allowed the engineers to quickly and easily configure the channels of the system, including calibrations for each of the transducers. Once the setup was complete, they disconnected the computer and, the data acquisition system ran on its own.

The on-track testing took place on the Canadian Pacific Railroad tracks between St. Paul, MN and Superior, WI, a distance of about 150 miles. Over a period of about two weeks, SPE ran two unloaded trips and two loaded trips. Sensor readings were taken at a sample rate of 200 Hz. Despite being unattended, the test ran without a hitch, and the data acquisition system recorded between 2 and 4 Gigabytes of data on each run.