Finite element modelling of the Ilizarov external fixation system.

This study describes a computational method for predicting the mechanical response of any configuration of the Ilizarov external fixation system. Mechanical testing of each of the individual components (ring, threaded rod, and wire) of the Ilizarov system was used to determine the stiffness of each component. Finite element (FE) analysis was then used to model each of the individual components. Each model was tuned to match the mechanical testing. A modular FE modelling system, using a master input file, was then developed where the tuned FE models of the individual components could be generated, positioned, and interconnected to replicate a range of fixator configurations. The results showed that the stiffness predications from the FE modelling of the fixator configurations were consistently 10 per cent higher than the stiffness values obtained from the mechanical testing. The FE modelling system can be used to predict the characteristic response of the fixator configurations and clearly shows the relative changes in that response for variations in the number of components used.

Wire tension in the Ilizarov system: accuracy of the wire-tensioning device.

The Ilizarov fixator consists of tensioned wires that attach bone segments to a modular frame. The aim of this study was to establish the accuracy and precision of the wire-tensioning device supplied with the Ilizarov external fixation system. The device was used to tension a wire in direct opposition to a calibrated load cell. Five subjects tested three devices, at each of their four tension settings, in two separate sessions. Subjects could not see the true tension during the test. There were significant differences between the results for different subjects (p < 0.01) and instruments (p < 0.01) but not for different tension settings or between the two sessions. Overall mean measured tensions were 4.9 per cent (standard deviation, 4.4 per cent) below intended values. Tensions obtained at the maximum edge (completely occluded) on the scale markings were significantly (p < 0.001) closer to the nominal values (mean discrepancy, 3.6 per cent) than those at the minimum edge (mean discrepancy, 17.6 per cent). Several factors influence wire tension. Tensioning devices are not identical and the results obtained with them depend on the user. If the scale markings are completely occluded, the discrepancy between intended and actual tensions of around 5 per cent is likely to be adequate for clinical practice since surgeons do not select the most suitable tension following quantitative data assessment, but rather it is a judgement based on surgical experience and consideration for the patient weight and expected level of activity.

Investigation into the material properties of beech wood and cortical bone.

When testing medical implants it is very important to be able to test the implant using a suitable material. In the case of orthopaedic implants the optimum material is bone. Beech wood is considered a suitable substitute for bone as it has a similar Young's modulus in tension. Although it is widely used, no actual comparison of the two materials has been undertaken. The aim of this study was to compare the material properties of beech wood and cortical bone using conventional compression tests. Cortical bone samples 4 mm in diameter and 20 mm in length, were prepared from the tibia of an amputated leg. Beech wood samples were prepared to the same specifications. In compression, the Young's modulus for cortical bone was found to be 27+/-9.9 GPa (mean +/- standard deviation) and for beech wood 2.6+/-1.7 GPa. The failure load for cortical bone was 911+/-207 N and 732+/-62 N for beech wood. Although beech wood has been used as a substitute for bone in some studies, this study has shown that there are significant differences in the properties of the two materials when they are subjected to compression.

Design of a retractable intramedullary nail for the humerus.

This paper describes the design of a retractable intramedullary nail for the humerus that does not require inter-locking screws. The developed nail has a series of fins which open out from the casing to grip the medullary canal of the bone, thus securing it in position. Prototypes of the nail have been mechanically tested using static compression, dynamic compression and static torsion tests. During the compression tests no nails were found to fail. Statically a mean force of 809 N was withstood. During dynamic testing all the nails survived 1 million cycles, with a maximum applied load of 400 N. In torsion the two nails failed at 1.0 and 2.2 N m. Based on the mechanical testing, the retractable intramedullary nail would appear strong enough to withstand the expected loading conditions in the human body.

The effect of clamping a tensioned wire: implications for the Ilizarov external fixation system.

This study demonstrates that clamping a tensioned wire can cause a reduction in wire tension. Tension (about 1275 N) was applied to a wire that was subsequently clamped, using cannulated bolts, to the steel half-ring of an Ilizarov external fixator. The tension in the wire was monitored before, during and after clamping. The apparatus was disassembled and the deformations in the wire caused by the clamps were measured. This experiment was repeated 15 times. When the wire was clamped to the frame, the wire tension was reduced by 22 +/- 7 per cent (mean +/- standard deviation, SD). The drop in wire tension was linearly correlated (r = 0.96; p < 0.001) with the deformation caused by the bolts. A finite element (FE) model of the wire was also constructed. The model was pre-stressed (tensioned), and the clamping effect replicated. This analysis showed that clamping the wire could be considered to squeeze the wire outwards (like toothpaste from a tube) and so reduce its tension during fixator assembly. To assess the magnitude of this effect in the clinical situation, the FE model analysis was repeated to replicate clamping a 1.8-mm-diameter wire to a 180-mm-diameter steel Ilizarov ring component. The analysis showed that for these conditions the tension reduced by 8-29 per cent. The results of this study highlight a general engineering problem: how can a tensioned wire be secured to a structure without an appreciable loss of tension? If the performance of the structure depends on the wire tension, this performance will change when the wire is secured.

