Input variables

Impactor tips with five different stiffnesses (S_tip ) were tested (tip 1 to tip 5, Fig. 3. See supplementary material for measurement of tip stiffness), with three different base plate stiffnesses S_base (‘Soft’: 0.5 N/m, ‘Stiff’: 5.0 N/m, ‘Rigid’: >10⁵ N/m) for two different impaction energies (E_High=2.2 J, E_Low=1.1 J). 1.1 J is prescribed for similar testing of modular components by the ASTM standard (ASTM F2009, 2011).^17,18 It represents a typical head-stem assembly using a hammer with a mass of 0.5 kg dropped from a height of 20 cm. Each of the 30 possible combinations was tested, and repeated three times (n = 3). A total of 20 heads and 40 tapers were available and were re-used a maximum of two times (no effect of repeated use was measurable in preliminary tests).

Fig. 3

The tips were screwed to the end of the impactor and rested on the bearing surface of the head. The tip stiffness was determined by measuring the force-displacement characteristic between the top of the impactor tips and the bottom of the head, and estimated by dividing the peak force by the peak displacement. The peak force considered was that applied for the 2.2 J energy level for each tip (see supplementary material for method of determining tip stiffness). Mean and standard deviation (sd) are given for three measurements of each tip. PE, Polyethylene; POM, Polyoxymethylene.

Output variables

The impactor force impulse was measured just above the tip of the impactor using an impulse sensor (capacity 22 kN, Model 208C05; PCB Piezotronics Inc., Depew, New York, sampling rate 500 kHz).

Parameters employed in the analysis were the peak assembly force F_assembly (measured with the impulse sensor in the impactor) and the time period of the assembly force T_assembly, defined between 10% of F_assembly and F_assembly (Fig. 4).

Fig. 4

Representation of the evaluation scheme for analysis of the peak force (F_assembly or F_base for the applied impulse and base plate force, respectively) and the time period (T_assembly or T_base for the applied impulse and base plate force, respectively) of the dynamic impaction assembly process under all test conditions.

The pull-off force F_off of the head from the taper was measured to assess the taper fixation strength of the modular connection. F_off was measured quasi-statically at 0.008 mm/s (according to ISO 7206-10 and ASTM F2009, 2011)^17,18 using a uni-axial testing machine (capacity 10 kN, Zwick/Roell Z010; Zwick GmbH & Co. KG, Ulm, Germany) with a custom grip^9-11 (according to ISO 7206-10 and ASTM F2009, 2011)^17,18 (Fig. 2).

The force in the springs F_base, and the time period T_base of their deflection, were derived from the measurement of their deflection using a laser sensor (range 10 mm, accuracy 0.005 mm, Model PT660021; ipf electronic gmbh, Lüdenscheid, Germany). F_base and T_base were calculated from the resulting force-time data according to Figure 4. For the ‘rigid’ case, the springs were replaced by metal tubes with very high stiffness (estimated analytically to be 10⁵ N/m). In this case, no displacement could be measured so an impulse sensor below the taper (PCB Piezotronic Model 208C05) was used to measure the force F_base. This sensor remained in the system for all experiments, also when not in use, during measurements using springs.

Applied peak force F_assembly, base plate force F_base, pull-off force F_off, applied time period T_assembly and base plate time period T_base were compared statistically between impactor tip stiffness S_tip, base stiffness S_base and impaction energy E, using a parametric analysis (two-way analysis of variance (ANOVA)) and post hoc test (Tukey), or a non-parametric analysis (Kruskal-Wallis) with the probability of a Type I error set to α = 0.05 (SPSS Statistics 22; IBM Corp., Armonk, New York). All results are reported as mean and sd.

