- Email: [email protected]
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame
Injury, Int. J. Care Injured (2006) 37, 941—945
www.elsevier.com/locate/injury
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame N.A. Osei a,*, B.M. Bradley b, P. Culpan a, J.B. Mitchell c, M. Barry a, K.E. Tanner c a
Department of Orthopaedics and Trauma, The Royal London Hospital, Whitechapel, London E1 1BB, United Kingdom b St. Bartholomew’s and The Royal London School of Medicine and Dentistry, Whitechapel, London E1 1BB, United Kingdom c Department of Materials, Queen Mary University of London, Mile End Road, London E1 4NS, United Kingdom Accepted 14 February 2006
KEYWORDS Ilizarov external fixator; Wire tension; Slippage torque; Locking-bolts; Frame biomechanics
Summary The wire—bolt interface in an Ilizarov frame has been mechanically tested. The optimal torque to be applied to the frame locking-bolts during physiological loading has been defined. The set-up configuration was as is used clinically except a copper tube was used to simulate bone. The force—displacement curves of the Ilizarov wires are not altered by locking-bolt torque. The force in the bone model at which pre-tension is lost increases as the locking-bolts are tightened to 14 Nm torque, but decreases if torque exceeds 14 Nm. Thus, 14 Nm is the optimal lockingbolt torque in frame. The relationship between pre-tension versus load for different locking-bolt torques arises because at low and high clamping torques poor wire holding and plastic deformation respectively occur. Wire damage was seen under light and electron microscopy. Clinically, over or under-tightening locking-bolts will cause loss of pre-tension, reduction in frame stiffness and excessive movement at the fracture site, which may be associated with delayed union. # 2006 Published by Elsevier Ltd.
Introduction The configuration of an Ilizarov frame has been handed down from surgeon to surgeon following * Corresponding author. 49 Noel Murless Drive, Newmarket, Suffolk CB8 7SA, United Kingdom. Tel.: +1 163 866 1608; fax: +1 207 247 1747. E-mail address: [email protected] (N.A. Osei). 0020–1383/$ — see front matter # 2006 Published by Elsevier Ltd. doi:10.1016/j.injury.2006.02.019
clinical trial and error. Ilizarov suggested adjusting wire tension according to the height and the weight of a patient, to provide a scientific approach to frame construction. However, there is a complex and as yet incompletely defined mathematical interaction between wire tension, ring diameter, general configuration of hardware including the properties of the wire itself that determines how effectively a frame will function.2,5,6,9,11
942 A very important, but overlooked, area is the wire—bolt interface. This interface is greatly influenced by the torque applied to the nuts that hold the wires in place. Recently this torque setting has generated interest4,10 and it is now recognised that the application of torque in the clinical setting has been subject to large variations in the past and has been a function of the strength of the operating surgeon.10 Given that frame stiffness influences fracture site motion3 and it is recognised that small axial micromovements of the order of 0.5 mm are beneficial to bone mineralisation, but large movements, that is greater than 2 mm are detrimental7 it is important that frames are configured carefully to optimise fracture healing and remain so throughout the healing process. Although the minimum torque to hold the wires without slippage (loss of pre-tension) has been established at 10 Nm10 controversy still remains as to the optimum tightening torque for the lockingbolts in an Ilizarov frame to allow it to withstand physiological loads. The literature produces conflicting reports on an upper limit to tighten the bolts. Some authors report shearing of the bolts at high torques,4 whereas others advocate torques up to and in excess of 20 Nm with no upper limit suggested.1 Additionally, the tension applied to the wire prior to clamping each wire will affect the stiffness and slippage of the wires under physiological loads. It is therefore the purpose of this study to determine if there is an upper limit to which the locking-bolts should be tightened to optimise frame performance and ascertain the consequences for frame performance if a higher torque is applied.
