Fatigue Assessment for Novel Hip Replacement Designs Using Aerospace Analysis Techniques

Abstract

Representative pre-clinical analysis is essential to ensure that novel prosthesis concepts offer an improvement over the state-of-the-art. Proposed designs must, fundamentally, be assessed against cyclic loads representing common daily activities [Bergmann 2001] to ensure that they will withstand conceivable in-vivo loading conditions. Fatigue assessment involves:

1. –

   cyclic mechanical testing, representing worst-case peak loads encountered in-vivo, typically for 10 million cycles, or

2. –

   prediction of peak fatigue stresses using Finite Element (FE) methods, and comparison with the material's endurance limit.

Cyclic stresses from gait loading are super-imposed upon residual assembly stresses. In thick walled devices, the residual component is small in comparison to the cyclic component, but in thin section, bone preserving devices, residual assembly stresses may be a multiple of the cyclic stresses, so a different approach to fatigue assessment is required.

Modular devices provide intraoperative flexibility with minimal inventories. Components are assembled in surgery with taper interfaces, but resulting residual stresses are variable due to differing assembly forces and potential misalignment or interface contamination. Incorrect assembly can lead to incomplete seating and dissociation [Langdown 2007], or fracture due to excessive press-fit stress or point loading [Hamilton 2010]. Pre-assembly in clean conditions, with reproducible force and alignment, gives close control of assembly stresses. Clinical results indicate that this is only a concern with thick sectioned devices in a small percentage of cases [Hamilton 2010], but it may be critical for thin walled devices.

A pre-clinical analysis method is proposed for this new scenario, with a case study example: a thin modular cup featuring a ceramic bearing insert and a Ti-6Al-4V shell (Fig. 1). The design was assessed using FE predictions, and manufacturing variability from tolerances, surface finish effects and residual stresses was assessed, in addition to loading variability, to ensure physical testing is performed at worst case:

1. –

   assembly loads were applied, predicting assembly residual stress, verified by strain gauging, and a range of service loads were superimposed.

The predicted worst-case stress conditions were analysed against three ‘constant life’ limits [Gerber, 1874, Goodman 1899, Soderberg 1930], a common aerospace approach, giving predicted safety factors. Finally, equivalent fatigue tests were conducted on ten prototype implants.

Taking a worst-case size (thinnest-walled 48 mm inner/58 mm outer), under assembly loading the peak tensile stress in the titanium shell was 274 MPa (Fig. 2). With 5kN superimposed jogging loading, at an extreme 75° inclination, 29 MPa additional tensile stress was predicted. This gave mean fatigue stress of 288.5 MPa and stress amplitude of 14.5 MPa (R=0.9). Against the most conservative infinite life limit (Soderberg), the predicted safety factor was 2.40 for machined material, and 2.03 for forged material, or if a stress-concentrating surface scratch occurs during manufacturing or implantation (Fig. 3). All cups survived 10,000,000 fatigue cycles.

This study employed computational modelling and physical testing to verify the strength of a joint prosthesis concept, under worst case static and fatigue loading conditions. The analysis technique represents an improvement in the state of the art where testing standards refer to conventional prostheses; similar methods could be applied to a wide range of novel prosthesis designs.