Applied force versus tip stiffness and energy

The peak force magnitude of the applied impulse F_assembly increased with tip stiffness S_tip (p < 0.001), but with decreasing gains (Fig. 5a). The applied time period T_assembly decreased with increasing tip stiffness S_tip (p < 0.001) (Fig. 5b). The applied force F_assembly was found to increase with the magnitude of the applied energy E (p < 0.001), but not to be statistically dependent on the stiffness of the base S_base (p = 0.864) (Fig. 5a). Conversely, the time period of the applied force T_assembly was not found to depend on the applied energy E (p = 0.145), it decreased with increasing stiffness of the base S_base (p < 0.001) and was not dependent on the impactor stiffness S_tip (p = 0.225) (Fig. 5b). Since the applied force F_assembly increased with increasing tip stiffness S_tip (Fig. 5a) as the time period T_assembly decreased (Fig. 5b), an increasing loading rate with increasing tip stiffness S_tip is indicated.

Fig.

All plots are mean/sd plotted against the same impactor tip stiffness S_tip scale on the horizontal axis and for three base plate stiffnesses S_base (blue: ‘soft’ 0.5 N/m; red: ‘stiff’ 5.0 N/m; green: ‘rigid’ > 10⁵ N/m) representing the mechanical compliance of the patient and two energy levels (open points: 1.1 J; solid points: 2.2 J). Tips 3 to 5 are labelled. Tip 4 is a commercially available plastic tip, and tip 5 is metal and was the stiffest tested. Note that the sd bars that are not visible are too small to be observed. a) Peak force F_assembly of the impulse applied to the head. b) Time period T_assembly of the impulse applied to the head, defined in Figure 3. c) Pull-off force F_off of the head from the taper. Exponential functions were fitted to each curve, according to Table II. d) Peak force F_base generated in the base plate springs. e) Time period T_base of the force generated in the base plate springs defined in Figure 4.

The base plate time period T_base decreased with the stiffness of the base S_base (p < 0.001), increased with applied energy E (p < 0.001), and was not statistically dependent on the impactor stiffness S_tip (p = 0.988) (Fig. 5e).

Research questions

Increasing the stiffness of the impactor tip was found to increase the taper fixation strength of a cobalt chrome head on a titanium stem taper. Furthermore, for a sprung base representing the mechanical compliance of the patient, an increasing tip stiffness did not increase forces below the taper for a given applied impaction energy. Increasing the impaction energy increased the fixation strength but also increased the force below the taper which represents greater load transfer through the tissues. This suggests that taper fixation strength should be maximized by increasing the stiffness of the impactor tip, rather than the applied energy, which could damage the patient’s tissues. However, the strategy of maximizing tip stiffness is limited by potentially damaging the head during impaction.

The impactor shaft is made of metal and is stiff relative to the standard plastic impactor tip. The impactor tip is sufficiently soft to protect the bearing surface of the head. It has a concave surface to locate on the surface of the femoral head. The tip could potentially be made stiffer, by using a material with a higher elastic modulus, by decreasing its thickness or by conforming its geometry to the spherical surface of the head. The maximum practical elastic modulus for the impactor tip would be that for a ceramic-type material, which is roughly double that of steel or cobalt chrome, and roughly two orders of magnitude greater than that for the plastic tips used in current designs. The thickness could be reduced, in principal, to a very thin layer. The greater the conformity of the tip to the head (conical or spherical), the greater its stiffness will be since Hertzian contact deformation is reduced. These strategies could be used to maximize impactor stiffness by being incorporated in a tool providing a pre-defined impulse energy, to prevent damage to the head.

How much strength can be gained by increasing tip stiffness?

In previous studies, the fixation strength of the head-neck junction has been related to the peak force with apparently linear relationships, both under quasi-static and dynamically applied assembly forces using a rigid base plate.^9-11,19 In the current study, the peak force was applied over a greater range, revealing a steep increase in pull-off force for low tip stiffness, becoming flatter at higher tip stiffnesses (Fig. 5c). An exponential function relating pull-off force to tip stiffness revealed a good fit (R² > 0.94) and suggested decreasing gains in strength with increasing tip stiffness until a maximum is achieved (given by constant a₃ in Table I). The metal tip (tip 5) generated pull-off forces that were within 10% of the projected maximum and the plastic POM tip (tip 4), used in commercially available designs, generated pull-off forces of within 30% of the projected maximum. Thus, current systems employing hard plastic tips may already be close to generating the maximum possible seating for a given impaction energy.