Materials and methods The Ilizarov frame was set up as it would be used clinically (Fig. 1). The bone was represented by a cylindrical copper tube measuring 150 mm in length and 27 mm in diameter, with a wall thickness of 3 mm. The ends of the tube were closed with nylon discs with a concave bevel in the centre. The modulus of elasticity of copper is higher than that of bone, but, given that the tube was thinner than the typical tibial or femoral shaft, it provided an equivalent structural stiffness to a tibia or femur. Its circular structure allowed accurate pre-drilling of holes to accommodate smooth Ilizarov wires crossing at 908. The copper tube was fixed between two parallel 180 mm carbon-fibre Ilizarov rings separated by 90 mm. This system models half an external fixator with no contact or load transfer, across the fracture gap. The load path through a fractured
N.A. Osei et al.
Figure 1 Testing rig with the Ilizarov set with the copper tube, base plate driven hydraulic actuator and the load cell used to measure force.
bone with an external fixation is down the bone, through the wires or pins of the external fixator, through the support bars of the fixator at the level of the fracture back through the wires or pins and into the bone.3 Thus, this set-up is mechanically symmetrical about the fracture. When there is contact between the bone ends the load taken by the bone will be shared between the fixator and the fracture, thus reducing the load transmitted through the fixator. Two 1.8 mm stainless steel Ilizarov wires with known breaking stress were then crossed 908 to each other at each of the two levels. For each wire, torque was first applied to one locking-bolt using a torque wrench (Draper-model 3004A) calibrated over an 80 Nm range. A torque of 8, 12, 14, 16 or 18 Nm was applied to the cannulated lockingbolts. The wires were then pre-tensioned using a calibrated tension device. Once the required pre-tension of 90, 110 or 130 kilograms force (kgf) (880, 1080 or 1280 N, respectively) was attained the opposite locking-bolt was secured using the torque wrench before the tension device was removed. The points of intersection of the wire with the Ilizarov rings was marked with 5 mm wide masking tape attached to the Ilizarov
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame
943
Figure 2 Ariel view showing perpendicular wire crossing angle, juxtaposition of tape against Ilizarov ring and central loading bevel, with base plate in background.
wires, two tape attachments per wire. Loss of pretension was defined as slippage of any wire by at least 1 mm at this point so that the tape was visibly separated from the side of the ring (Figs. 2 and 3). The slippage distance was chosen as it was the smallest which could easily be seen with the naked eye. Once configured the frame was screwed onto a metal base plate by four threaded stainless steel supporting rods (Fig. 1). Each rod passed through holes in both Ilizarov rings, forming a square support to the system. The base plate was attached to the actuator of an MTS 810 (MTS Minneapolis, MN, USA) servo-hydraulic mechanical test machine. A 10 kN load cell was connected between the machine crosshead and the upper end of the copper tubing via the nylon end cap. The inferior end of the tube was free to move. The copper tube was approximated against the stainless steel mount of the cross-beam manually. The load cell was then zeroed and the displacement noted. The hydraulic actuator displaced the base plate and the load generated was measured by the load cell. The frame was ramped at a rate of 5 N s 1 in 50 N intervals. MTS Test Star software running on a PC plotted the relationship between force and displacement. At each 50 N load the operators identified if the wires had slipped. The system was then taken to 200 N above this level. Each wire was used only once and the cannulated bolts were examined for damage after each test and replaced if necessary. At the end of the experiments the wires were examined under scanning electron (JOEL JSM-6300)
Figure 3 Ariel view showing slippage of wire out of locking-bolt and wire visualised between tape and ring.
and light microscopes, for evidence of scoring, and plastic deformation. The deformation of the wire caused by clamping it to an Ilizarov ring was demonstrated through optical microscopy at 12 magnification. A non-tensioned wire was clamped to a ring at both 12 and 18 Nm assessed the effect of lockingnut torque and the resultant wire deformation was then imaged. To further assess if the loss of pretension was due to slippage through the lockingbolts, wires were assessed for damage under the scanning electron microscope a JOEL 3500 scanning electron microscope (SEM) at 20 kV accelerated voltage. The wires were conductive by virtue of being stainless steel and thus surface coating was not required. Each experiment was performed at 90, 110 or 130 kgf wire tension and at 8, 12, 14, 16 or 18 Nm bolt clamping torque. For each tension—torque combination three repeat tests were performed to establish a mean and standard deviation. Force—displacement curves and load—pre-tension curves for the applied torque were generated and statistics performed using Microsoft Excel. The light and electron microscopic features of the tested wires were analysed for damage to the wires.
