All mechanical tests were done on an MTS™ servohydraulic machine (MTS Model Bionix 858, MTS Systems Corporation, Eden Prairie, MN). To simulate a fracture, the Delrin cylinder was divided into halves. The unilateral DynaFix™ fixator was used to immobilize the Delrin cylindrical halves simulating the bone segments allowing no contact at the fracture site so that all loads would be transmitted through the pins to the fixator frame. Two metal balls were attached to each cylinder, one at the distal end and one at the proximal end, to allow rotation but constrain translation when placed in the custom loading jig. In the loading jig, the cylinder was oriented to simulate the anatomic axis of the femur 6° from the loading axis of the MTS™ testing machine. As a result of this orientation, the fixator was exposed simultaneously to mediolateral bending and axial compression. Such a loading configuration simulates a more realistic (clinical) condition than a simple axial configuration.
Of the eight new fixators tested, three were tested under a high-load of 900 N cycled with a minimum compression of 150 N at a frequency of 5 Hz (the high-load low-cycle test). Because all test specimens failed within simulated one-use under this loading condition, the remaining five fixators were used for a low compressive load of 450 N cycled with a minimum compression of 100 N at a frequency of 10 Hz (the low-load high-cycle test). Even though the test construct could be regarded as elastic, loading frequency could not be increased because of the hydraulic capacity of the testing machine. One use was defined as 1 million cycles for the low-load high-cycle test and 500,000 cycles for the high-load low-cycle test. 2,10,15 The low-load high-cycle test was used to simulate normal loading after fracture immobilization under the recommended partial weightbearing restrictions. The high-load low-cycle test was used to simulate abnormal loading conditions in which the patient would resume full weightbearing under external fixation against a physician’s advice. The frame deformation and load levels were monitored throughout the cycles via a linear variable displacement transducer and a load cell. Tests were done to simulate three uses unless abrupt fracture or permanent deformation was observed.
All fixator screws and joints were tightened with a digital torque wrench (Proto 6320, Stanley-Proto Industrial Tools, Covington, GA) to a torque of 27 N-m at the beginning of a new simulated use. The initial loosening torque for each screw was measured before the beginning of a new simulated use and the screws were tightened to 27 N-m. To facilitate fixator component inspection for possible damage during the fatigue test sequence, intermediate (half-use) and final (one-use) disassemblies of the fixator were done. At every half- (500,000 cycles) and full-use (1 million cycles) interval, all key components of the device were disassembled and inspected under a microscope (Stereoscopic zoom microscope, Cole-Palmer Instrument Co, Vernon Hills, IL) before reloading for the remaining cycles. The screw loosening torque was measured at the time of inspection, and each screw was retightened to the previously measured torque at every half-use interval. After every simulated use, all fixators were cleaned (STS duoTEK Inc, Model A98–095, Rush, NY) and sterilized (STS duoTEK Inc, Model M95–1580) according to the manufacturer’s recommendation.
Statistical analyses of deformation and stiffness were done using analysis of variance (ANOVA) treating the cycle number as the independent variable and the change in component deformation and fixator stiffness values as the dependent variables to study the proposed hypothesis with the Tukey-Kramer post hoc test. Significance was set at a probability less than 0.05.
Structural Failure Reflected by Stiffness Reduction of the Fixator Construct
In the low-load high-cycle test, average frame deformation at 2 and 3 million cycles increased by 12.7% and 21.6%, respectively (Table 1), and the average stiffness at 2 and 3 million cycles decreased by 5% and 17% of the initial average stiffness, respectively. The average deformation and stiffness values at 0 to 0.5 million cycles and 0.5 to 1 million cycles were significantly different than values at 2.5 to 3 million cycles.
In the same low-load high-cycle test, initial loosening torque did not vary remarkably between the three simulated uses, although it generally decreased. Throughout the cycles, the most frequent sites of loosening were the inner screw of the proximally placed pin clamp and the outer screw of the distally placed pin clamp. Two fixators had loosening torque too small to be measured with the torque wrench (< 5 N-m): one had torque loosening at the outer screw of the distal pin clamp (1 million cycles) and at the inner screw of the proximally place pin clamp (2 million cycles); the second had torque loosening at the proximal serrated joint screw on the telescoping mechanism side and at the distal serrated joint screw on the rotary joint side, both at 3 million cycles. There was some plastic deformation observed at the tip and at the tip end threads of the rotary joint set screw and on the telescoping mechanism set screw, but these did not produce adverse effects on fixator stiffness performance.
Component Damage Under Fatigue Test
In the high-load low-cycle test, the maximum deformations throughout the cycles, produced at the last loading cycle before failure, were 3.7, 4.9, and 5.1 mm (Table 2). These values were more than twice those in the last 1 million cycles of the high-cycle test. The deformities seen in the low-load high-cycle test caused by set screw wears were not detected in the high-load low-cycle test.
