Optimisation of the mechanical and handling properties of an injectable calcium phosphate cement.

Calcium phosphate cements have the potential to be successful in minimally invasive surgical techniques, like that of vertebroplasty, due to their ability to be injected into a specific bone cavity. These bone cements set to produce a material similar to that of the natural mineral component in bone. Due to the ceramic nature of these materials they are highly brittle and it has been found that they are difficult to inject. This study was carried out to determine the factors that have the greatest effect on the mechanical and handling properties of an apatitic calcium phosphate cement with the use of a Design of Experiments (DoE) approach. The properties of the cement were predominantly influenced by the liquid:powder ratio and weight percent of di-sodium hydrogen phosphate within the liquid phase. An optimum cement composition was hypothesised and tested. The mechanical properties of the optimised cement were within the clinical range for vertebroplasty, however, the handling properties still require improvement.

In vitro simulation and quantification of wear within the patellofemoral joint replacement.

Quantification of the wear rate in vitro is now considered an essential step in the development of a new joint replacement prior to clinical trials. However, little research exists around in vitro simulation of wear in the patellofemoral joint (PFJ) despite over 200,000 being implanted annually within the European Union. A method to simulate wear in the laboratory using four input degrees of freedom within the PFJ of total knee replacement (TKR) has been developed. Wear simulation was validated through comparison of functional kinematics and patellar surface damage modes produced in vitro to clinical outcomes. The technique has been shown to replicate the prescribed in vivo kinematics in a reproducible and repeatable manner. The wear scar areas were similar to those found in vivo. However, geometrical measurements of wear were not reliable due to creep and geometry changes. As has been found previously with tibial inserts, geometrical determination of wear volume was not found to be an effective method of comparing wear from simulators and retrievals. Change in volume calculated gravimetrically was seen to be the most repeatable measure of patellar wear in vitro.

The biomechanical response of spinal cord tissue to uniaxial loading.

The spinal cord is an integral component of the spinal column and is prone to physical injury during trauma or more long-term pathological insults. The development of computational models to simulate the cord-column interaction during trauma is important in developing a proper understanding of the injury mechanism. Such models would be invaluable in seeking both preventive strategies that reduce the propensity for injury and identifying specific treatment regimes. However, these developments are hampered by the limited information available on the structural and mechanical properties of this soft tissue owing to the difficulty in handling this material in a cadaveric situation. The purpose of the present paper is to report the rapid deterioration in the quality of the tissues once excised, which provides a further challenge to the successful elucidation of the structural properties of the tissue. In particular, the tangent modulus of the tissue is seen to increase sharply over a period of 72 h.

Numerical model of self-propulsion in a fluid.

We provide initial evidence that a structure formed from an articulated series of linked elements, where each element has a given stiffness, damping and driving term with respect to its neighbours, may 'swim' through a fluid under certain conditions. We derive a Lagrangian for this system and, in particular, we note that we allow the leading edge to move along the x-axis. We assume that no lateral displacement of the leading edge of the structure is possible, although head 'yaw' is allowed. The fluid is simulated using a computational fluid dynamics technique, and we are able to determine and solve Euler-Lagrange equations for the structure. These two calculations are solved simultaneously by using a weakly coupled solver. We illustrate our method by showing that we are able to induce both forward and backward swimming. A discussion of the relevance of these simulations to a slowly swimming body, such as a mechanical device or a fish, is given.

Measurement of canal occlusion during the thoracolumbar burst fracture process.

Post-injury CT scans are often used following burst fracture trauma as an indication for decompressive surgery. Literature suggests, however, that there is little correlation between the observed fragment position and the level of neurological injury or recovery. Several studies have aimed to establish the processes that occur during the fracture using indirect methods such as pressure measurements and pre/post impact CT scans. The purpose of this study was to develop a direct method of measuring spinal canal occlusion during a simulated burst fracture by using a high-speed video technique. The fractures were produced by dropping a mass from a measured height onto three-vertebra bovine specimens in a custom-built rig. The specimens were constrained to deform only in the impact direction such that pure compression fractures were generated. The spinal cord was removed prior to testing and the video system set up to film the inside of the spinal canal during the impact. A second camera was used to film the outside of the specimen to observe possible buckling during impact. The video images were analysed to determine how the cross-sectional area of the spinal canal changed during the event. The images clearly showed a fragment of bone being projected from the vertebral body into the spinal canal and recoiling to the final resting position. To validate the results, CT scans were taken pre- and post-impact and the percentage canal occlusion was calculated. There was good agreement between the final canal occlusion measured from the video images and the CT scans.