944
N.A. Osei et al.
Figure 4 Force—displacement curve for 14 Nm clamping torque at different wire pre-tensions.
Results The gradient of the force—displacement curves of the wires at a given locking-bolt torque were similar (0.006 mm/N, S.D. 0.0006) at each tension (Fig. 4). This would suggest that there was no change in pretension by increasing the torque at the locking-bolts and the wires behaved elastically. The loads to produce slippage were reproducible for each tension—torque combination with low standard deviations. The slippage on the load—pre-tension curves (Fig. 5) for different torques shows that wires with larger pre-tension forces can withstand greater forces transmitted to them via a loaded bone without loss of slippage as the torque on the bolts is increased from 8 Nm up to 14 Nm. The greatest axial force in the model bone tolerated by the frame wires was 1650 N (S.D. 28.87) with wires pre-tensioned to 130 kgf and clamped at 14 Nm. At 16 Nm clamping torque the amount of axial load that can be applied to the system before slippage, falls. The greatest tolerated load at 16 Nm clamping torque is 1100 N (S.D. 57.7) and occurs at 110 kgf pre-tension. At 18 Nm clamping torque the load—pre-tension response becomes a nearly horizontal line. Torques
Figure 5 Slippage load at each wire pre-tension for different clamping torques.
Figure 6 Plastic deformation of clamped Ilizarov wire seen under the light microscope 12 magnification. Superior wire clamped at 12 Nm and inferior wire at 18 Nm.
above 18 Nm caused shearing of the bolts. Moreover, at this torque there is considerable mechanical deformation of the wires and pre-tension cannot be maintained. At torques above 14 Nm pre-tension cannot be maintained at any of the clinically relevant levels, and once the slippage point of the bolts is reached wire pre-tension is lost, thus changing the fixator stiffness and its response to load. These findings correlate with microscopic examination. Under the light microscope the wires are plastically deformed (Fig. 6). This effect is also evident under the electron microscope (Fig. 7) there are morphological features consistent with slippage such as scoring. The plastic deformation damage of the wire surfaces increases with increasing clamp torque.
Figure 7 Scoring of clamped Ilizarov wire seen under the electron microscope at 20 kV, marker bar = 10 mm. Long axis of wire is horizontal, vertical marks are from machining of wire, the score marks are horizontal.
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame
Discussion The optimal torque to apply to the locking-bolts securing the wires to an Ilizarov ring is an essential factor in the mechanical behaviour of the external fixator. Mullins et al.10 suggest that a torque of at least 10 Nm is required when setting up the Ilizarov frame to prevent slippage. Although this is a safe torque, Ilizarov wires secured by locking-bolts at 14 Nm torque can withstand greater fixator loads without loss of pre-tension. Aronson and Harp1 suggest torques up to and beyond 20 Nm without failure of the clamping bolts. However, their loading was applied along the length of an individual wire, whereas in this study the load was applied to the bone model in a physiological direction. The current study shows that wires secured with torques in excess of 14 Nm lose pre-tension at lower loads. The reasoning of Mullins et al.10 and Aronson and Harp1 is correct up to 14 Nm clamping torque, however increasing the locking-bolt torque securing the Ilizarov wires beyond this point causes loss of pretension at lower loads as evidenced by slippage of the wires. Davidson et al.4 reports shearing of bolts at high torques, we found that this to be a problem with torques in excess of 18 Nm. Given the load— pre-tension curve for the torque illustrated in this work (Fig. 6) there is no indication to apply excessive torques. In fact it is detrimental to frame performance and thus ultimately to fracture healing. This apparent paradox can be explained because clamps exert a compressive stress on the section of wire they contact and cause plastic deformation of the wire. When bone is loaded in the Ilizarov the wires are deformed, which increases the loads along the wire that is already pre-tensioned. Watson et al.12 suggests that clamping the wires reduces tension within the wires by between 8 and 29% this result is consistent with our findings. Unfortunately, the torque Watson et al. to clamp their wires is not mentioned in their paper. From the present study it appears that if wires are pre-tensioned and then clamped at 14 Nm torque this gives the minimum loss of pre-tension. If one applies too low a torque the wires are poorly gripped and will likewise slip at low loads. One reason a safe upper limit was not reported by Mullins et al.10 or Aronson and Harp1 is that their methods did not allow for testing where a tensioned wire is deflected. Aronson and Harp1 mentions micro-deformation in their work but did not consider it relevant for clinical wire tensions. We disagree and the morphological changes at optical and electron micrographic levels show that considerable deformation takes place.
