Fractography Study of Explanted Intramedullary Nails

IOAN CRISTESCU, IULIAN ANTONIAC, MIHAI BRANZEI*, BRANDUSA GHIBAN, ROBERT CIOCOIU, ION CIUCĂ, AUGUSTIN SEMENESCU, SEBASTIAN GRADINARU*, DANIEL POP, ANCA DANIELA RAICIU 1 Clinical Emergency Hospital Bucharest, Department Orthoped & Traumatol, 8 Floreasca Av., 014461 Bucharest, Romania 2 University Politehnica of Bucharest, 313 Splaiul Independentei Str., 060042, Bucharest, Romania 3 Carol Davila University of Medicine and Farmacy, 37 Dionisie Lupu Str., 030167, Bucharest, Romania 4 Titu Maiorescu University, Faculty Pharmacy, Dept. Pharmacognosy Phytochem Phytoterapy, 16 Gh. Sincai Str., Bucharest, Romania

The fractographic examination is the main tool to obtain this vital information.

Experimental part Materials and methods
In this study the failure analysis of six failed metallic implants gathered over several years is presented. The six in vivo failed implants were retrieved according to hospital settled protocol, disinfected, marked and photographed for traceability, sealed and transferred to be studied.
The received specimens were cut using a metallographic cutting machine to obtain the regions of interest (the fracture surfaces) and metallographic samples for light microscopy studies. The fracture surfaces were studied by naked eye and using a stereomicroscope Olympus SZX7, then using scanning electron microscopes ESEM Philips XL 30 TMP and FEI Quanta Inspect F, both equipped with energy dispersive X-ray spectrometers (EDS).
The chemical composition of the implant was studied in respect to main alloying elements by EDS. Metallographic specimen preparation was performed according to the studied alloy and light microscopy studies were performed using the Reichert UnivaR and Olympus BX51 microscopes.
All results were used to infer failure mode and cause.

Results and discussions
In following paragraphs the six case studies on explanted failed intramedullary nails are presented. Three implants are made of stainless steel 316L and the other of Ti-6Al-4V alloy, a reason to divide the study in two parts: first part groups the stainless steel and the second the Ti-6Al-4V implants.