A method for shape optimization of a hip prosthesis to maximize the fatigue life of the cement.

The long term success of total joint replacement can be limited by fatigue failure of the acrylic cement and the resulting disruption of the bone-cement interface. The incidence of such problems may be diminished by reduction of the fatigue notch factor in the cement, so that stress concentrations are avoided and the fatigue crack initiation time maximized. This study describes a method for numerical shape optimization whereby the finite element method is used to determine an optimal shape for the femoral stem of a hip prosthesis in order to minimize the fatigue notch factor in the cement layer and at interfaces with the bone and stem. A two-dimensional model of the proximal end of a femur fitted with a total hip prosthesis was used which was equivalent to a simplified three-dimensional axisymmetric model. Software was developed to calculate the fatigue notch factor in the cement along the cement/stem and cement/bone interfaces and in the proximal bone. The fatigue notch factor in the cement at the cement/stem interface was then minimized using the ANSYS finite element program while constraining the fatigue notch factor at the cement/bone interface at or below its initial level and maintaining levels of stress in the proximal bone to prevent stress shielding. The results were compared with those from other optimization studies.

Comparison of gas plasma and gamma irradiation in air sterilization on the delamination wear of the ultra-high molecular weight polyethylene used in knee replacements.

Early failure of knee replacements is thought to be due to the combination of sterilization by gamma irradiation in air and the high cyclic stresses that they endure during use. Such failures are shown through delamination and permanent deformation of the ultra-high molecular weight polyethylene (UHMWPE) component. This study investigated whether gas plasma sterilization, as an alternative to gamma irradiation in air, would give better performance after ageing in a knee replacement using a metal pin on polymer plate wear test. Fourier transform infrared (FTIR) analysis was performed on the components to assess oxidation levels and a finite element stress analysis model is presented to estimate strain at failure in the UHMWPE. Delamination occurred in the majority of the gamma-irradiated plates but did not occur in any of the gas-plasma-sterilized plates. The FTIR analysis showed that the plates gamma irradiated in air were highly oxidized when compared with the gas-plasma-sterilized plates. Plastic strain at failure was determined for the gamma-irradiated plates and found to be less than 2.4-14 per cent.

Prediction of plastic strains in ultra-high molecular weight polyethylene due to microscopic asperity interactions during sliding wear.

Studies of explanted femoral heads have shown that scratches caused by bone cement, bone or metallic particles are present on the rubbing surface. This damage has been cited as a cause of increased wear of ultra-high molecular weight polyethylene (UHMWPE) acetabular cups and it is known that the particulate wear debris produced leads to osteolysis. A series of explanted Charnley femoral heads have been surface characterized using a Talysurf 6 profilometer and found to have scratches with lip heights in the size range 0.1-3.25 microns with an average height of 1 micron giving an average aspect ratio (defined as height/half-width) of 0.1. These geometries were incorporated into a finite element model of a stainless steel asperity sliding over UHMWPE under conditions similar to those in an artificial hip system. It was found that as the aspect ratio of the asperity lip increased, the plastic strains both on and below the surface of the UHMWPE increased non-linearly, but that the magnitude of the strain was independent of the asperity height. The asperity aspect ratio was also found to affect the position of the maximum sub-surface strain, as the asperity aspect ratio was increased the maximum strain rose to the surface. The high plastic strains predicted offer an explanation for the highly elevated wear rates in scratched counterface tests and the aspect ratio of scratch lips is therefore a critical determinant of plastic strain.

A two-dimensional model of cyclic strain accumulation in ultra-high molecular weight polyethylene knee replacements.

As new methods of sterilization of the ultra-high molecular weight polyethylene (UHMWPE) component in knee replacements are introduced, reported incidents of delamination will decrease. The prediction of plastic strain accumulation and associated failure mechanisms will then become more important in knee replacement design. The finite element analysis reported in this paper aims to advance the modelling of strain accumulation in UHMWPE over repeated gait cycles and seeks to determine the effects of the knee replacement design variables of geometry and kinematics. Material testing was performed under cyclic and creep conditions to generate the elastic, viscoplastic material model that has been used in this time-dependent analysis. Non-conforming geometries were found to accumulate plastic strains at higher rates than conforming geometries. The anatomical motion known as rollback initially produced lower strain rates, but predictions of the long-term response indicated that designs which allow rollback may produce higher strains than static designs after only about a week of loading for a knee replacement patient.

Comparison of wear, wear debris and functional biological activity of moderately crosslinked and non-crosslinked polyethylenes in hip prostheses.