945
During weight bearing load passes perpendicular to the plane of fixation of the Ilizarov wires to the rings, and interacts with stretching and bending of the wires to alter the predicted displacement demonstrated by Hillard et al.8 These forces summate so that the relationship between load and pre-tension for given torques is bell-shaped and not linear.
Conclusions Currently, most surgeons are not optimising their frames at the wire-clamp interface.10 Mullins et al. have shown that surgeons generate approximately 8 Nm when using spanners in the operative setting, however this torque is sub-optimal. We advocate the use of a torque wrench to tighten the clamps to 14 Nm when configuring a frame because lower and higher torques reduce the frames ability to maintain pre-tension and resist slippage in the wires. At torques in excess of 18 Nm the bolts will fail.
References 1. Aronson J, Harp JH. Mechanical considerations in using tensioned wires external fixation system. Clin Orthop 1990;250: 50—7. 2. Calhourn JH, Li F, Ledbetter BR, Gill CA. Biomechanics of the Ilizarov fixator for fracture fixation. Clin Orthop 1992;280: 15—22. 3. Churches AE, Tanner KE, Evans M, Gwillim J. Fracture healing assessment with external fixation. Eng Med 1985;14:13—20. 4. Davidson AW, Mullins M, Goodier D, Barry M. Ilizarov wire tensioning and holding methods: a biomechanical study. Injury Int J Care Injured 2003;34:151—4. 5. Fleming B, Paley D, Kristiansen T, Pope M. A biomechanical analysis of the Ilizarov external fixator. Clin Orthop 1989;241:95—105. 6. Gasser B, Bomam B, Wyder D, Schneider E. Stiffness characteristics of the circular Ilizarov device as opposed to other conventional external fixators. J Biomech Eng 1990;112: 15—20. 7. Goodship AE, Kenwright J. The influence of induced micromovement upon the healing of experimental tibial fractures. JBJS (Br) 1985;67B:650—5. 8. Hillard PJ, Harrison AJ, Atkins RM. The yielding of tensioned fine wires in the Ilizarov frame. Proc Inst Mech Eng 1998;212H:37—47. 9. Kummer FJ. Biomechanics of the Ilizarov external fixator. Bull Hosp J Dis Orthop Inst 1989;49:140—7. 10. Mullins MM, Davidson AW, Goodier D, Barry M. The biomechanics of wire fixation in the Ilizarov system. Injury Int J Care Injured 2003;34:155—7. 11. Orbay GO, Kummer FJ, Frankel VH. The effect of wire configuration on the stability of the Ilizarov external fixator. Clin Orthop 1992, Jun;279:299—302. 12. Watson MA, Mathias KJ, Maffulli N, Hukins DWL. The effect of clamping a tensioned wire: Implications for the Ilizarov external fixation system. Proc Inst Mech Eng 2003;217H2: 91—8.