Why does assembly force increase with tip stiffness but not with base stiffness?

The applied energy is provided by the gravitational potential of the hammer which is directly proportional to its mass and height above the impactor and is converted to the kinetic energy of the impaction. The hammer strikes the metal impactor shaft, which accelerates the impactor tip against the head. The softer the tip, the longer the time period over which the impactor shaft will accelerate. Assuming energy losses are low, the contact force F_assembly between impactor tip and head will decrease as the period T_assembly increases (Figs 5a and b, respectively), since momentum is conserved and the area below the force-time impulse function remains constant. The impaction force F_assembly was unaffected by the sprung base stiffness S_base because the head-neck taper is very stiff and seating is completed more rapidly (within a period of the same order of magnitude as the applied impulse T_assembly) than the deflection of the much softer, and more massive sprung base system (T_assembly is an order of magnitude lower than T_base. Compare Fig. 5b with Fig. 5e).

Standard Testing

It is noted that the ASTM testing procedure for modular components specifies an assembly energy but does not address the force-time characteristic of the applied impulse. This study clearly demonstrates that very different seating behaviour can be achieved for a given applied energy (E_low = 1.1 J in the current study is specified by the ASTM standard)^17,18 due to the dependency of the applied force on the stiffness of the impactor. However, although it is recommended in the ASTM standard that the apparatus be placed on a soft base, it is also noted here that seating is independent of the stiffness of the base plate, so that simple rigid fixation of the modular components should be adequate for studies of taper strength.

Why does force transmitted to surrounding tissues increase with energy but not with tip stiffness?

Assembly forces are modulated by the tip stiffness due to their effects on the acceleration of the impactor. However, assuming small losses due to friction between head and stem tapers, for example, the assembly energy must be transferred through the modular connection to the springs, which deform according to the magnitude of the energy applied. The taper connection seats rapidly due to its high stiffness, regardless of the applied assembly force (tip stiffness). In contrast to the experimental construction with a sprung base described above, the force transmitted to the rigid base plate increased with tip stiffness (Fig. 5d) because the rate of loading is more similar to the fundamental frequency of the base, thereby providing dynamic coupling.

How does the base plate stiffness relate to the mechanical compliance of patients?

The rigid base plate construction (with the springs exchanged for metal tubes) can be assumed to represent an unrealistically high reference stiffness in which it was shown that the force transferred to the base plate (Fig. 5d) was slightly greater than the applied peak force (Fig. 5a) for the two stiffest tips (< 130%). However, although a representative stiffness and mass for a patient on the operating table are not known, it can be assumed that the rigid condition does not exist and that the force transmitted to the surrounding tissues (Fig. 5d) is lower than the applied assembly force (Fig. 5a), as observed for the sprung experiments. Purely subjective observation of the intraoperative situation by the authors suggests that a patient’s leg, which has been incised at the hip and has had the femoral head removed in preparation for implantation, is at least as soft in the direction of head impaction as the stiffer springs tested experimentally. This would lead to tissue forces that are at least an order of magnitude lower than the applied force (Fig. 5d). According to the data presented, the compliance of the patient may indeed help to maintain safe stress magnitudes in the soft tissues. It is noted that although there are reports of periprosthetic fracture due to cavity preparation or stem impaction into the bone,^13-16 there are no reports of intraoperative tissue damage specifically due to impaction of the head onto the stem taper. However, if safe seating of the head is to be ensured, applied energy levels may need to be increased.

