The influence of friction and interference on the seating of a hemispherical press-fit cup: a finite element investigation

The formation of gaps in the polar region of acetabular cups is seen as a drawback of press-fit fixation of non-cemented acetabular cups. Recent findings indicate a link between long-term polar gaps and the gaps present directly after implantation. In this study the process of Dress-fitting is simulated with a linear-elastic two-dimensional axisymmetric finite-element model. The aim of this paper is to investigate the possible importance of friction and interference on the formation of these gaps. A range of cup-bone friction coefficients (mu = 0.1-0.5) is assigned to the cup-bone interface in order to represent the unknown amount of friction occurring during press-fitting. The cup is modeled with a radius of 27 mm, whereas the radius of the cavity is varied between 26.50 and 26.75 mm, thus, creating 0.50 and 0.25 mm radial interference fits. The difference in cavity radius represents the discrepancy between the radius of the last-reamer-used and radius of the cavity it creates. The subchondral plate is considered as being completely removed during reaming. The effects of impact blows via the surgeon's mallet during surgery are modeled as a series of four load pulses, in which peak force is gradually increased from 0.5 to 4.0 kN. The effects of load removal as well as those of load application are investigated. On load application, the cup penetrates into the cavity, and on load removal, the cup rebounds. Depending on the friction, interference and load applied, the position of the cup after the load pulse is somewhere between its position at peak force and its position at the beginning of the pulse. Although the simplifications and conditions involved in the creation of the model necessitate caution when interpreting the results for all clinical cases, it is found that the seating of hemispherical cups in trabecular bone could be more satisfactory for intermediate values of friction (mu = 0.2-0.3) and smaller interference fits (0.25 mm). (C) 1999 Elsevier Science Ltd. All rights reserved.