www.elsevier.com/locate/injury
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame N.A. Osei a,*, B.M. Bradley b, P. Culpan a, J.B. Mitchell c, M. Barry a, K.E. Tanner c a
Department of Orthopaedics and Trauma, The Royal London Hospital, Whitechapel, London E1 1BB, United Kingdom b St. Bartholomew’s and The Royal London School of Medicine and Dentistry, Whitechapel, London E1 1BB, United Kingdom c Department of Materials, Queen Mary University of London, Mile End Road, London E1 4NS, United Kingdom Accepted 14 February 2006
KEYWORDS Ilizarov external fixator; Wire tension; Slippage torque; Locking-bolts; Frame biomechanics
Summary The wire—bolt interface in an Ilizarov frame has been mechanically tested. The optimal torque to be applied to the frame locking-bolts during physiological loading has been defined. The set-up configuration was as is used clinically except a copper tube was used to simulate bone. The force—displacement curves of the Ilizarov wires are not altered by locking-bolt torque. The force in the bone model at which pre-tension is lost increases as the locking-bolts are tightened to 14 Nm torque, but decreases if torque exceeds 14 Nm. Thus, 14 Nm is the optimal lockingbolt torque in frame. The relationship between pre-tension versus load for different locking-bolt torques arises because at low and high clamping torques poor wire holding and plastic deformation respectively occur. Wire damage was seen under light and electron microscopy. Clinically, over or under-tightening locking-bolts will cause loss of pre-tension, reduction in frame stiffness and excessive movement at the fracture site, which may be associated with delayed union. # 2006 Published by Elsevier Ltd.
Introduction The configuration of an Ilizarov frame has been handed down from surgeon to surgeon following * Corresponding author. 49 Noel Murless Drive, Newmarket, Suffolk CB8 7SA, United Kingdom. Tel.: +1 163 866 1608; fax: +1 207 247 1747. E-mail address: [email protected] (N.A. Osei). 0020–1383/$ — see front matter # 2006 Published by Elsevier Ltd. doi:10.1016/j.injury.2006.02.019
clinical trial and error. Ilizarov suggested adjusting wire tension according to the height and the weight of a patient, to provide a scientific approach to frame construction. However, there is a complex and as yet incompletely defined mathematical interaction between wire tension, ring diameter, general configuration of hardware including the properties of the wire itself that determines how effectively a frame will function.2,5,6,9,11
942 A very important, but overlooked, area is the wire—bolt interface. This interface is greatly influenced by the torque applied to the nuts that hold the wires in place. Recently this torque setting has generated interest4,10 and it is now recognised that the application of torque in the clinical setting has been subject to large variations in the past and has been a function of the strength of the operating surgeon.10 Given that frame stiffness influences fracture site motion3 and it is recognised that small axial micromovements of the order of 0.5 mm are beneficial to bone mineralisation, but large movements, that is greater than 2 mm are detrimental7 it is important that frames are configured carefully to optimise fracture healing and remain so throughout the healing process. Although the minimum torque to hold the wires without slippage (loss of pre-tension) has been established at 10 Nm10 controversy still remains as to the optimum tightening torque for the lockingbolts in an Ilizarov frame to allow it to withstand physiological loads. The literature produces conflicting reports on an upper limit to tighten the bolts. Some authors report shearing of the bolts at high torques,4 whereas others advocate torques up to and in excess of 20 Nm with no upper limit suggested.1 Additionally, the tension applied to the wire prior to clamping each wire will affect the stiffness and slippage of the wires under physiological loads. It is therefore the purpose of this study to determine if there is an upper limit to which the locking-bolts should be tightened to optimise frame performance and ascertain the consequences for frame performance if a higher torque is applied.
