**3. The use of PU foams in orthopaedic implant testing**

The number and variety of implants for osteosynthesis and joint replacement has increased dramatically over the past few decades along with the use of biomechanical testing of these implants to evaluate their performance. The most obvious material of choice will be fresh or embalmed cadaveric human or animal bones as they have the unique viscoelastic properties and internal structures of real bone. However such studies are often beset with a number of other problems such as issues in handling biological samples and huge variety in size, shape and material properties even in matched pairs to name a few. If reproducibility of experiments is important and comparable not absolute results are required, synthetic bones made from PU-foam can provide a great alternative to the real bones (Figure 4).

174 Polyurethane

cycles.

behaviour was similar to cancellous bones and an appropriate density of PU foams can be determined for a particular modulus value. However the shear response showed some discrepancy and concluded that caution is required when simulating other behaviours than elastic behaviour with these foams. Calvert and coworkers (Calvert et al.) evaluated cyclic compressive properties of PU foams and examined the mechanical properties in terms of microstructural features. They found that microstructural properties such as cell size and volume were uniform and increased with decreasing density. And their cyclic testing revealed hysteresis in the low density foams but consistent modulus up to 10

**Figure 3.** Compressive stress-strain curves for several relative densities (ρ\*

As the use of PU foams in orthopaedic implant testing and their use as bone analogs increased, the American Society for Testing and Materials (ASTM) developed ASTM F1839- 97, "Rigid polyurethane foam for use as a standard material for testing orthopaedic devices and instruments." The aim of this standard is to provide a method for classifying foams as graded or ungraded based on the physical and mechanical behaviour with a given density. This standard has been revised twice since 1997 when it was originally introduced in order to include a wide range of properties and nominal densities(American Society for Testing and Materials, 2008a). As such the number of studies that used PU-foam in testing implant materials and function has increased dramatically after the introduction of the standard. The

(modified from Figure 11-5 in Gibson and Ashyby, 1997)

next section will give review of those studies.

/ρs)of wet cancellous bone

(a) plastic cortical shell with cellular PU-foam inside

(b) rigid PU-foam cortical shell with cellular PU-foam inside

(c) transparaent plastic cortical shell with cellular PUfoam inside

(d) Solid PU-foam throughout

**Figure 4.** Various synthetic bone material combinations with different types of PU foams and plastics (from www.sawbones.com)

The major use of PU foam blocks is comparatives studies for quantitatively measuring some important functional parameters of orthopaedic implants such as pull out strength, stability and stiffness. Bredbenner et al. (Bredbenner and Haug, 2000) investigate the suitableness of synthetic bone made of PU-foam in testing rigidity of fracture fixations by comparing pull out strength from cadaveric bones, epoxy red oak and PU foams. They found out that PUfoam bone substitutes generated comparable results to cadaveric bones, concluding that PUfoams can be used in mechanical investigation of human bones.

Indeed many researchers have used PU-foam in comparative studies measuring pullout strength of fixation screws. Calgar et al.(Caglar et al., 2005) performed biomechanical comparative studies of different types of screws and cables using Sawbone models and found out that the load to failure of screws was significantly greater than that of the cables. Farshad et al. (Farshad et al., 2011)used PU foam blocks to test bone tunnels drilled during anterior cruciate ligament reconstruction. They found that screw embossed grafts achieved higher pull out strengths. Krenn et al. (Krenn et al., 2008)investigated the influence of thread design on screw fixation using PU-foam blocks with different densities.

