Medical Device Link.

Originally Published EMDM November/December 2004

ENGINEERING INSIGHT

Although tibial trays rarely fracture, it can be catastrophic for the patient when one does. A fractured retrieval tray is pictured here.

Researchers at the Centre for Biomedical Engineering (CBE), University College London, recently concluded a study of knee implant designs. Sponsored by the UK Department of Trade and Industry, the project examined failure modes and test procedures, ultimately developing an alternative test method to the one that is conventionally used. A finite-element analysis (FEA) program developed by MSC.Software GmbH (München) was instrumental in producing meaningful data and accelerating the study.

Failure modes exhibited in some early knee implant designs prompted CBE to conduct research into the design of tibial components. While a fracture of the tibial tray can be catastrophic for the patient, it is thankfully rare. CBE research resulted in setting a benchmark against which new designs of knee replacement can be tested.

The ISO 14879 test method for assessing fractures of tibial trays was published about 10 years ago, but does not contain any loading criteria. The method calls for the use of a simplified structural model of the joint: one-half of the tibial tray is clamped while the other half receives the load. The test has not been scientifically verified, so CBE’s first goal was to try to determine appropriate loading conditions.

“I used tibial trays with a history of failure and trays that performed successfully, which were provided by manufacturers involved in our project,” says researcher Sunita Ahir, PhD. “I developed models in MSC.Marc software to establish the testing method. Computer-aided design geometry was imported from Unigraphics into MSC.Marc, where it was automatically meshed,” she explains.

Ahir was familiar with MSC.Marc, which she had used in past projects, and determined it to be a good fit for the needs of this study. In particular, she was impressed by the software’s sophisticated meshing capabilities. Automatic hex-meshing is especially useful, she adds, when dealing with bone models that are awkward to mesh by hand.

Predicting Failure

Most of the study relied on the use of conventional FEA. Partner companies provided data on their tibial trays, making it possible to compare different tray designs. The FEA simulations were verified by physical testing.

A joint was considered to have failed when the stress surpassed fatigue limits. The predictive data generated by MSC.Marc correlated with the results of the physical tests. Cracks appeared in the trays at the precise locations that had been predicted.

“In reality, loads vary with patient weight and size, explains Ahir. “But when you’re trying to establish a standard test procedure, you can’t really allow for this kind of variation. Research has shown that the peak walking load is approximately equal to three times a person’s body weight, so a good average figure is 2000 N,” she says.

Ahir sought to establish a failure load by analyzing joints that failed as well as implants that were successful. Based on MSC.Marc simulations, this came to about 900 N, less than the 2000-N load that is generally regarded as a safe design figure.

In the ISO test, half of the tray is clamped, while the other unsupported half carries the load. This would be a severe condition if it were to occur in a patient, requiring a reduction in the physiological load. In real life, both sides of the tibial component provide support, which is why many artificial knee joints survive actual loads of the order of 2000 N even if they are deemed to have failed according to the ISO test method.

The first part of the study concluded that the ISO test can be used with a load of 900 N to validate new designs and to ensure compliance.

The study went on to investigate what happens to the tray in an actual tibia. A model was created of the tibia with an implanted tray. Variations in material properties were used to simulate various arthritic conditions, and contact analysis was performed between the tibia and the base plate.

Under normal bone conditions, the results showed that even trays that failed according to the ISO test would survive in use, says Ahir. “But once fibrous tissue or arthritic conditions develop, the survival of the implant is compromised. Fibrous tissue can develop under the base plate, leading to increased bending of the tray caused by the more-compliant support in comparison with hard bone,” says Ahir. “This can lead to a fracture of the tibial component in poorly designed implants.” This part of the study showed that the implant’s stress-concentrating design features, its thickness, and the fatigue limit of the material play an important part in eliminating fatigue fracture of the tibial base plate, explains Ahir.

The third part of the study focused on developing an alternative test to the ISO method. “I used MSC.Marc again to analyze the base plate supported on two blocks of material,” says Ahir. The blocks were made of materials in varying degrees of stiffness to simulate bone and fibrous tissue. The study showed that by using blocks made from polymers such as Delrin or polyurethane, a load of 2000 N could be used to distinguish between clinically successful and unsuccessful trays.

Although the study could have been done entirely by means of physical testing, it would have required substantially more tests to cover the full range of design variations. FEA simulations rapidly produced reliable and meaningful data.

To learn more about MSC.Marc, go to RequestLink at www.devicelink.com/emdm.

Copyright ©2003 European Medical Device Manufacturer