In the low-load high-cycle test, all five fixators completed the three simulated uses (3 million cycles) with no evidence of component fractures, angulation at the joints, sliding at the distracted telescoping mechanism, or permanent pin deformation. However, after two uses, hairline cracks were detected in four of the five fixators (there are four serrated joints in each fixator) on the anodized Al surface of the serrated joint component (Fig 3). Eight of the 10 hairline cracks occurred on the tension side under bending. Despite these hairline cracks and the reduction of structural stiffness that occurred during the third-use sequence in the low-load high-cycle test, the frames did not fail through excessive deformation or loss of set screw or joint tightening torque after the simulated three-uses test.
In the high-load low-cycle test, all three fixators failed before completing the first simulated use of 500,000 cycles (approximate values, 150,000, 430,000, and 180,000 cycles). Major failures, such as joint angulation and material fracture of serrated joints, were observed in all three fixators (Fig 4). Hairline cracks also were observed on the anodized surface of supporting part of the serrated joint.
Estimated Reuse Criteria
Using the low-load high-cycle test condition, no appreciable component damage was seen in any of the five test specimens after two simulated uses (2 million cycles). The amount of structure stiffness decrease and screw and joint loosening torque reduction was insignificant during this testing period. However, consistent hairline cracks were observed on the anodized Al serrated joint components in four of five fixators during the third and final use test sequence. Therefore, it is reasonable to suggest that under normal partial weightbearing loading conditions, the DynaFix™ fixator could be reused for two consecutive applications in fracture treatment. Although all fixators went through three complete uses of 3 million cycles under low-load high-cycle conditions with minimal stiffness changes, the hairline cracks that occurred in the third use would strongly prevent reuse without replacement of the key components of the fixator to avoid catastrophic failure of the device.
In abnormally high loading conditions, none of the three test fixators completed the simulated short application of 500,000 cycles and the tests were terminated because of structural failure of excessive deformation and component fracture. Therefore, when a patient is suspected to have experienced full weightbearing activities, that fixator should be discarded regardless of how long it was used or how the end results of the treatment evolved. In no circumstance should the pins be reused regardless of the loading conditions and time involved.
The testing configuration used in the current study exposed the fixator frame to a severe loading condition by combining axial and bending loads. Furthermore, the construct permitted rotational freedom, introducing relatively large deformation for the fixator and putting the fixator frame in a high-loading environment, especially when the load level was increased. The load magnitude used for this study was based on the average body weight of an adult man. 1,2,15 However, in the clinical setting, patients with a lower extremity external fixator are unlikely to bear the 900 N load simulated in the current study because of pain and discomfort. The load of 450 N was considered practical for partial weightbearing on a lower limb with an external fixator.
The cyclic load of 900 N produced fatigue failures of the serrated joint. Therefore, it was determined that the high-load low-cycle test could identify the structural part most prone to fatigue fracture in the fixator frame. However, the fatigue fracture occurred in the early stage of cyclic loading even before the completion of a simulated one-use under nonphysiologic, high-loading conditions. Such testing is necessary to determine which structural part may fail under high-loading conditions. However, such an excessive condition may not provide accurate predictions in normal clinical use to assess the reusability of an external fixator. Therefore, the data obtained from the high-load low-cycle test may be used to determine which component of the fixator should be checked and replaced when reuse is being considered.
In the low-load high-cycle test, the accumulated wear and deformation of tightening screws on the telescoping mechanism and the rotary joint were detected even though the load magnitude was smaller than that in the high-load low-cycle test. These fatigue behaviors seem to be related to the micromotion at joints and they might have been cycle-dependent rather than load-dependent. In the clinical setting, excessive loading may not occur in a routine fashion because of patients’ controlled responses regulated by pain. Considering normal use, the number of loading cycles seems to be an appropriate test parameter of external fixator reuse. However, the high-load test should be used in combination with the low-load test result in the overall evaluation of fixator reuse. Therefore, the number of loading cycles is of interest in determining the fatigue life of the external fixator and the timing for refurbishing or replacing damaged fixator components under normal loading whereas high-load test data will be used to determine which component of the fixator need to be replaced in recommending any reuse possibility.