PU-foams are also extensively used in biomechanical studies for finding optimal surgical parameters in orthopaedic surgeries. For example, osteotomy is a surgical procedure where a bone is cut to shorten or lengthen to rectify abnormal alignment. One variation of that technique is Weil osteotomy where the knuckle bone in the foot is cut to realign the bones and relieve pain. There are two separate independent studies on Weil osteotomy involving PU-foams and cadaver bones. Melamed et al. (Melamed et al., 2002) used 40 PU-foams to find the optimal angle for the osteotomy and found that an angle of 25° to the metatarsal shaft give the best result. This result was confirmed a year later by Trnka et al. (Trnka et al., 2001)who performed the similar study on fresh frozen cadaver feet. They also found that the range between 25°-35° give the optimal results, confirming that the use of PU-foams in such studies. Nyska et al. (Nyska et al., 2002) analyzed osteotomy for Bunion deformity using 30 PU-foams and found that displacement osteotomies provided good correction for middle and intermediate deformity. Acevedo et al. (Acevedo et al., 2002) compared five different first metatarsal shaft osteotomies by analyzing the relative fatigue endurance. They used 74 polyurethane foam synthetic bones to determine the two strongest of the five osteotomy techniques and they found that Chevron and Ludloff osteotomies showed superior endurance than the other techniques.

Use of Polyurethane Foam in Orthopaedic Biomechanical Experimentation and Simulation 177

mixture of glass fibres and epoxy resin was pressure injected around the foam to mimic cortical bone. Chong et al. (Chong et al., 2007a, Chong et al., 2007b) performed extensive mechanical testing with these synthetic bones and found out that the fourth-generation material has better fatigue behavior and modulus, strength and toughness behaviours a lot closer to literature values for fresh-frozen human bones than previous composite bones. Heiner (Heiner, 2008) tested stiffness of the composite femurs and tibias under bending, axial and torsional loading as well as measuring longitudinal strain distribution along the proximal-medial diaphysis of the femur. She found out that the fourth-generation composite bones average stiffness and strains that were close to values for natural bones (Table 1). Papini et al.(Papini et al., 2007) performed an interesting study where they compared the biomechanics of human cadaveric femurs, synthetic composite femurs and FE femur models by measuring axial and torsional stiffness. They found that composite femurs represents mechanical behaviours of healthy rather than diseased femur (e.g. osteoporosis), hence

caution is required in interpreting the data from experiment with composite bones.

**Figure 5.** Various composite bones made up of PU-foam core covered with a cortical shell of short fiber

(a) Femur (b) Humerus (c) Tibia

Anterior flexural rigidity (N m2) Natural 317

Lateral flexural rigidity (N m2) Natural 290

Axial stiffness (N/μm) Natural 2.48

Torsional rigidity (N m2/deg) Natural 4.41

**Table 1.** Structural properties of natural human and 4th generation femurs (modified from Table 2 of

Property Bone Type Value

Composite 241

Composite 273

Composite 1.86

Composite 3.21

filled epoxy (from www.sawbones.com)

(Heiner, 2008))

Nasson and coworkers (Nasson et al., 2001)used eight foam specimens for tibia and talus to evaluate the stiffness and rigidity of two different arthrodesis techniques where artificial joint ossification is induced between two bones either with bone graft or synthetic bone substitutes. They performed arthrodesis on these artificial bones made up of PU-foams and tested rotation and bending strength and recommended that the use of crossed screws for the strength, simplicity , speed and minimal tissue dissection.

Another interesting development in the use of PU foam in orthopaedic biomechanics is the development of so called composite bones. Since PU foam closely resembles cancellous bone structure and properties, a composite material made up of epoxy resin with fibre glass along with PU foam was used to create a synthetic bone where cortical and cancellous bone materials are simulated with epoxy resin and PU foam respectively (Figure 5). Zdero et al (Zdero et al., 2007, Zdero et al., 2008) tested the performance of these composite bones by measuring bone screw pullout forces in such composite bones and comparing them with cadaver data from previous literature. They found out that composite bones provide a satisfactory biomechanical analog to human bone at the screw-bone interface.