Materials and methods The Ilizarov frame was set up as it would be used clinically (Fig. 1). The bone was represented by a cylindrical copper tube measuring 150 mm in length and 27 mm in diameter, with a wall thickness of 3 mm. The ends of the tube were closed with nylon discs with a concave bevel in the centre. The modulus of elasticity of copper is higher than that of bone, but, given that the tube was thinner than the typical tibial or femoral shaft, it provided an equivalent structural stiffness to a tibia or femur. Its circular structure allowed accurate pre-drilling of holes to accommodate smooth Ilizarov wires crossing at 908. The copper tube was fixed between two parallel 180 mm carbon-fibre Ilizarov rings separated by 90 mm. This system models half an external fixator with no contact or load transfer, across the fracture gap. The load path through a fractured
N.A. Osei et al.
Figure 1 Testing rig with the Ilizarov set with the copper tube, base plate driven hydraulic actuator and the load cell used to measure force.
bone with an external fixation is down the bone, through the wires or pins of the external fixator, through the support bars of the fixator at the level of the fracture back through the wires or pins and into the bone.3 Thus, this set-up is mechanically symmetrical about the fracture. When there is contact between the bone ends the load taken by the bone will be shared between the fixator and the fracture, thus reducing the load transmitted through the fixator. Two 1.8 mm stainless steel Ilizarov wires with known breaking stress were then crossed 908 to each other at each of the two levels. For each wire, torque was first applied to one locking-bolt using a torque wrench (Draper-model 3004A) calibrated over an 80 Nm range. A torque of 8, 12, 14, 16 or 18 Nm was applied to the cannulated lockingbolts. The wires were then pre-tensioned using a calibrated tension device. Once the required pre-tension of 90, 110 or 130 kilograms force (kgf) (880, 1080 or 1280 N, respectively) was attained the opposite locking-bolt was secured using the torque wrench before the tension device was removed. The points of intersection of the wire with the Ilizarov rings was marked with 5 mm wide masking tape attached to the Ilizarov
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame
943
Figure 2 Ariel view showing perpendicular wire crossing angle, juxtaposition of tape against Ilizarov ring and central loading bevel, with base plate in background.
wires, two tape attachments per wire. Loss of pretension was defined as slippage of any wire by at least 1 mm at this point so that the tape was visibly separated from the side of the ring (Figs. 2 and 3). The slippage distance was chosen as it was the smallest which could easily be seen with the naked eye. Once configured the frame was screwed onto a metal base plate by four threaded stainless steel supporting rods (Fig. 1). Each rod passed through holes in both Ilizarov rings, forming a square support to the system. The base plate was attached to the actuator of an MTS 810 (MTS Minneapolis, MN, USA) servo-hydraulic mechanical test machine. A 10 kN load cell was connected between the machine crosshead and the upper end of the copper tubing via the nylon end cap. The inferior end of the tube was free to move. The copper tube was approximated against the stainless steel mount of the cross-beam manually. The load cell was then zeroed and the displacement noted. The hydraulic actuator displaced the base plate and the load generated was measured by the load cell. The frame was ramped at a rate of 5 N s 1 in 50 N intervals. MTS Test Star software running on a PC plotted the relationship between force and displacement. At each 50 N load the operators identified if the wires had slipped. The system was then taken to 200 N above this level. Each wire was used only once and the cannulated bolts were examined for damage after each test and replaced if necessary. At the end of the experiments the wires were examined under scanning electron (JOEL JSM-6300)
Figure 3 Ariel view showing slippage of wire out of locking-bolt and wire visualised between tape and ring.
and light microscopes, for evidence of scoring, and plastic deformation. The deformation of the wire caused by clamping it to an Ilizarov ring was demonstrated through optical microscopy at 12 magnification. A non-tensioned wire was clamped to a ring at both 12 and 18 Nm assessed the effect of lockingnut torque and the resultant wire deformation was then imaged. To further assess if the loss of pretension was due to slippage through the lockingbolts, wires were assessed for damage under the scanning electron microscope a JOEL 3500 scanning electron microscope (SEM) at 20 kV accelerated voltage. The wires were conductive by virtue of being stainless steel and thus surface coating was not required. Each experiment was performed at 90, 110 or 130 kgf wire tension and at 8, 12, 14, 16 or 18 Nm bolt clamping torque. For each tension—torque combination three repeat tests were performed to establish a mean and standard deviation. Force—displacement curves and load—pre-tension curves for the applied torque were generated and statistics performed using Microsoft Excel. The light and electron microscopic features of the tested wires were analysed for damage to the wires.
