Research paper
Bone–cement interfacial behaviour under mixed mode loading conditions

https://doi.org/10.1016/j.jmbbm.2010.03.001Get rights and content

Abstract

Interfacial behaviour of the bone–cement interface has been studied under tensile, shear and mixed mode loading conditions. Bovine cancellous bone was used to bond with acrylic bone cement to form bone–cement interface samples, which were mechanically tested under selected tensile, shear and mixed mode loading conditions. The influence of the loading angle and the extent of the cement penetration on the interfacial behaviour were examined. The failure mechanisms with regard to loading mode were examined using micro-focus computed tomography. The measured tensile and shear responses were utilized in a cohesive zone constitutive model, from which the pre-yield linear and the post-yield exponential strain softening behaviour under mixed mode loading conditions was predicted. The implications of the work on the studies of cemented joint replacements are also discussed.

Introduction

The success of cemented total hip replacements depends critically on the lasting integrity of the bond between the bone and the cement (Stocks et al., 1995, Thanner et al., 1999). Retrieval studies reveal that, although microcracks are observed in the cement mantle, it is the debonding between the cement and the bone that often defines the final failure of cemented hip replacements, particularly on the acetabular side. This may be illustrated at the revision surgeries by the easy removal of the acetabular cups with cement mostly attached to the cup. Recent in vitro experiments on cemented bovine acetabular implants under constant amplitude peak hip contact force (Heaton-Adegbile et al., 2006, Zant et al., 2007) and physiological loading conditions using a hip simulator (Zant et al., 2008, Wang et al., 2009) also support this argument, where failure at the bone–cement interface has been identified as the predominant failure mode in cemented acetabular replacements.

Although numerous studies have been carried out on bone–cement interfacial strength, using experimental (Krause et al., 1982), analytical (Clech et al., 1985) or numerical (Tong et al., 2007) methods, the most complete studies have been carried out by Mann and his associates (Janssen et al., 2008, Mann et al., 1997a, Mann et al., 1997b, Mann et al., 1999, Mann et al., 2001, Kim et al., 2004, Mann et al., 2009). Their work includes mechanical testing of cadaveric bone–cement interface samples under tensile (Mann et al., 1997a), shear (Mann et al., 1999), mixed mode (Mann et al., 2001), creep-fatigue (Kim et al., 2004) loading conditions, constitutive modeling (Mann et al., 1997b), and more recently, micromechanics studies of the interface (Janssen et al., 2008, Mann et al., 2009). Much valuable information has been generated from these well-designed studies, although some of the fundamental aspects of the topic remain to be further explored. For example, what is the role of cement interdigitation in defining the bone–cement interfacial strength? Published works do not appear to suggest a significant correlation between the bone–cement interfacial strength and the depth of cement penetration (R2=0.22 (Krause et al., 1982)), although the methods of quantification of cement penetration tend to vary. The quantity of bone interdigitated with the cement has been used to improve the correlation (Mann et al., 1997a, Mann et al., 1999, Mann et al., 2001, Kim et al., 2004), albeit with only moderate success (R20.5 for tensile; R20.50.7 for mixed mode cases). The main failure mechanisms were identified (Mann et al., 1997a, Mann et al., 1999), although not correlated with the loading angle. In general, there is a large variation in the results between specimens tested under similar loading conditions, possibly due to the variation in cadaver bone tissues, and the difference in the amount of cement infiltration.

Modeling of the bone–cement interface has been attempted using interfacial fracture toughness (Tong et al., 2007), although the measured interfacial fracture toughness was found to be crack length dependent, indicating the necessity of using a nonlinear fracture mechanics approach. The cement–bone interface is capable of carrying substantial loads after the peak load is reached (Mann et al., 1997a) due to the interdigitation between the cement and the porous cancellous bone. Such behaviour may be described using cohesive zone models (Clech et al., 1985) so that post-yield softening as well as pre-yield behaviour may be characterized (Mann et al., 1997b). More recently, Moreo et al., 2006, Moreo et al., 2007 extended the analysis of Mann et al. (1997b) to simulate fatigue failure of bone–cement interfaces using a coupled viscoplastic damage model.

In this paper, bone–cement interfacial behaviour under remote mixed mode loading conditions has been studied using both experimental and numerical approaches. Bovine cancellous bones were used to bond with acrylic bone cement and mechanically tested under mixed mode conditions. The failed samples were examined using micro-focus computed tomography to identify the failure mechanisms associated with the loading angle. A finite element model of the interface sample was built and cohesive zone analyses were carried out, utilizing the characteristic parameters including the initial stiffness, fracture energy and apparent strength as obtained from the experimental traction–separation curves in tension and shear, and the model was subsequently used to predict the traction–separation behaviour under mixed mode loading conditions. The main motivation of the work was to explore the relationship between the measurable parameters, such as the fracture strength and the failure mode, and the operational parameters, such as loading mode and cement penetration, so that the key parameters that dictate the interfacial behaviour may be identified. The effectiveness of cohesive modeling as a numerical tool was also examined and its predictive capacity was evaluated with regard to the mixed mode loading cases. It is hoped that such a tool might be useful eventually for predictive purposes of cement fixation in joint replacements, when the bio-mechanical characteristics of the interface become better understood.

Section snippets

Materials and specimens

Bovine cancellous bone coupons were harvested from fresh bovine pelvis obtained from a local abattoir. Samples were carefully cleaned to remove fatty soft tissues and selected to ensure relative uniform cancellous bone morphology. The cross-section area of the specimens was 5 mm by 10 mm, the same as that used in Mann et al. (1997a). The bone coupons were then placed in a mold, to which mixed bone cement (CMW1, DePuy CMW) was applied. A constant pressure was maintained during curing to ensure

Cohesive zone model formulation

The cohesive zone model adopted here follows the classical continuum damage mechanics theories (Dugdale, 1960; Lemaitre, 1985) where the mechanical response of bone–cement interface may be described using a force–displacement relationship coupled with damage formulation. The bone–cement interfacial behaviour may be described by an initial elastic behaviour followed by the initiation and the evolution of damage, as the crack propagates along the interface. The elastic behaviour may be defined by

Concluding remarks

The results obtained from this work may be useful in the finite element analysis of bone–cement interface in joint replacements with or without defects. The fracture process may be modeled using the cohesive zone model based on the constitutive behaviour of the bone–cement interface determined experimentally. A full model of a cemented acetabular reconstruction may be analysed to assess the initiation and propagation of defects at the bone–cement interface. In vitro mechanical testing may then

Acknowledgements

The authors gratefully acknowledge the contribution of Mr. T. Bender in the experimental work; Dr. J.-G. Hussell and Dr. P. Heaton-Adegbile for helpful discussion with regard to clinical applications. The bone cement was donated by DePuy CMW, UK.

References (29)

  • J.Y. Wang et al.

    Damage evolution in acetabular replacements under long-term physiological loading conditions

    J. Biomech.

    (2009)
  • N.P. Zant et al.

    Fatigue failure in the cement mantle of a simplified acetabular replacement model

    Int. J. Fatigue

    (2007)
  • ABAQUS, Inc. 2008. ABAQUS Analysis User’s Manual Version...
  • D.J. Bean et al.

    Sustained pressurization of polymethylmethacrylate: a comparison of low- and moderate-viscosity bone cements

    J. Orthop. Res.

    (1988)
  • Cited by (0)

    View full text