Why did axle fail? FEA helps find out.

An FEA program from MSC.Software, Santa Ana, Calif. (www.mscsoftware.com) helped track down the problem. The axles are part of a 450-seat turntable that takes visitors through exhibits at the Hoover Dam. David Dearth, an engineering consultant and president of Applied Analysis and Technology, Huntington Beach, Calif. ([email protected]) got the job of determining causes for the repeated failures. He worked with metallurgist Larry McKnight of McKnight Labs, Santa Fe Springs, Calif.

The 84-ft-diameter table is supported by 66 10-in. wheels and turns at an outer-radius speed of 45 fpm. The support wheels turn on 2.5-in. steel axles and support the nearly 900,000-lb platform. Dearth reexamined the initial design using hand calculations, sizing estimates, and MSC.visualNastran for Windows.

After a metallographic analysis, McKnight suggested that cyclic fatigue led to the frequent catastrophic axle failures. FEA analysis models back up his theory. "FEA pinpointed crack initiation in the fatigue damage. The software also showed the locations of the stress concentrations, and I used the results to generate my own fatigue analysis," he says.

Dearth ran at least a dozen different linear-static scenarios with different load cases. Physical tests showed many sources of crack initiation in the original design, indicating that loading conditions were greater than expected. "This type of cracking is caused by impact loads," says Dearth. "These came from waviness in the turntable, like the curves in a potato chip."

Input from Allan Battenfield, a design consultant, helped Dearth generate a revised modular design that solved the problems. "The original axles used 1020 Series steel with a 79-ksi ultimate strength. Results suggested using a 4340 bar stock, which has a 180-ksi ultimate-stress rating and better fatigue characteristics," says Dearth.