The wear, wear debris and functional biological activity of non-crosslinked and moderately crosslinked ultrahigh molecular weight polyethylene (UHMWPE) acetabular cups have been com pared when articulating against smooth and intentionally scratched femoral heads. Volumetric wear rates were determined in a hip joint simulator and the debris was isolated from the lubricant and characterized by the percentage number and volumetric concentration as a function of particle size. The volumetric concentration was integrated with the biological activity function determined from in vitro cell culture studies to predict an index of specific biological activity (SBA). The product of specific biological activity and volumetric wear rate was used to determine the index of functional biological activity (FBA). On smooth femoral heads the crosslinked UHMWPE had a 30 per cent lower wear rate, but it had a greater percentage volume of smaller, more biologically active particles, which resulted in a similar index of FBA compared with the non-crosslinked material. On the scratched femoral heads the volumetric wear rate was three times higher for the moderately crosslinked UHMWPE and two times higher for the non-crosslinked UHMWPE compared with the smooth femoral heads. This resulted in a higher wear rate for the moderately crosslinked material on the scratched femoral heads. All the differences in wear rate were statistically significant. There were only small differences in particle volume concentration distributions, and this resulted in similar indices of FBA which were approximately twice the values of those found on the smooth femoral heads. Both materials showed lower wear and FBA than for previously studied aged and oxidized UHMWPE gamma irradiated in air. However, this study did not reveal any advantage in terms of predicted FBA for moderately crosslinked UHMWPE compared with non-crosslinked UHMWPE.

Quantification of third body damage to the tibial counterface in mobile bearing knees.

Fourteen pairs of explanted low contact stress (LCS) tibial interface components: six rotating platform (RP), six meniscal (MN) and two anterior-posterior (AP) glide designs, have been analysed with particular attention paid to the condition of the tibial counterfaces. The average surface roughness, Ra, for the tibial trays ranged from 0.01 to 0.087 micron, significantly greater than the unworn control measurement of 0.008 micron. The scratch geometry analysis showed that the scratch peaks were found to be consistently of a lower aspect ratio than the scratch valleys and under 1 micron in height (average asperity height Rp = 0.52 micron, aspect ratio delta p = 0.01, average asperity depth Rv = 1.10 microns, delta v = 0.05). The largest scratches were 3-4 microns in both Rp and Rv. In vitro tests have shown that ultra-high molecular weight polyethylene (UHMWPE) wear increases in the presence of counterface scratches perpendicular to the direction of motion. In these explants, the unidirectional motion produced scratches parallel to the direction of sliding which is predicted to produce a smaller increase in UHMWPE wear. Other designs in mobile bearing knees have less constrained motion at the tibial counterface and this has been shown to accelerate wear; it may also lead to a further increase in wear in the presence of third body scratches. It may be possible in future knee designs to reduce this type of wear damage by introducing alternative materials or coatings which are more resistant to scratching and surface roughening.

Material optimisation of the femoral component of a hip prosthesis based on the fatigue notch fatigue approach.

Previous studies have optimised the shape of a cemented stainless steel stem in order to minimise the fatigue notch factor Kf in the cement whilst at the same time maximising Kt in the proximal medial bone to prevent bone resorption [1]. The present study firstly describes the effect of changes in the modulus of elasticity of the stem material for both the original Charnley stem and the optimised shape on Kf as predicted by a 2D finite element (FE) model of the implanted prosthesis. The paper further describes a method for parametric optimisation to determine the best material properties of a layered composite femoral stem consisting of a core material (stainless steel) and an outer layer of a different material, the elastic modulus of which is used as a design variable. The overall objective of the optimisation was to maximise Kf in the proximal bone whilst at the same time constraining Kf at all cement interfaces to be no greater than its initial value. The results of the first study suggest that Young's moduli of about 145 and 210 GPa are optimal for the monolithic Charnley and optimised stems, respectively. A composite prosthesis with a layer of modulus 31 GPa added to the optimised stainless steel stem in the proximal region only was found to significantly increase the stresses in the proximal bone and reduce Kf in the cement whilst retaining the advantages of an outer stem profile very similar to that of the original Charnley prosthesis.

The wear of oriented UHMWPE under isotropically rough and scratched counterface test conditions.