The mechanical compliance of the patient on the operating table was modelled using springs below the base plate. This is a simple but defined approach in which the mass of the sprung system, as well as the spring constant, are known. The input and output parameters in the current study can therefore be used to validate mechanical models, leading to a clearer understanding of the load transfer mechanism. The assembly load is transferred from the head to the stem, then to the femur, then to the soft tissues, and finally to the operating table and to ground. Three-dimensional geometry, non-linear time-dependent tissue properties and complex boundary conditions will certainly lead to a force-displacement response that is not represented by the mass on elastic springs employed in the current study. However, it seems fair to assume that the acceleration of the impactor mass combined with the deformation of the tissues will be high enough to allow a few millimetres of motion of the tissues under the applied load. This is provided experimentally by the springs, in contrast to the less realistic rigid condition. The two spring stiffnesses employed in the current study merely provided a parametric variation, demonstrating that less force is transferred though the tissues of an effectively softer patient (Fig. 5d) (albeit for a longer period, Fig. 5e). Whether the stiffness and mass parameters relate to larger or smaller patients, different surgical approaches, or are modified by the position of the patient on the table, is unclear. Complete modelling of the patient would probably necessitate 3D geometry and dynamic stress-strain properties, as well as mass distributions of the tissues. Damage to tissues may be dependent not only on stress magnitudes, but also on their duration.^13-16

In other studies, the mechanical compliance of the patient has been represented by a foam pad below the base plate but its influence on the taper fixation strength was not discussed.^20,21

The clinical situation

The current experiment was conducted using a standard cobalt chrome head on a replica titanium stem taper. Similar combinations have recently been reported to fail due to corrosion of the cobalt chrome component, leading to metal ion release and biological reaction or fracture of the neck.^7,22-24 Failure of the titanium taper has also been noted due to massive wear.¹ Both processes are related to abrasion of the metal surfaces, which occurs due to relative motion between the components.⁷ Increasing the initial press fit by increasing the assembly force has been shown to decrease corrosion.^12,22 Increasing the impactor stiffness and the applied energy will increase fixation strength and may thereby decrease corrosion and wear. These trends should hold for all metal-on-metal head-neck modular taper junctions but critical assembly force magnitudes for each design remain to be determined.

Measurements of tissue loading during component assembly would be necessary, in combination with dynamic models of the patient, in order to estimate energy levels that maintain tissue stresses within safe magnitudes. Note that the energy applied is theoretically proportional to the mass of the hammer and the height from which it is dropped. Doubling the applied energy doubled the deflection of the base plate springs and, therefore, the force acting in them. This finding should be interpreted as a warning to the surgeon.

Other limitations

It is noted that the force below the taper was derived differently for the sprung and rigid bases. For the sprung base, the force was calculated from the measured deflection and the spring constant. This is the force acting in the spring, representing the compliance of the patient. However, for the rigid system, the springs were replaced by stiff metal tubes and no displacement could be measured. The force was therefore measured directly below the taper by an impulse load cell and the effective mass and stiffness of the base were assumed to tend to infinity, providing an extreme reference.

In conclusion, this study demonstrates that increasing the tip stiffness of the impactor increases the fixation strength of the head on a neck taper and that this does not influence the forces transmitted through the patient’s tissues below the modular components. The impactor stiffness should therefore be as high as possible without damaging the surface of the head but this limit would need to be determined experimentally and implemented by the manufacturers. On the other hand, it seems that tip designs currently used may generate interface strengths that are within 30% of an apparent maximum possible for a given applied energy. To increase the strength further, the applied energy can be increased, but this will increase forces transmitted through the patient’s tissues. The surgeon is responsible for energy application, which is theoretically proportional to the mass of the hammer (and the surgeon’s arm), and to the square of the impaction velocity (or to the drop height). The assembly forces required to eliminate dangerous levels of corrosion are yet to be determined. They will be necessary, along with the determination of maximum safe levels of applied energy, to prevent corrosion-related pathology, wear and dislocation, patient tissue damage and damage to the bearing surface, for a particular implant system and given patient. An impaction device could be employed that delivers a predefined impulse, providing a sufficient assembly force, without damaging the head or the patient.