It is probable that the severe wearing of the fixator parts was a product not only of the large number of cycles in the high-cycle test, but also of the repetitive tightening and loosening procedures (five before completing 3 million cycles), which should be considered when evaluating reuse of the fixator. In the clinical scenario, tightening of each screw should be monitored so that none are over-tightened or exceed its elastic limits. A previous study showed that there is an optimum tightening torque to prevent deleterious effects caused by under-tightening or over-tightening. 11
Although all fixators endured the 3 million cycles of simulated three-uses test in the low-load high-cycle conditions, the low loosening torque measured at several set screw sites and the hairline cracks observed on the surface of the serrated joints only after 2 million cycles of loading may suggest a possibility of impending failure which must be considered in recommending fixator reuse more than once under normal clinical applications involving lower extremity weightbearing conditions. 3,14
Each patient and each fracture are different; therefore, fixators may have various degrees of load transmission. Duration of fixation, levels of weightbearing, and the patient’s weight and level of activity all must be considered in judging fixation frame reusability. However, inspecting fixator components and certifying their performance standards are time consuming and involve additional cost and device liability concerns. For practical and safety reasons, the DynaFix™ unilateral fixator may be considered for limited reuse based on the current study results. In a low-load condition, the current data suggest that the DynaFix™ unilateral fixator frame could be considered for one-time reuse provided that critical components containing the serrated joints are replaced. Under high-load conditions, reuse of the fixator should be discouraged. Fixator pins should not be reused regardless of the application conditions.
The low-load high-cycle test results could be used to provide guidelines for fixator reusability under normal clinical application. The high-load low-cycle test provides information on which parts of the fixator are vulnerable to fatigue failure and they should be replaced when reuse is being considered. These tests are necessary to assess the reusability of a well-made and highly versatile external fixator based on biomechanical considerations and cost containment issues. 5
1. Behrens F, Johnson WD, Koch TW, Kovacevic N: Bending stiffness of unilateral and bilateral fixator frames. Clin Orthop 178:103–110, 1983.
2. Chao EY, Hein TJ: Mechanical performance of the standard Orthofix external fixator. Orthopedics 11:1057–1069, 1988.
3. Chao EYS, Aro HT: Biomechanics of Fracture Fixation. In Mow VC, Hayes WC (eds). Basic Biomechanics. Ed 2. Philadelphia, Lippincott-Raven 317–351, 1997.
4. De Bastiani G, Aldegheri R, Brivio LR: The treatment of fractures with a dynamic axial fixator. J Bone Joint Surg 66B:538–545, 1984.
5. Dirschl DR, Smith IJ: Reuse of external skeletal fixator components: Effects on costs and complications. J Trauma 44:855–858, 1998.
6. Drijber FL, Finlay JB: Applying tribological principles to improve the performance of the Hoffmann external fixator’s universal joint. Proc Inst Mech Eng [H] 204:253–258, 1990.
7. Drijber FL, Finlay JB: Joint slippage in the Hoffmann external fixator. No effect of loading rate in bench experiments. Acta Orthop Scand 62:535–537, 1991.
8. Drijber FL, Finlay JB: Universal joint slippage as a cause of Hoffmann half-frame external fixator failure. J Biomed Eng 14:509–515, 1992.
9. Drijber FL, Finlay JB, Moroz TK, Rorabeck CH: Source of the slippage in the universal joints of the Hoffmann external fixator. Med Biol Eng Comput 28:8–14, 1990.
10. Kasman RA, Chao EY: Fatigue performance of external fixator pins. J Orthop Res 2:377–384, 1984.
11. Mastuura M, Kim YH, Inoue N, Walulik S, Chao EYS: Mechanical performance of a serrated joint in a unilateral fixator design. Proceedings of Seventh Biannual Meeting, International Society of Fracture Repair. Hong Kong, A-7, 2000.
12. Moroz TK, Finlay JB, Rorabeck CH, Bourne RB: External skeletal fixation: Choosing a system based on biomechanical stability. J Orthop Trauma 2:284–296, 1988.
13. Probe R, Lindsey RW, Hadley NA, Barnes DA: Refracture of adolescent femoral shaft fractures: A complication of external fixation: A report of two cases. J Pediatr Orthop 13:102–105, 1993.
14. Saikko VO: Wear of polyethylene acetabular cups against alumina femoral heads: Five prostheses compared in a hip simulator for 35 million walking cycles. Acta Orthop Scand 64:507–512, 1993.
15. Shiba R, Chao EYS, Kasman R: Fatigue properties of the Hoffman-Vidal external fixation apparatus. Orthopedics 7:443–456, 1984.
16. Skaggs DL, Leet AI, Money MD, et al: Secondary fractures associated with external fixation in pediatric femur fractures. J Pediatr Orthop 19:582–586, 1999.
© 2003 Lippincott Williams & Wilkins, Inc.
17. Wiss DA, Johnson DL, Miao M: Compression plating for non-union after failed external fixation of open tibial fractures. J Bone Joint Surg 74A:1279–1285, 1992.