Unidirectional wear tests of UHMWPE against smooth counterfaces show that molecular chains at the surface of virgin material become oriented parallel to the sliding direction giving low wear rate. It is postulated that under more abrasive conditions and predominantly unidirectional motion as in knee prostheses, it may proof beneficial to provide molecular orientation of the bulk material. Therefore strips of UHMWPE were oriented by die drawing at elevated temperature and the resulting anisotropic material subjected to tensile tests, small punch tests and also unidirectional wear tests both parallel and perpendicular to the draw direction. The tensile tests showed that, in the parallel direction, the oriented UHMWPE became stiffer and less ductile compared to the virgin UHMWPE. In the perpendicular direction, there were reductions in yield stress, 5% proof stress and energy to failure compared to the virgin material. The small punch test showed that the oriented UHMWPE exhibited apparent hardening when tested in both parallel and perpendicular directions but the mechanical behaviour in the perpendicular direction was comparable to the virgin UHMWPE. The wear tests demonstrated that the oriented UHMWPE did not show any significant improvement of wear resistance for sliding against either isotropically rough or scratched counterfaces. There was no clear dependency between the mechanical properties and wear factors of the oriented UHMWPE.

Effect of FE idealisation, load conditions and interface assumptions on the stress distribution and fatigue notch factor in the human femur with an endoprosthesis.

The load transferred through the hip joint is one of the major forces occurring in the human body. After the replacement of this joint in THR arthroplasty, the load is transferred through the implant to the femoral bone. Loosening of the fixation of the implant and the fatigue failure of prosthetic stems create problems for both patient and surgeon. Both problems can be reduced by the use of Finite Element (FE) analysis to predict stresses and fatigue lifes but the results are sensitive to assumptions regarding the loading conditions and the idealisation of the components. Consequently the stress distributions and resulting fatigue notch factors in the human femur with an endoprosthesis have been determined for different assumptions regarding the form of the idealisation, the load conditions, and the interface conditions. The FE results show that a realistic loading condition without a tension banding force always produces the highest fatigue notch factor and von Mises stresses. An equivalent 2D plane stress model obtained by varying the thickness is likely to give more realistic stresses because it predicts more realistic strains than other 2D approximations. The full bonded interface is a satisfactory approximation for the real interface conditions because it predicts stress distributions of the correct form without excessive stress concentration.

Shape optimisation of a Charnley prosthesis based on the fatigue notch factor.

The femoral component of the artificial hip joint implanted in a patient's femur is subjected to a complex set of forces exerted due to normal life activities. It is thought that high values of stress in the cement in a cemented prosthesis can lead to fractures of the cement mantle and loosening of the stem. The incidence of such problems may be diminished by reduction of the fatigue notch factor in the cement, such that stress concentrations are avoided and the crack initiation time maximised. This study describes a method for numerical shape optimisation to determine an optimal shape for the femoral stem of a hip prosthesis in order to minimise the fatigue notch factor in the cement layer at the interfaces with the bone and stem whilst at the same time maintaining or increasing the stress levels in the proximal medial femoral bone to help prevent stress shielding. The method when used to optimise the shape of a stainless steel Charnley stem was found to be extremely efficient and effective. The resulting optimal shape was heavily waisted in the proximal region below the neck but distally was quite similar to the original design. The fatigue notch factors in the cement were reduced by 16% and 19% for medial and lateral cement/stem interfaces, respectively, and by 8% and 2% for the corresponding cement/bone interfaces. The fatigue notch factor in the proximal medial bone was increased by 57% which indicates that the general stress level in this region is markedly increased. Thus the optimised design should increase the fatigue life of the cement and at the same time reduce stress shielding in the proximal bone. Both of these effects may help prevent loosening of the femoral component and hence reduce the need for early revision operations.

Simulation of tibial counterface wear in mobile bearing knees with uncoated and ADLC coated surfaces.

A multidirectional pin-on-plate reciprocating machine was used to compare the wear performance of UHMWPE sliding against cast cobalt chrome (CoCr) plates that were either untreated or coated with Amorphous Diamond Like Carbon (ADLC). The test conditions were based on a 1/5 scale model representative of in vivo motion at the tibial counterfaces of unconstrained mobile bearing knees. The average +/- STERR wear rates were 13.78+/-1.06 mm3/Mcycles for the ADLC counterfaces and 0.504+/-0.12 mm3/Mcycles for the control CoCr counterfaces. All of the pins run on the ADLC counterfaces exhibited the same patterns of blistering along the central axis, and severe abrasion elsewhere to the extent that all of the original machining marks were removed after just one week of testing. The average value of friction coefficient was 0.24 for the ADLC counterfaces and 0.073 for the control CoCr counterfaces. The factor of 3.5 increase was statistically significant at p < 0.05. In the tribological evaluation of ADLC coatings for tibial trays in mobile bearing knees, this study shows that this specific Physical Vapour Deposition (PVD) ADLC showed significantly poorer frictional and wear performance than uncoated surfaces which was sufficient to negate any potential benefits of improved resistance to third body damage.

