Component fatigue testing is performed before clinical use to validate the safety of the total joint replacements against fatigue failure. Fatigue test prediction can aid the design of fatigue resistant components and is required in the planning of an efficient fatigue-testing program.

The objective of this study was to evaluate fatigue test prediction methods based on finite element analysis (FEA) for the standard fatigue testing of total hip replacement components. The term fatigue test prediction includes life prediction at a given load level and strength prediction at a given life.

In order to simplify this initial study the ISO standard hip stem fatigue test without torsion, which defines constant amplitude loading, and a forged, titanium alloy hip stem with plain geometry were chosen for the evaluation. The validation of the fatigue test predictions was supported by static testing of strain-gauged hip stems and an extensive component fatigue testing program. Through effective planning and analysis, the component fatigue test results could be described with 3-parameter Weibull distributions of life at two stress levels, and log-Normal distributions of fatigue strength at various lives.

An accurate stress-force relationship was found with large deflection FEA including stem-cement debonding. The classical stress-life and strain-life methods were investigated, including Neuber's rule to correct FEA stress to the cyclic stress-strain curve and the Morrow, Smith-Watson-Topper and Goodman mean stress correction methods. Strain-controlled material fatigue tests provided the required cyclic material properties. Estimates of the crack propagation life found that crack propagation was a small percentage of the total life. The component fatigue strength at 5 million cycles could be predicted with an error from 1% to 9%. The life predictions using the strain-life method ranged between a non-conservative factor of 9 to a conservative factor of 7 from the median fatigue life determined from the component tests. It was concluded that the strain-life method was successful for the transition and infinite life regions;​ whereas the classical stress-life method was only appropriate near the endurance limit.

Through the combination of modelling and testing methods the development engineer can deliver high quality and competitive engineering components that fulfil fatigue strength requirements. This was demonstrated using hip stems.