Measurement of the bony anatomy of the humerus using magnetic resonance imaging.

This study was undertaken during the development of a new humeral intramedullary nail. To determine the overall size of the nail it was necessary to have the dimensions of the humerus and intramedullary canal. The dimensions of the bony anatomy of the humerus were obtained using an open magnetic resonance imager. The right arm of 20 volunteers was scanned and the length and the dimensions of both the intramedullary canal and the cortical bone were measured. The diameter of the canal was found to be 12.1+/-2.6 mm (mean +/- standard deviation) with the middle 50 per cent of measurements between 10 and 14 mm. The overall diameter of the bone was 19.3+/-2.3 mm, with the middle 50 per cent of measurements between 18 and 21 mm. This study has shown that magnetic resonance imaging (MRI) is an effective method of collecting dimensional data from any part of the skeleton for the development of medical devices.

The effect of torsion on the appearance of the rupture surface of the ACL of rabbits.

This study investigates how torsion affects the mechanical properties and appearance of the rupture surfaces of the anterior cruciate ligament (ACL). Rupture of the ACL is frequent among athletes and twisting of the knee during rupture is often implicated as a major factor in the injury. The problem is compounded by the ACL having limited ability to heal when ruptured. Thirty-two specimens were stretched to failure at tension--torsion rates of 10 mm/min and 1 degree/s, 10 mm/min and 40 degree/s, 500 mm/min and 1 degree/s and 500 mm/min and 40 degree/s. Results from these experiments were compared with those from pure tension tests performed at displacement rates of 10 and 500 mm/min. The changes that occurred during the rupture process were monitored by video recordings and the appearance of the rupture surfaces studied by scanning electron microscopy (SEM). We found significant differences in stiffness between tests carried out with and without torsion. The macro and microscopic appearance of the ruptures from specimens tested with torsion at different displacement rates were similar although, for the slow displacement rate, the rupture surfaces were different from those from tests carried out without torsion. The failure modes for all tests carried with torsion were similar, with all the specimens failing in the ligament substance by fibre pullout. The modes of failure for tests carried at 500 mm/min with torsion were different from those tested without torsion. Half of the specimens tested without torsion failed by tibial avulsion, whereas all specimens tested with torsion failed in the ligament substance. These results suggest that torsion affects both mechanical and structural properties of the ACL and should, therefore, be considered when designing or selecting replacements for the ACL.

Preliminary comparison of the rupture of human and rabbit anterior cruciate ligaments.

OBJECTIVE: This study was aimed at examining ruptures of the human anterior cruciate ligaments by scanning electron microscopy and video imaging and comparing the appearance of the rupture surfaces with those from rabbit anterior cruciate ligaments. DESIGN: The specimens were tested to failure as femur-anterior cruciate ligament-tibia complexes using an Instron 8511 materials testing machine. BACKGROUND: Rupture of the anterior cruciate ligament is a major clinical problem, leading to instability of the knee joint. Due to the frequency and potential severity of the injuries, a need still exists for information on the biomechanical properties of ligaments under loading conditions, which occur at the time of trauma. METHODS: Four human femur-anterior cruciate ligament-tibia complexes were loaded to failure at a displacement rate of 0.008 m/s. Video recordings of the tests were used to study the progress of the ruptures and to compare the modes of failure of the ligaments. Scanning electron microscopy was employed to study the appearance of collagen fibres at the rupture surfaces. RESULTS: The modes of failure of the rabbit anterior cruciate ligament and appearance of the rupture surfaces were similar to those of the human anterior cruciate ligaments. CONCLUSION: The rabbit anterior cruciate ligament provides a model for investigating failure of the human ligament during trauma. RELEVANCE: The results will be of significance since most studies on ligaments are carried out on animal models with the intention of applying the deductions from the results to human ligaments. Examining the appearance of collagen fibres at these surfaces may help us to understand more about what actually happens during and after ligament rupture.

The effect of flexion angle on the macro and microscopic appearance of the rupture surface of the ACL of rabbits.