Comparative wear and wear debris under three different counterface conditions of crosslinked and non-crosslinked ultra high molecular weight polyethylene.

The wear debris generated from ultra high molecular weight polyethylene (UHMWPE) have been recognised as one of the major causes of failure in total hip replacements (THR). It is essential to reduce the wear debris generated from UHMWPE acetabular cups in order to minimise this problem. Debris in the submicron size range is believed to have greater osteolytic potential. It is now known that crosslinked UHMWPE acetabular cups have reduced volumetric wear rates but little is known about the influence of crosslinking on the size and morphology of the wear debris. In this study, the wear of grade GUR 1020 crosslinked (vacuum gamma irradiated), GUR 1120 crosslinked (acetylene enhanced irradiated) and non cross linked (ethylene oxide sterilised) GUR 1020 UHMWPE was compared in multidirectional pin-on-plate wear tests under three different counterface conditions (smooth, isotropically rough and scratched counterfaces). Multidirectional motion was chosen because this motion was closer to the relative motion in the natural hip. From this study, better wear resistance of crosslinked UHMWPE compared with non-crosslinked UHMWPE was demonstrated for the smooth counterface conditions. However, in the rough and scratched counterface conditions, the vacuum gamma irradiated crosslinked material produced significantly higher wear rates than the non-crosslinked material. The analysis of the wear debris showed that the majority of the volume of the acetylene enhanced crosslinked UHMWPE wear debris was in the most biologically active size range (0.1 to 0.5 microm). In contrast, the non-crosslinked material and the vacuum gamma irradiated crosslinked material had a greater proportion of the volume of the debris in the larger size ranges which are less biologically active. This has important implications for its osteolytic potential.

An experimental model of tibial counterface polyethylene wear in mobile bearing knees: the influence of design and kinematics.

Current designs of mobile bearing knees have different kinematics at the tibial counterface articulation; unidirectional represented by linear tracks and rotating platform designs, and multidirectional represented by reduced constraint designs with motion of the tibial surface in A-P and M-L directions simultaneously. One fifth scale experimental models of the tibial counterface articulation have been developed with mean contact stresses of 0.6 MPa. The unidirectional model had a linear reciprocating motion with a 10 mm stroke, the multidirectional model had a reciprocating motion with a 10 mm stroke and simultaneous rotation of +/- 7.5 degrees. Six specimens of GUR415 polyethylene were tested for each model, sliding on polished cobalt chrome counterfaces with Ra < 0.01 micron in 25% bovine serum lubricant. The mean +/- STERR wear rates were: unidirectional 0.045 +/- 0.015 mm3/million cycles and multidirectional 0.44 +/- 0.15 mm3/million cycles. Applying the scaling factor of 5, the predicted wear rates in actual knee prostheses were: unidirectional 0.23 mm3/million cycles and multidirectional 2.2 mm3/million cycles. The order of magnitude increase in wear rate was statistically significant (p = 0.05).

A comparison of the wear and debris generation of GUR 1120 (compression moulded) and GUR 4150HP (ram extruded) ultra high molecular weight polyethylene.

The wear debris generated from UHMWPE (ultra high molecular weight polyethylene) has been recognised as one of the major causes of failure in THR (total hip replacement). GUR 1120 (compression moulded) and GUR 4150HP (ram extruded) which are currently the most frequently used materials in THR were studied in pin-on-plate wear test. The wear particles generated from this test were observed by scanning electron micrograph and analysed by image analysis. The results from this study showed that GUR 4150HP had superior wear resistance than GUR 1120 under relatively high wear factor conditions. These results also highlighted the importance of multidirectional motion and its effect on the wear rates of UHMWPE. The multidirectional motion tended to show a higher wear factor than previous studies using unidirectional motion conducted under otherwise similar conditions. The wear debris analysis conducted with the wear particles collected from unidirectional (relatively rough) pin-on-plate wear tests (GUR 1120 and GUR 4150HP) showed that the greatest number of particles had a size range of 0.1-0.5 micron followed by 0.5-1.0 micron, 1.0-5.0 microns and 5.0-10.0 microns, in both GUR 1120 and GUR 4150HP. However, comparing the masses of the wear particles, the bigger size range of greater than 10 microns, had the highest percent mass followed by 1.0-5.0 microns, 0.5-1.0 micron, 0.1-0.5 micron and 5.0-10.0 microns.