Journal of Orthopaedic Research 9798-808 Raven Press, Ltd., New York 0 1991 Orthopaedic Research Society

Mechanical Characteristics of the Stem-Cement Interface Kenneth A. Mann, Donald L. Bartel, Timothy M. Wright, and Anthony R. Ingraffea Cornell-Hospital f o r Special Surgery Program in Biomechanical Engineering, Cornell University, Zthaca, New York, U.S.A.

Summary: The mechanical characteristics of the interface between a metallic stem and the surrounding poly(methy1 methacrylate) bone cement were determined from experimental tests and finite element analyses. Push-through-stem tests of straight and tapered titanium alloy stems, surrounded by cement columns, were performed and the resulting load-displacement behavior and strain distribution on the surface of the cement column were measured for loading, unloading, and reloading. Test geometries were modelled using nonlinear, axisymmetric, finite element analyses, which incorporated Coulomb friction elements at the titanium alloy-cement interface. Initial residual stresses, due to curing of the cement column, were modeled by thermal contraction of the cement. Good agreement was obtained between load-displacement curves and surface strains predicted from the nonlinear analysis and those obtained from experiments, when a coefficient of friction of 0.3 was assumed for the stemcement interface. These results show that, in the absence of chemical adhesion, the load-displacement behavior of a stem-cement composite can be described completely in terms of the friction at the interface and the residual stresses normal to the interface. Key Words: Total joint replacement-Finite element analysis-Poly(methy1 methacrylate)-Implant-cement interface.

limited to experimental studies of interface strength and to numerical investigations of the role of a degraded stem-cement interface on behavior of the bone-cement-implant system. Studies of interface strength have relied on experimental tests, which compare the ultimate strength of metal-cement preparations, using various chemical and mechaniResults cal alterations (3,5,6,7,14,15,17,19,20). from these investigations provide a method for comparing relative strengths of metal-cement interfaces, but are not descriptive of the mechanical behavior of the interface in terms of load-displacement characteristics of the test specimens. Other investigators (1 1,13) have used stress analysis to describe the structural behavior of the bone-implant system after interface conditions have been altered. Although these models are helpful in describing changes in load transfer and stress fields when loosening has occurred, they do not describe the rnech-

Loosening of total joint replacements continues to be a major mode of failure in cemented total joint arthroplasty (18,21,22). Loss of fixation at the stemcement interface may contribute to the loosening of the implant (10). The role of this interface in loosening remains unclear due to lack of information about the mechanical behavior of the interface and to limited ability to correlate clinical loosening with radiographic findings at the stem-cement interface. To understand the possible role of the stem-cement interface in loosening of the implant, mechanical characteristics of the interface must be determined. The contribution of the stern-cement interface to the integrity of the bone-implant system has been Received March 16, 1990; accepted November 13, 1990. Address correspondence and reprint requests to Dr. D. L. Bartel at Department of Mechanical and Aerospace Engineering, Upson Hall, Cornell University, Ithaca, New York 14853, U.S.A.

798

799

STEM-CEMENT INTERFACE MECHANICS

anisms responsible for initial stem-cement interface debonding. In preliminary studies in our laboratory (23), straight and tapered push-through-stem tests were performed under axial monotonic loading. Finite element models of test specimens were developed, which included interface elements at the stemcement interface. These elements were used to evaluate structural effects of a boundary layer, adhesion, strain softening, and Coulomb friction on overall load-displacement behavior of the stemcement structure. The single monotonic load step used in these preliminary tests did not provide enough information to determine which of these mechanical characteristics of the stem-cement interface controls overall load-displacement behavior. In the present study, we performed pushthrough-stem tests for the same specimen geometries used in the monotonic loading tests (23), but applied several loading-unloading cycles to the specimens. Finite element analysis of test specimens was used to determine mechanical characteristics of the interface, which produced the observed load-displacement behavior. This behavior was accurately modeled when friction at the interface and residual stresses, due to curing of the cement, were included.

MATERIALS AND METHODS Experimental Six tapered and nine straight push-through-stem test specimens were prepared for testing (Fig. 1). The diameters and tapers of the titanium alloy (Ti6A1-4V) stems are similar to those used in conventional femoral components. Titanium alloy stems were reused for each test and were washed in warm soapy water, rinsed in isopropyl alcohol, and air dried between specimen preparations. The poly(methyl methacrylate) cement (Zimmer, Warsaw, IN, U.S.A.) was prepared by mixing 120 gm prepolymerized powder with 60 cc methylmethacrylate liquid. To increase working time, the monomer and powder were mixed in a container surrounded by an ice-water bath. The cement was mixed for one minute before introduction into an aluminum mold. The stem was then inserted into the cement and centered distally by a guiding pin, and proximally by a polyethylene cap placed over the top of the stem. Specimens were allowed to cure for 24 h before testing. Displacement of the top of the stem, relative to the base of the specimen, was measured using three, direct current linear variable differential transformers (DCDT), located circumferentially around the specimen (Fig. 1). The global load-

Tapered Stem

Stralght Stem

I18 r

114r

DCDT 14

CDT 11-3 FIG. 1. Longitudinal cross-sections of axisymmetric straight and tapered push-through-stem test specimens. DCDTs Nos. 1,2, and 3 measure displacement of the top of the stem relative to the base of the cement column and DCDT No. 4 measures relative displacement between the top of the stem and the top of the cement column. Locations of strain gages applied to three of the straight-stem specimens are also shown. Dimensions are shown in millimeters.

737

c

Strain Gages

J Orthop Res, Vol. 9, No. 6 , 1991

800

K . A . MANN ET A L .

displacement curve was determined from the average of the three DCDT measurements. Use of the terms “global load-displacement’’ and “global stiffness” is common in finite element analysis (9) and describes the overall load-displacement behavior of the entire finite element structure. For the most part, individual DCDT measurements did not vary by more than three percent of the average value of the three DCDT measurements. DCDT signals were filtered using an analog two-pole, low pass filter, with a cut-off frequency of 5 Hz. DCDT signals required filtering before digital sampling, to remove 60 Hz noise from the DCDTs. A cut-off frequency of 5 Hz was sufficient to remove higher frequency noise, while still allowing any displacement transients to be measured. By itself, global load-displacement behavior was insufficient to describe the mechanism of load transfer between stem and cement column. To further clarify behavior at the stem-cement interface, we added a fourth DCDT to measure relative displacement between the top of the stem and the top of the cement column. These additional data, along with structural analyses, suggested that slip-at the stem-cement interface-began at the tip of the stem and progressed upward as the applied load was increased. To examine this hypothesis, six singleelement resistance strain gages (Fig. 1) were applied to the outer surface of the cement column, on three of the straight-stem specimens, after the cement had fully cured. The gages were applied to measure longitudinal strain during mechanical testing of the specimens and provided additional experimental data, which were compared with finite element results. Straight-stem specimens were used for this portion of the experiment since, based on structural analyses, anticipated strains were within the limits of the gages (

Mechanical characteristics of the stem-cement interface.

The mechanical characteristics of the interface between a metallic stem and the surrounding poly(methyl methacrylate) bone cement were determined from...
910KB Sizes 0 Downloads 0 Views