Stainless steel nails
Case 1: Seidel centro -medullar nail A centro-medullar nail was extracted from a patient after failure according to hospital settled protocol. The catastrophic failure occurred at the middle of the length of the implant, as depicted by Fig. 1. Arrow points to failure.
The major alloying elements were determined by EDS, their concentration is shown in table 1. According to the composition from table 1 the alloy is a 316L stainless steel and it conforms to ASTM F 138 and ISO 5832-1 specifications for chemical composition. Metallographic specimens were obtained from an intact region of the implant and prepared using the standard procedure for stainless steel in order to study their microstructure.  The micrograph shown in Fig. 2 shows a typical solid solution with twins suggesting a FCC system, the austenite. The inclusion content determined prior etching established that the steel is clean (low inclusions content) this way a possible cause of failure can be eliminated.
Further investigations employed a macroscopic study of the fracture surface using the stereomicroscope. The images shown in Fig. 3 reveal general fracture type aspects. In fig. 3.a the general view of the fracture surface is presented, the section shows no deformations, thus no overload can be assumed. Fatigue failure was the failure mechanism culprit, fatigue striations can be seen in fig. 3.b to 3.d, most obvious in the last. Galling is present in fig 3.b, a form of wear between sliding surfaces. Crack growth separated the surfaces and under compressive load some material was pulled with the contacting surface.
In Fig. 3.d a region which resembled to a brittle failure was inferred at this the low magnification. In this area final fracture occurred; the section area was small enough and unable to withstand the applied load.
Using the scanning electron microscope key fracture aspects were studied in an attempt to find crack origin and infer its growth direction. a.
b.   Case study 2: tibial interlocking Kuntscher nail The second failed implant presented in this study shows a similar failure mode as the first, in the middle of the length of the centro-medullar nail. The explant is shown in fig. 7, immediately after extraction and disinfection. Using EDS the concentration of the major alloying elements was determined and it is presented in table 2. According to this composition a 316L stainless steel is the implant material with alloying element conforming to ASTM F 138 and ISO 5832-1 range. From a sound region of the implant metallographic specimens were obtained and prepared accordingly to study the microstructure.
The microstructure shown in Fig. 8 reveals a typical austenitic structure with fine grains and low inclusion content. Regarding the microstructure and inclusion content the steel conforms to specifications for its use.
Fracture surface was first observed using the stereomicroscope; details from this investigation are shown in fig. 9. The fracture aspects depicted in Fig. 9.a suggest a failure in bending, the striations and smooth surface followed by a different fracture morphology observed in Fig. 9 b through d are typical signs of fatigue. The implant failed by fatigue in a bending loading scenario.
By SEM crack origin and directions were easily located in this case.
In Fig. 10.a it can be seen how the fracture suddenly changes planes in its growth. Also signs of galling are visible near the exterior edge of the implant. Again a smooth surface occupies a large amount which typical for crack growth by fatigue followed by a small region with crystalline morphology where catastrophic failure occurred [21].
In fig. 10.b the crack origin can be easily located by tracing backwards the branched river-lines towards the outside edge of the implant. Aspects of galling can be observed, by compression loading the fractured surfaces were in sliding contact.
In fig. 11.a details from a galled region are shown, the sliding direction coincides with crack growths on a NE-SV path. Branching micro-cracks can be observed parallel to this direction. In the region studied and shown in fig. 11.b same crack growth direction was observed but no galling was present.
In this case a fatigue bending failure was inferred with the origin of the crack located at implant surface.
Case study 3: tibial interlocking Kuntscher nail The third implant studied failed in a region where section is varied, at the screw hole, as depicted in fig. 12.  The alloy is and AISI 316L which conforms to both ASTM F 138 and ISO 5832-1 standard specifications. Metallographic specimens obtained from the implant were studied by light microscopy and the microstructure shown in fig. 13.a and b in transverse and longitudinal direction reveals a typical austenitic structure with fine grains and a slight preferred orientation.
Observations performed using the stereo-microscope on the fractured surface revealed mixed fracture aspects, brittle and ductile, shown in fig. 14.c and predominant brittle failure in fig. 14.b. According to fracture aspects a fatigue failure can be inferred, fatigue striation and river-lines converge to crack initiation site [22] in the NE region of fig. 14.a. The loading which caused the fracture appears to be in bending.
Using the scanning electron microscope fracture site details were observed at higher magnifications and galling can be observed in fig 15.a and b. The arrows point towards crack initiation site. Sliding direction can be observed in fig. 15.b on a NE-SV direction. a.
b. The dimpled aspect presented in fig. 16 is typical for a ductile fracture and voids can be observed at the bottom. The overload causes microvoid coalescence and the elongated dimples seen in fig. 15.a suggest shearing while in fig. 16.a enhanced by artificial height coloring in fig. 16.b dimple morphology suggests a tensile loading. a. b. The mixed fracture aspects confirm a bending fatigue failure, where a region is loaded in tension and another in compression, while the in the in-between region shear appears. The crack most likely originated at a defect near the surface, unfortunately galling covered all features that could point the exact initiation site.