This study investigated how knee flexion angle affects the appearance of the rupture surface of the anterior cruciate ligament (ACL) using rabbit knees. Specimens were failed at flexion angles of 45 degrees and 90 degrees at displacement rates of 10 and 500 mm/min. Video recordings and scanning electron microscopy (SEM) studied the appearance of the rupture surfaces. At the displacement rate of 10 mm/min, the major mode of failure was fibre pullout for all flexion angles. At displacement rate of 500 mm/min, which more closely approximated loading conditions during trauma, we found that all modes of failure were exhibited. At 45 degrees flexion, the majority of specimens tested failed by avulsion. The reverse was true for specimens tested at 90 degrees where the majority of failures were fibre pullouts. At 45 degrees, there were more pulled-out fibres on rupture surfaces of specimens tested at 10 mm/min than at 500 mm/min. At 90 degrees, little difference was seen in the appearance of ruptures from both rates. SEM revealed that the waviness of collagen fibres was more pronounced at 10 mm/min. Therefore, the appearance of rupture surfaces of ACLs are affected by both flexion angle and displacement rate.

External ring fixators: an overview.

External fixation is widely used in the fixation of fractures and limb deformities. The mechanical characteristics of a specific external fixator are major factors in determining the biomechanical environment at a fracture/osteotomy site and, hence, affect the healing process. Although the optimal biomechanical environment for healing of a fracture or an osteotomy is unknown, a specific range of interfragmentary motion exists which promotes healing. It is therefore desirable that the mechanics of an external fixator can be manipulated to enable the surgeon to control the range of interfragmentary motion. The characteristics of an external fixator are defined by a large number of variables. Therefore, to gain control over the degree of interfragmentary motion, an understanding of the effect of each variable and how it interacts with the others to determine the overall characteristics of the device is required. For the past two decades, individual components and whole-frame configurations have been studied in depth. This article provides a summary of previous work concerning the mechanics of external ring fixators and how they affect the biomechanical environment at the fracture/osteotomy site.

Design of spinous process hooks for flexible fixation of the lumbar spine.

A prototype flexible fixation system for the lumbar spine was subjected to tensile testing to failure and cyclic tensile testing in order to determine any regions of weakness. The system consisted of a spinous process hook and two laminar hooks made of stainless steel (316L). Each laminar hook was attached to the spinous process hook by a loop of polyester braid secured by a crimped metal sleeve. In five tensile tests, the system failed by irreversible deformation of the spinous process hook at 2.5 +/- 0.3 kN (mean +/- standard deviation). In three cyclic tests, in which the applied tension varied sinusoidally between 0.04 and 0.4 kN at a frequency of 5 Hz, failure occurred after less than 400,000 loading cycles. This occurred as a result of fatigue crack initiation and propagation in the spinous process hook. A finite element model showed a stress concentration in the region where the crack occurred, which raised the applied stress above the tensile fatigue strength of this stainless steel. The spinous process hook was redesigned for manufacture in a titanium alloy (Ti-6AI-4V ELI) to minimize artefacts in magnetic resonance imaging. Further finite element models showed no unacceptable stress concentrations.

Mechanical testing of a flexible fixation device for the lumbar spine.

The aim of this study was to test mechanically a new flexible fixation system for the lumbar spine. This device incorporates loops of polyester braid which are secured by a crimped titanium sleeve. In tensile tests, all loops failed, by slippage through the crimped sleeve, at 434 +/- 25 N (mean +/- standard deviation from five loops) for a single crimp and 415 +/- 15 N (from five loops) for two crimps. The intact system was then tested according to the ASTM standard. In a static test, all five specimens failed by slippage of the braid through the sleeve. Initial slippage occurred between 600 and 700 N, but the mean maximum load sustained was 1090 +/- 140 N. Dynamic tests were performed on ten constructs at a frequency of 5 Hz, under a range of loading conditions. The maximum load applied in any of the tests was 825 N. Two constructs did not complete the required 5 x 10(6) test cycles because of fracture of their spinous process hooks. However, other tests, under the same conditions, showed no signs of failure. Fracture occurred as a result of fretting damage from the recommended stainless steel roll pins.

In vitro monitoring of rabbit anterior cruciate ligament damage by acoustic emission.