When such composite bone material is shaped according to the actual bony shapes of human bones such as femur and tibia, they can be a great alternative for cadaver bones in research and experiment. Since they closely mimic both geometry and material properties of actual bones and yet have consistency that is lacking in cadaver bones, they can lower variability significantly, offering a more reliable testing bed. Composite replicates of femur and tibia were first introduced in 1987 and then have undergone a number of design changes over the years. The currently available composite bones are fourth-generation composite bones where a solid rigid PU-foam is used as cancellous core material while a mixture of glass fibres and epoxy resin was pressure injected around the foam to mimic cortical bone. Chong et al. (Chong et al., 2007a, Chong et al., 2007b) performed extensive mechanical testing with these synthetic bones and found out that the fourth-generation material has better fatigue behavior and modulus, strength and toughness behaviours a lot closer to literature values for fresh-frozen human bones than previous composite bones. Heiner (Heiner, 2008) tested stiffness of the composite femurs and tibias under bending, axial and torsional loading as well as measuring longitudinal strain distribution along the proximal-medial diaphysis of the femur. She found out that the fourth-generation composite bones average stiffness and strains that were close to values for natural bones (Table 1). Papini et al.(Papini et al., 2007) performed an interesting study where they compared the biomechanics of human cadaveric femurs, synthetic composite femurs and FE femur models by measuring axial and torsional stiffness. They found that composite femurs represents mechanical behaviours of healthy rather than diseased femur (e.g. osteoporosis), hence caution is required in interpreting the data from experiment with composite bones.

(a) Femur (b) Humerus (c) Tibia

176 Polyurethane

endurance than the other techniques.

the strength, simplicity , speed and minimal tissue dissection.

PU-foams are also extensively used in biomechanical studies for finding optimal surgical parameters in orthopaedic surgeries. For example, osteotomy is a surgical procedure where a bone is cut to shorten or lengthen to rectify abnormal alignment. One variation of that technique is Weil osteotomy where the knuckle bone in the foot is cut to realign the bones and relieve pain. There are two separate independent studies on Weil osteotomy involving PU-foams and cadaver bones. Melamed et al. (Melamed et al., 2002) used 40 PU-foams to find the optimal angle for the osteotomy and found that an angle of 25° to the metatarsal shaft give the best result. This result was confirmed a year later by Trnka et al. (Trnka et al., 2001)who performed the similar study on fresh frozen cadaver feet. They also found that the range between 25°-35° give the optimal results, confirming that the use of PU-foams in such studies. Nyska et al. (Nyska et al., 2002) analyzed osteotomy for Bunion deformity using 30 PU-foams and found that displacement osteotomies provided good correction for middle and intermediate deformity. Acevedo et al. (Acevedo et al., 2002) compared five different first metatarsal shaft osteotomies by analyzing the relative fatigue endurance. They used 74 polyurethane foam synthetic bones to determine the two strongest of the five osteotomy techniques and they found that Chevron and Ludloff osteotomies showed superior

Nasson and coworkers (Nasson et al., 2001)used eight foam specimens for tibia and talus to evaluate the stiffness and rigidity of two different arthrodesis techniques where artificial joint ossification is induced between two bones either with bone graft or synthetic bone substitutes. They performed arthrodesis on these artificial bones made up of PU-foams and tested rotation and bending strength and recommended that the use of crossed screws for

Another interesting development in the use of PU foam in orthopaedic biomechanics is the development of so called composite bones. Since PU foam closely resembles cancellous bone structure and properties, a composite material made up of epoxy resin with fibre glass along with PU foam was used to create a synthetic bone where cortical and cancellous bone materials are simulated with epoxy resin and PU foam respectively (Figure 5). Zdero et al (Zdero et al., 2007, Zdero et al., 2008) tested the performance of these composite bones by measuring bone screw pullout forces in such composite bones and comparing them with cadaver data from previous literature. They found out that composite bones provide a

When such composite bone material is shaped according to the actual bony shapes of human bones such as femur and tibia, they can be a great alternative for cadaver bones in research and experiment. Since they closely mimic both geometry and material properties of actual bones and yet have consistency that is lacking in cadaver bones, they can lower variability significantly, offering a more reliable testing bed. Composite replicates of femur and tibia were first introduced in 1987 and then have undergone a number of design changes over the years. The currently available composite bones are fourth-generation composite bones where a solid rigid PU-foam is used as cancellous core material while a

satisfactory biomechanical analog to human bone at the screw-bone interface.