Part 2: Titanium alloys nails
Along stainless steels titanium alloys are used for centro-medullar nails manufacture. The second part of the study focuses on failed implants made of titanium alloy Ti-6Al-4V.
Case study 4: Gamma3 intramedullary nailing system The first example is a Gamma 3 centro-medullar nail which was retrieved and studied after failure. In fig. 17 the failed implant is presented. http://www. revmateriale plastice.ro  According to ASTM F1108 and ASTM F1472 in the chemical composition of such alloys 5.5-6.75% Al and 3.5 -4.5%V are required, the studied implant complies with standard specifications regarding major alloying elements concentrations.
First studies were performed by light microscopy; the metallographic sample preparation procedure was the one specific for titanium alloys [23]. a. b. The microstructures depicted in fig. 18 a. and b. reveal a microstructure comprised of two phases, alpha and beta, specific for this type of alloy. Phase morphology and orientation suggests a texture produced by plastic deformation of the alloy.
Fracture surface studies were firstly performed using the stereo-microscope, in fig. 19 .a to d. details of the fracture surface are shown. In fig. 19.a fracture surface is observed from a top perspective, the features observed led to a first inference that bending was the main loading type. In fig. 19.b a corroded region was observed along brittle fracture morphology, cleavage planes and steps become obvious in fig. 19.d at higher magnifications. Deep grooves and scratches were created by the physician in his attempt to extract the implant, as observed in fig. 19.c and d.
Scanning electron microscopy studies performed on fracture surface revealed fatigue failure aspects observable in fig.  20.a and b. with a crack origin located at the surface, caused most likely by a surface defect [24]. Fine cracks are also observable in fig. 21.a and b which are most likely to originate from alpha phase grains, given its structure and characteristics [23,25].
The implant failed by fatigue, crack origin was found on the implant region where tensile stresses [26] are dominant.
Case study 5: tibial interlocking Kuntscher nail The centromedullary nail shown in Fig.22 was studied given its different loading mechanism when compared with Gamma 3 nail. The failure occurred, as pointed by the arrow in fig. 22 in the distal region, at the hole.  According to ASTM F1108 and ASTM F1472 the Al and V contents are within standard specified limits. Studies performed by light microscopy revealed a two phase microstructure specific to a Ti-6Al-4V alloy, a mixture of alpha plus beta.
The morphology and distribution of alpha and beta tends towards a bi-modal microstructure resulting from thermomechanical processing [23], the lamellar alpha phase is deformed and oriented.
Studies of the fracture surface using the stereomicroscope revealed fatigue failure details similar to those presented in [27] as shown in fig. 24 a. and b. Galling appears by sliding surface contact in compression, the details are easier to identify in fig. 24.b. Scanning electron microscopy studies revealed a regions with a faceted fracture appearance, as shown in fig. 25.
The failure of this implant occurred by fatigue in by bending.
Case study 6: tibial interlocking Kuntscher nail The last failed implant from the titanium alloys series is shown in fig. 26, the failure occurred in the proximal end, opposed to the implant studied previously.  The concentration of Al and V both comply to ASTM F1108 and ASTM F1472 specifications.
The studies using light microscopy revealed a typical microstructure formed by an alpha and beta phase mixture (Fig.  27 ), typically for Ti-6Al-4V. Large inclusions and other in-homogeneities were not observed in the microstructure. Fracture surface observed using the stereomicroscope revealed fatigue fracture details, as shown in Fig. 28. a, and mixed fracture morphologies specific to ductile and brittle failure are observed in Fig. 28.b and c. Fracture morphology shown in Fig. 28 suggests a bending load. Studies by scanning electron microscopy [28][29][30][31][32] reveal regions galling, Fig. 29.a, fatigue striations similar to those presented in appear and river lines diverge from fracture initiation site observable in Fig. 29.c. Crack growth changed planes resulting in ledges as seen in Fig. 29.b, while a faceted cleavage surface appears with fine cracks oriented in the same direction in Fig. 29.d.
Fracture in this implant was caused by fatigue in bending. Out of six mechanically heavy loaded specimens all failed by fatigue, five in bending and one in torsion loading. Whenever observable, crack origin was located near the implant surface and it was inferred that either it originates at a surface roughness feature or a dent caused by improper handling when implanted in the patient.
Another detail observed by this comparative study was the small amount of galling in the fracture surface of the titanium alloy Ti-6Al-4V, comparable with the results presented by Antoniac et al [33]. Our best guess, at current stage, was that given the lower elastic modulus of Ti-6Al-4V, 114GPa, almost half that of 316L, 210GPa, stresses are taken also by the bone. In a dissimilar materials assembly the rigid one is more stressed, thus in a titanium alloy -bone configuration the implant is less stressed than in a stainless steel -bone configuration.
Bone fracture location and implant position are important in nail failure: when distal fractures were treaded 3 out of 6 implants failed catastrophically. From a mechanical point of view the distance from bone fracture to the locking hole is very important, its increase should alter load distribution and mitigate high stresses in thin sections. Implant sizing is essential, and based upon the failure locations in this current study a section increase ore hole diameter decrease in the upper region of the implant should prevent implant failure.

Conclusions
In accordance with this and other studies performed by the authors, the implant success influence factors described previously are appreciated: -the material and processing factors have a low incidence on implant failure. Current materials used for medical devices are regulated by restrictive standards, the inclusion content is low and the mechanical characteristics are adequate. Still these factors cannot be disregarded when failure analysis is performed. -the design factors are important, implant sizing (nail diameter and length, hole diameters) should be appropriate for each fracture case. The need of personalized implants is obvious. -the choice, placement and placement technique of the implant are inherently important for the success. The distance from bone fracture to locking holes and implant size are essential for load transfer. Regarding the placement technique it is the most difficult parameter to control: each clinical case has its own particularities and difficulties and the surgeon needs to adapt to onsite conditions. The scratches or indentations created on the implant surface become crack initiation sites and need to be avoided at all costs, careful screw placement and adequate handling procedures should reduce their occurrence. -patient cooperation is needed. Sudden implant failures caused by overload were not observed in this study and had low occurrence in previous studies. Heavy implant loading speeds fatigue failure and to avoid this scenario a proper protocol associated with healing stage should be presented to the patient. In current implant failure cases material defects and failure by overload appear seldom as failure cause while fatigue remains the main failure cause in load bearing implants.