Anterior cruciate ligament (ACL) rupture is a major clinical problem leading to instability and degeneration of the knee joint. The problem is compounded by the limited ability of the ACL to heal when ruptured. The existing knowledge regarding the way the ACL ruptures is limited, and this investigation is an attempt to understand the nature of the ruptures using the rabbit as a model. A total of 16 rabbit tibia-ACL-femur complexes were stretched in tension to complete rupture. Four specimens were stretched to failure at a displacement rate of 0.5 mm/min and 12 specimens at 10 mm/min. Acoustic emission (AE) transducers were placed on both the tibia and the femur, and stress wave signals generated during the tensile test were recorded. Fibre fractures produced the highest amplitude signals with a relatively longer rise time. Other failure modes such as matrix failure and debonding produced lower amplitude signals with shorter rise times. We also noted that few events were recorded during the initial period of tensile loading (the elastic phase). The activity then increased significantly after maximum load was reached. The location information provided by the acoustic emission system was consistent with the final site of rupture. We have shown that AE can be used to characterise ligament damage, with fibre pull-outs and fibre fracture producing the highest signal amplitudes.

Design of an occipito-cervical fixation device.

Metal plates may be used to stabilize the cervical spine. The plates are attached to the posterior of the vertebra by placing screws into the lateral masses. The plating may be extended, in the form of rod or plate, to connect with and support the occiput. Several problems, such as screw loosening and the plate obscuring the surgeon's view as a screw is being inserted, have been identified with present plate systems. This paper describes the initial design for a cervical fixation device to overcome these problems, and the design and development that was undertaken to enable a prototype device to be manufactured.

Development of surgical instruments for implanting a flexible fixation device for the lumbar spine.

A new flexible fixation device for the lumbar spine has been developed. This paper describes the development and evaluation of two surgical instruments required for implanting this device. Prototypes were designed, manufactured and then evaluated for use in surgery. Further evaluation was performed, if necessary, and the design finalized, in accordance with BS EN 12011. This process involved close collaboration between engineers and surgeons.

Macro and microscopic examination of the ruptured surfaces of anterior cruciate ligaments of rabbits.

In a study combining tissue mechanics and fracture morphology for the first time, we examined the ruptured surfaces of anterior cruciate ligaments of rabbits and related their appearance to the initial loading conditions. Sixteen specimens were stretched to failure at rates of displacement of 10 and 500 mm/min. We used video images to study the changes which occurred during the fracture process and SEM to examine the appearance of the ruptured surfaces. The surfaces of ligaments tested at 10 mm/min had more pulled-out collagen fibres and the fibres had more pronounced waviness compared with those tested at 500 mm/min. We have shown that the macroscopic appearance of ruptured ligaments can be related to their microscopic appearance and that it is possible to deduce whether failure was by gradual tearing of the fibres or catastrophic failure.

Spinous process strength.

STUDY DESIGN: Mechanical testing of cadaveric lumbar spines and dual energy radiograph absorptiometry scanning were performed. OBJECTIVES: To devise a technique to measure the strength of lumbar spinous processes and to determine the bone mineral density of the vertebrae used. SUMMARY OF BACKGROUND DATA: The spinous process has been identified as the weakest part of the anatomy to which a flexible fixation device can be attached. It was unknown if the spinous processes could withstand the forces applied by the device. METHODS: A hook was fitted to the spinous process of 32 lumbar vertebrae. A custom-built rig was designed to secure a vertebra to a materials testing machine. A loop of cord was passed over a bar mounted on the crosshead of the machine and around the two bollards of the hook. As the crosshead was raised, a tension was applied to the cord. Each vertebra was tested to failure. The bone mineral density of each vertebra was then measured using dual energy radiograph absorptiometry. RESULTS: Failure of the specimens occurred by failure of the spinous process, pedicles, or vertebral body. The logarithm (base 10) of the load (N) at which failure occurred was 2.53 +/- 0.3, which corresponded to a mean failure load of 339 N. The bone mineral density of each vertebral body varied between 0.263 and 0.997 g/cm2. A significant linear correlation was found between bone strength and bone mineral density (P < 0.0001). CONCLUSIONS: Specimens with a bone mineral density in the range of 0.263-0.997 g/cm2 failed at a mean load of 339 N when the load was applied through the spinous process hook of a flexible fixation device.

Hip joint prosthesis design: effect of stem introducers.

The stress levels in the femoral component of a total hip prosthesis (Corin Taper Fit, Corin Medical Ltd, Cirencester, Gloucestershire, UK) were calculated by finite element (FE) analysis. This prosthesis has two holes drilled in the shoulder to engage a stem introducer. There were no unacceptable stress levels around these holes. Instead the maximum stresses were around the periphery of the shaft of the stem, as has been observed for FE analyses of conventional designs. Three prostheses were also subjected to cyclic mechanical testing (peak load 2.3 kN) according to the appropriate British Standard. The holes were examined for cracks, before and after testing, by stereomicroscopy. All three specimens were able to withstand 5 million loading cycles with no evidence of damage. Thus it is possible to design a femoral component with holes in the shoulder, to accommodate a stem introducer, without creating unacceptable stress concentrations.