944
N.A. Osei et al.
Figure 4 Force—displacement curve for 14 Nm clamping torque at different wire pre-tensions.
Results The gradient of the force—displacement curves of the wires at a given locking-bolt torque were similar (0.006 mm/N, S.D. 0.0006) at each tension (Fig. 4). This would suggest that there was no change in pretension by increasing the torque at the locking-bolts and the wires behaved elastically. The loads to produce slippage were reproducible for each tension—torque combination with low standard deviations. The slippage on the load—pre-tension curves (Fig. 5) for different torques shows that wires with larger pre-tension forces can withstand greater forces transmitted to them via a loaded bone without loss of slippage as the torque on the bolts is increased from 8 Nm up to 14 Nm. The greatest axial force in the model bone tolerated by the frame wires was 1650 N (S.D. 28.87) with wires pre-tensioned to 130 kgf and clamped at 14 Nm. At 16 Nm clamping torque the amount of axial load that can be applied to the system before slippage, falls. The greatest tolerated load at 16 Nm clamping torque is 1100 N (S.D. 57.7) and occurs at 110 kgf pre-tension. At 18 Nm clamping torque the load—pre-tension response becomes a nearly horizontal line. Torques
Figure 5 Slippage load at each wire pre-tension for different clamping torques.
Figure 6 Plastic deformation of clamped Ilizarov wire seen under the light microscope 12 magnification. Superior wire clamped at 12 Nm and inferior wire at 18 Nm.
above 18 Nm caused shearing of the bolts. Moreover, at this torque there is considerable mechanical deformation of the wires and pre-tension cannot be maintained. At torques above 14 Nm pre-tension cannot be maintained at any of the clinically relevant levels, and once the slippage point of the bolts is reached wire pre-tension is lost, thus changing the fixator stiffness and its response to load. These findings correlate with microscopic examination. Under the light microscope the wires are plastically deformed (Fig. 6). This effect is also evident under the electron microscope (Fig. 7) there are morphological features consistent with slippage such as scoring. The plastic deformation damage of the wire surfaces increases with increasing clamp torque.
Figure 7 Scoring of clamped Ilizarov wire seen under the electron microscope at 20 kV, marker bar = 10 mm. Long axis of wire is horizontal, vertical marks are from machining of wire, the score marks are horizontal.
Relationship between locking-bolt torque and load pre-tension in the Ilizarov frame
Discussion The optimal torque to apply to the locking-bolts securing the wires to an Ilizarov ring is an essential factor in the mechanical behaviour of the external fixator. Mullins et al.10 suggest that a torque of at least 10 Nm is required when setting up the Ilizarov frame to prevent slippage. Although this is a safe torque, Ilizarov wires secured by locking-bolts at 14 Nm torque can withstand greater fixator loads without loss of pre-tension. Aronson and Harp1 suggest torques up to and beyond 20 Nm without failure of the clamping bolts. However, their loading was applied along the length of an individual wire, whereas in this study the load was applied to the bone model in a physiological direction. The current study shows that wires secured with torques in excess of 14 Nm lose pre-tension at lower loads. The reasoning of Mullins et al.10 and Aronson and Harp1 is correct up to 14 Nm clamping torque, however increasing the locking-bolt torque securing the Ilizarov wires beyond this point causes loss of pretension at lower loads as evidenced by slippage of the wires. Davidson et al.4 reports shearing of bolts at high torques, we found that this to be a problem with torques in excess of 18 Nm. Given the load— pre-tension curve for the torque illustrated in this work (Fig. 6) there is no indication to apply excessive torques. In fact it is detrimental to frame performance and thus ultimately to fracture healing. This apparent paradox can be explained because clamps exert a compressive stress on the section of wire they contact and cause plastic deformation of the wire. When bone is loaded in the Ilizarov the wires are deformed, which increases the loads along the wire that is already pre-tensioned. Watson et al.12 suggests that clamping the wires reduces tension within the wires by between 8 and 29% this result is consistent with our findings. Unfortunately, the torque Watson et al. to clamp their wires is not mentioned in their paper. From the present study it appears that if wires are pre-tensioned and then clamped at 14 Nm torque this gives the minimum loss of pre-tension. If one applies too low a torque the wires are poorly gripped and will likewise slip at low loads. One reason a safe upper limit was not reported by Mullins et al.10 or Aronson and Harp1 is that their methods did not allow for testing where a tensioned wire is deflected. Aronson and Harp1 mentions micro-deformation in their work but did not consider it relevant for clinical wire tensions. We disagree and the morphological changes at optical and electron micrographic levels show that considerable deformation takes place.
