* Corresponding address. Laboratorio di Tecnologia Medica, Istituti OrtopediciRizzoli,Via di Barbiano 1/10, 40136 Bologna, Italy.Tel.: #39-051-6366864; fax:#39-051-6366863. E-mail address: cristofolini@tecno.ior.it (L. Cristofolini) Journal of Biomechanics 33 (2000) 279}288 Mechanical validation of whole bone composite tibia models Luca Cristofolini !,*, Marco Viceconti " ! Engineering Faculty, University of Bologna, Bologna, Italy " Laboratory for Biomaterials Technology, Rizzoli Orthopaedic Institutes, Bologna, Italy Received 29 September 1998; accepted 13 October 1999 Abstract Composite synthetic models of the human tibia have recently become commercially available as substitutes for cadaveric specimens. Their use is justi"ed by the advantages they o!er as a substitute for real tibias. The present investigation conc an extensive experimental validation of the mechanical behaviour of the whole bone composite model, compared to hum for di!erent loading conditions. The sti!ness of the tibias was measured with a torsional load applied along the long axis, a a bending load applied both in the latero-medial and in the antero-posterior direction. The bending sti!ness of the compos matched well with that of the cadaveric specimens. This was not true for the torsional sti!ness. In fact, the composite tibia sti!er than the cadaveric specimens, possibly due to the structure of the reinforcement material. The inter-specimen varia composite tibias was much lower than that for the cadaveric specimens. Thus, it seems that the composite tibias are suita cadaveric specimens for certain types of test, whereas they might be unsuitable for others, depending on the loading reg ( 2000 Elsevier Science Ltd. All rights reserved. Keywords: Tibia;Composite; Model;Material testing; Validation 1. Introduction Cadaveric bone segments are often used to test in vitro prosthetic components and "xation devices. Tibial im- plants such as intramedullary nails, "xation plates and knee prostheses need to be evaluated in vitro before they are implanted into patients. In most cases, comparative tests are performed, e.g.to compare the stress shielding e!ect or the primary stability ofan implant,or the sti!nessof an intramedullary nail. The variability of cadaveric specimens has always been a problem, requir- ing enormous sample sizes to obtain a satisfactory signi"- cance and power of statistical comparisons. The inter- specimen variability for cadaveric specimens can be esti- mated from the literature and from the data presented here to reach 100% of the mean (see for instance Henley et al.,1993;Crowninshield et al., 1990). For example, if a di!erence of 10% ofthe mean must be detected with a con"dence of 95%, a sample of several hundred speci- mens is required. Moreover, the use of human cadaveric specimenscausesproblemsin availability,handling, preparation, and preservation. For the reasons mentioned above, attention has been given to possible substitutes of human specimens. Syn- thetic models o!er the advantage of being more avail- able, and at the same time of allowing a standardisation of the tests (Viceconti et al., 1996). A larger research for synthetic substitutes exists for the femur, starting from the polyurethane models used by Uta (1992)to test fracture "xation devices. To account for the anisotropy and inhomogeneity of the bone properties, more accurate home-madecompositemodelshave been proposed (Ypma et al., 1982;McKellop et al., 1991). To the authors' knowledge, so far only four works have been published in which synthetic models of the tibia were used. First,Little and O'Keefe (1989) used a large- scale model of the tibia,made oflow-density polyethy- lene to represent the spongy bone (holes were drilled to simulate the porosity), and "lled epoxy to simulate the corticalbone.The modelwas used to investigate the stress distribution in the cement underlying a prosthetic tibial component. Kanakis and Cordey (1991)used plastic models of the tibia to test di!erent means of "xatio of comminuted fractures. Firoozhakhsh et al. (1996) used 0021-9290/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 1 - 9 2 9 0 ( 9 9 ) 0 0 1 8 6 - 4