**Figure 5.** Various composite bones made up of PU-foam core covered with a cortical shell of short fiber filled epoxy (from www.sawbones.com)


**Table 1.** Structural properties of natural human and 4th generation femurs (modified from Table 2 of (Heiner, 2008))

The advert of such biomechanically compatible bone analog greatly widened the use of PUfoam in orthopaedic biomechanics experiments as more mechanically meaningful parameters such as strength and surface strains were possible to be introduced in the design of the experiment.

Use of Polyurethane Foam in Orthopaedic Biomechanical Experimentation and Simulation 179

pelvis(Betz, 2002). Autograft has the advantages of being histocompatible and nonimmunogenic, it eliminates the risk of transferring infectious diseases and has osteoinductive and osteoconductive properties (Arrington et al., 1996). However, the harvesting procedure often requires a second incision to extract the graft from the donor site, which can extend operation time by up to 20 min (Russell and Block, 2000). The ensuing donor site morbidity is regarded as "a serious postoperative concern for both patient and surgeon" (Silber et al., 2003). Ross et al. (Ross et al., 2000) reported an overall complication rate of 3.4–49%, of which 28% suffered persistent pain, which can last as long as 2 years and

The reason for donor site pain remains unclear, however, it might be proportional to the amount of dissection needed to obtain the graft (Kurz et al., 1989). Conventional bone grafting tools usually require great exposure of the donor site with accompanied trauma to nerves and muscles. Damage to nerves and muscles may be reduced by using minimally invasive bone grafting techniques (Russell and Block, 2000), which shorten the incision length by approximately 60%, reduce the amount of dissection and are roughly two times faster than conventional methods (Burstein et al., 2000). Minimally invasive tools are usually rotational cutting tools, which include trephines, bone grinders (Burstein et al., 2000) and

The harvester collects the graft, i.e. bone chips consisting of cancellous bone fragments and bone marrow, in its barrel as it turns and penetrates deeper into the bone. Despite the advantages of using minimally invasive tools such as the bone graft harvester, cell morbidity is yet unavoidable, because both fracturing of the bone architecture and heat generation accompany every bone cutting process. However, mechanical and thermal damage could be

Many researchers have studied various cutting operations, such as orthogonal cutting (Jacobs et al., 1974), drilling (Saha et al., 1982, Natali et al., 1996) milling (Shin and Yoon, 2006) and sawing (Krause et al., 1982), in order to identify some of the critical parameters that influence heat generation and to gain an overall understanding of bone cutting

reduced by improving tool geometry and by applying appropriate cutting parameters.

often exceeds the pain from the primary operation.

bone graft harvesters as depicted in Figure 6 A.

**Figure 6.** Bone graft harvester and its major parameters

Agneskirchner et al. (Agneskirchner et al., 2006) investigated primary stability of four different implants for high tibial osteotomy with composite bones and found that the length and thickness as well as the rigidity of the material strongly influence the load to failure of tibial osteotomy. Gulsen et al. (Gulsen et al.) used composite bone in testing biomechanical function of different fixation methods for periprosthetic femur fractures and compared the yield points of these techniques. Cristofolini et al. (Cristofolini et al., 2003) performed in vitro mechanical testing with composite femurs to investigate difference between good design and bad design in total hip replacement femoral stems. They placed two different implants (one good design and the other bad design) to synthetic femurs and applied one million stair climbing loading cycles and measured interface shear between the stem and cement mantle to see the result of long term performances. Their set-up involving composite bones was sensitive enough to detect the result of design difference and was able to predict long term effects of different implant designs. Simoes et al. (Simoes et al., 2000) investigated the influence of muscle action on the strain distribution on the femur. They measured strain distributions for three loading conditions that involve no muscle force, abductor muscle force only and then 3 major muscle forces in the hip. They placed 20 strain gauges on the composite femur and applied muscle and joint forces accordingly. They found out that strain levels were lower when muscle forces were applied than when only joint reaction force was used, indicating that the need to constrain the femoral head to reproduce physiological loading conditions with joint reaction force only.

As discussed up till now, the use of PU-foam based material is almost limitless and the list discussed here is by no means an exhaustive survey of the use of PU-foam based materials in orthopaedic experiment. However, our group has also been working with the PU-foam materials in our orthopaedic biomechanics extensively. The following chapters give summary of these works.