945
During weight bearing load passes perpendicular to the plane of fixation of the Ilizarov wires to the rings, and interacts with stretching and bending of the wires to alter the predicted displacement demonstrated by Hillard et al.8 These forces summate so that the relationship between load and pre-tension for given torques is bell-shaped and not linear.
Conclusions Currently, most surgeons are not optimising their frames at the wire-clamp interface.10 Mullins et al. have shown that surgeons generate approximately 8 Nm when using spanners in the operative setting, however this torque is sub-optimal. We advocate the use of a torque wrench to tighten the clamps to 14 Nm when configuring a frame because lower and higher torques reduce the frames ability to maintain pre-tension and resist slippage in the wires. At torques in excess of 18 Nm the bolts will fail.
References 1. Aronson J, Harp JH. Mechanical considerations in using tensioned wires external fixation system. Clin Orthop 1990;250: 50—7. 2. Calhourn JH, Li F, Ledbetter BR, Gill CA. Biomechanics of the Ilizarov fixator for fracture fixation. Clin Orthop 1992;280: 15—22. 3. Churches AE, Tanner KE, Evans M, Gwillim J. Fracture healing assessment with external fixation. Eng Med 1985;14:13—20. 4. Davidson AW, Mullins M, Goodier D, Barry M. Ilizarov wire tensioning and holding methods: a biomechanical study. Injury Int J Care Injured 2003;34:151—4. 5. Fleming B, Paley D, Kristiansen T, Pope M. A biomechanical analysis of the Ilizarov external fixator. Clin Orthop 1989;241:95—105. 6. Gasser B, Bomam B, Wyder D, Schneider E. Stiffness characteristics of the circular Ilizarov device as opposed to other conventional external fixators. J Biomech Eng 1990;112: 15—20. 7. Goodship AE, Kenwright J. The influence of induced micromovement upon the healing of experimental tibial fractures. JBJS (Br) 1985;67B:650—5. 8. Hillard PJ, Harrison AJ, Atkins RM. The yielding of tensioned fine wires in the Ilizarov frame. Proc Inst Mech Eng 1998;212H:37—47. 9. Kummer FJ. Biomechanics of the Ilizarov external fixator. Bull Hosp J Dis Orthop Inst 1989;49:140—7. 10. Mullins MM, Davidson AW, Goodier D, Barry M. The biomechanics of wire fixation in the Ilizarov system. Injury Int J Care Injured 2003;34:155—7. 11. Orbay GO, Kummer FJ, Frankel VH. The effect of wire configuration on the stability of the Ilizarov external fixator. Clin Orthop 1992, Jun;279:299—302. 12. Watson MA, Mathias KJ, Maffulli N, Hukins DWL. The effect of clamping a tensioned wire: Implications for the Ilizarov external fixation system. Proc Inst Mech Eng 2003;217H2: 91—8.