* Corresponding author. Tel.: #1-315-464-5226; fax: #1-315-464- 6470. Journal of Biomechanics 32 (1999) 1251}1254 Brief communication Mechanical strength of the cement}bone interface is greater in shear than in tension K.A. Mann*, F.W. Werner, D.C. Ayers Department of Orthopaedic Surgery, SUNY Health Science Center at Syracuse, 750 East Adams Street, Syracuse, NY 13210, USA Received 20 August 1998; accepted 4 May 1999 Abstract The objective of this study was to determine the relative mechanical properties of the cement}bone interface due to tensile or shear loading. Mechanical tests were performed on cement}bone specimens in tensile (n"51) or shear (n"55) test jigs under the displacement control at 1 mm/min until complete failure. Before testing, the quantity of bone interdigitated with the cement was determined and served as a covariate in the study. The apparent strength of the cement}bone interface was signi"cantly higher (p(0.0001) for the interface when loaded in shear (2.25 MPa) when compared to tensile loading (1.35 MPa). Signi"cantly higher energies to failure (p(0.0001) and displacement before failure (p(0.01) were also determined for the shear specimens. The post-yield softening response was not di!erent for the two test directions. The data obtained herein suggests that cement}bone interfaces with equal amounts of tensile and shear stress would be more likely to fail under tensile loading.  1999 Elsevier Science Ltd. All rights reserved. Keywords: Total joint arthroplasty; Polymethylmethacylate (PMMA); Interface; Mechanical testing; Fracture mechanics 1. Introduction Excessive shear stresses at the cement}bone interface have been cited as a mechanism by which the cement column subsides within the proximal femur (Ling, 1992; Miles et al., 1990). However, "nite element analysis of cemented femoral hip components has shown that sub- stantial tensile stresses can develop across the ce- ment}bone interface (Chang et al., 1998). This suggests that it could be possible to initiate failure of the interface with tensile loading. To determine the mechanism of interfacial failure, be it due to tensile, shear, or a combi- nation of tensile or shear loading requires determination of the mechanical properties of the cement}bone inter- face under both tensile and shear loading. A direct comparison between the tensile and shear strength properties of the cement}bone interface is di$- cult due to confounding variables including the amount of trabecular bone left in the femoral canal after broach- ing, quality of the bone in the canal and in"ltration of cement into the trabecular spaces. An approach has recently been developed to quantify these variables and includes the use of quantitative computed tomo- graphy of the bone and direct measurement of the amount of interdigitation between the cement and bone (Mann et al., 1997a). These quantities can then be used to develop a relationship between the strength of the interface and the amount of bone interdigitated with cement. The purpose of this study was to determine if the cement}bone interface would be more likely to fail under tensile or shear loading. Mechanical test results from a previously reported tensile test study (Mann et al., 1997a) are compared to new data from shear tests. 2. Materials and methods Cement}bone test specimens were fabricated from six fresh-frozen human proximal femurs (mean age 58, range 25}77) that had been broached, cleaned with a saline lavage, and cemented using contemporary cementing techniques. Institutional Review Board approval at the 0021-9290/99/$ - see front matter  1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 1 - 9 2 9 0 ( 9 9 ) 0 0 1 0 7 - 4