Materials testing and characterization is often a lengthy process. It can take more than a year and billions of testing cycles for a manufacturer to characterize the properties of a new metal compound to be used in a critical application such as a component of an automotive or jet engine.

Typically, the testing regimen is prescribed by an OEM in accordance with procedures specified in industry organization standards; for example, one testing method for bending fatigue of gear steels is outlined in SAE Standard 1619 using a standard fixture and standard gear design. Companies manufacturing automotive and aerospace engines often refer to this and other standards as they specify the materials and designs of gears to be used in their applications, as do companies manufacturing test systems that verify production gears meet the standards.

The prescribed testing can also be used in a pre-production environment as companies evaluate changes in materials designed to improve the characteristics of end products; for example, in the case of engine gears, the goal with fuel consumption in mind is to reduce assembly weight without decreasing the strength of critical components. Another goal is to increase power density — the ability to deliver more power through the gear train without increasing the size of the gears involved. One company that is providing test systems for use in this field is Symbrium Inc. of Raleigh, NC.

Figure 1. A Symbrium gear test stand. Symbrium's test control system controls multiple test stands simultaneously.

When gear strength is tested, research labs apply both static and cyclic forces to gear teeth, looking to see when cracks start, and measuring how long it takes before a gear tooth fails. “Static forces are applied while measuring tooth strain or tooth deflection to determine yield and ultimate strength of the gear steel. Then dynamic forces are applied to determine the fatigue life or endurance limit,” said Wes Blankenship, President of Symbrium. “Our customers study different alloys and processes used in gear manufacture to try to increase fatigue strength.”

Test System Alternatives

There are two main approaches to testing automotive or aerospace components. One is to apply a general-purpose test platform that is connected to a load frame created for the specific device. The second approach is to use test systems that are totally customized for the particular application. Blankenship favors the latter approach. “General-purpose test machines are very expensive to be doing a simple test like exercising a gear tooth,” he said. “By building a dedicated machine to do gear testing (Figures 1 and 2), we can increase the speed of the testing operation and enable testing of multiple devices in parallel.”

Figure 2. Close-up view of a Symbrium gear testing fixture.

In order to operate multiple testing operations at the same time, Blankenship looked for an electrohydraulic motion controller that could control multiple motion axes simultaneously. Another critical need was a controller with a very short control loop time so that subtle changes in how the device under test responds to the test cycle can be detected between test cycles. After building a spreadsheet containing info on different alternative motion controllers, Blankenship selected the RMC200 motion controller (Figure 3) from Delta Computer Systems (Battle Ground, WA). The RMC200 is capable of running control loop times as fast as 250 microseconds.

Figure 3. Delta's RMC200 can control up to 32 motion axes simultaneously.

How the Controller is Used

“The motion controller drives a high-end servo valve, applying cyclic loading 55 to 100 times per second at 100+ kilo-newtons (20-some thousand pounds), which it controls to within ±1 newton,” said Blankenship.

The dynamic loading applied to the gear tooth is measured using a load cell connected to the motion controller, while the position of the hydraulic axis is measured using a non-contact laser displacement sensor that is capable of sensing distance down to a tolerance of two microns in order to detect a break.

Figure 4. A crack in the gear tooth is visible at the top center of the photo.

When the system detects a crack (Figure 4), it can either stop immediately or maintain the cyclic loading very accurately until the tooth breaks, enabling the tracking of crack propagation. “Being able to detect a microscopic crack prior to the sample actually failing is very important to our customers because they can do more advanced analysis on the gear,” said Blankenship. “It's the speed of the motion control processor that makes this possible.”

Besides closing the control loop and adjusting the load on the gear tooth under test over 1,000 times/second, the controller has enough spare horsepower to perform additional functions as well. “We used to use a PLC to handle safety checking, but we threw the PLC out and can do the machine 100% with the Delta controller outfitted with an I/O module,” said Blankenship. “I'm comfortable doing it and can get CE approval without the PLC. The controls are more streamlined with a single processor in the system.”

The RMC200 was programmed using the RMCTools software provided for free by Delta to support its motion controllers. “The controller was very easy to program,” said Blankenship. “The heavy lifting came in optimizing the control algorithm.” For that, the Symbrium engineers used special tools provided in Delta's RMCTools package, guided by assistance from the Delta team.

The Results

Figure 5. The Symbrium control module is capable of operating multiple test stands at the same time. On top of the module enclosure is an RMC200 controller.

Whereas other gear testers are capable of running 40-Hz test cycles, Blankenship's purpose-built load frame is running tests for Fortune 500 powertrain companies at speeds up to 100 Hz. “For the price of one off-the-shelf test frame, we can handle three test frames that each produce results three times faster,” he said. “That's a nine-times cost-effectiveness advantage over other testing approaches.”

The capacity of the Symbrium design is actually even higher. With the RMC200 included, the design can run up to six test stations off the same multi-axis controller (see Figure 5). “An additional benefit of automating the test process is that we can run the tests unattended,” said Blankenship. “We can detect very minute cracks and shut the machine down within one cycle without operator intervention.”

With the success of the metal gear test system, Symbrium has moved on to develop a smaller variant of the machine, producing 20 to 500 pounds of force that can be used to test the strength of gears made out of plastics.

This article was written by Reid Bollinger of Delta Computer Systems, Battle Ground, WA. For more information, visit here.