**4.1. Identification of optimal design parameters in bone grafting tools with PUfoams**

Bone grafting is a reconstructive orthopaedic procedure in which a bone substitute is used to fuse broken bones and to repair skeletal defects (Arrington et al., 1996, Lewandrowski et al., 2000). Bone grafting is performed worldwide around 2.2 million times per year, with approximately 450 000 procedures in the United States alone (Russell and Block, 2000). Most popular method is autograft where the graft material is extracted from the patient itself. The graft can be harvested from the patient's femur, tibia, ribs and the iliac crest of the 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 often exceeds the pain from the primary operation.

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 bone graft harvesters as depicted in Figure 6 A.

178 Polyurethane

of the experiment.

summary of these works.

**acetabular fractures** 

**foams** 

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

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

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

**4. Use of PU-foams in device and implant testing for bone grafting and** 

**4.1. Identification of optimal design parameters in bone grafting tools with PU-**

Bone grafting is a reconstructive orthopaedic procedure in which a bone substitute is used to fuse broken bones and to repair skeletal defects (Arrington et al., 1996, Lewandrowski et al., 2000). Bone grafting is performed worldwide around 2.2 million times per year, with approximately 450 000 procedures in the United States alone (Russell and Block, 2000). Most popular method is autograft where the graft material is extracted from the patient itself. The graft can be harvested from the patient's femur, tibia, ribs and the iliac crest of the

physiological loading conditions with joint reaction force only.

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

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

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 mechanisms. For a bone graft harvester, these parameters are the rake and point angles of tool, the rotational speed and the feed rate (Figure 6 B).

Use of Polyurethane Foam in Orthopaedic Biomechanical Experimentation and Simulation 181

**Figure 8.** Various chip types formed during orthogonal cutting of polyurethane foams of various

The same experiment was repeated with bovine fresh cancellous bone(Malak and Anderson, 2008) from the patella, the femur and the iliac crest. Similar orthogonal cutting experiments were conducted using the same device used in Part I to identify major parameters that influence the formation of chips after cutting. Three groups of experiments were done where the effect of the depth of the cut, rake angle and cutting speed. Similar chip types as the experiment with PU-foams in Part I were observed which were found to be dependent on

densities

rake angle and depth of cut (Figure 9).

**Figure 9.** Chip foramtion during orthogonal cut of cancellous bone.

The influence of such parameters can be measured by characterizing chip types formed when cutting the bone. Smaller chips imply more fracturing per volume of bone material collected and due to the linkage between fracturing of the bone architecture and cell morbidity, and larger chips are believed to act as "life rafts" for the bone cells and will increase the rate of survival for embedded living cells. We have conducted two-part study where we identified various chip types during orthogonal cutting process(Malak and Anderson, 2008, Malak and Anderson, 2005).

In Part I (Malak and Anderson, 2005), we used polyurethane foams of various densities and cell sizes to investigate chip formation and surface finish. An optical arrangement made up of dynamometer, microscope and camera system (Figure 7) was used to visually record the cutting process, while horizontal and vertical cutting forces were measured. A total of 239 measurement were performed using rake angles of 23°, 45° and 60° with depths of cut from 0.1 to 3 mm (increments of 0.1 and 0.2 mm). Cutting events were observed on the video and then linked to simultaneous force events by merging both sets of data into a combined video stream, generating force plot images.

**Figure 7.** Experimental set-up for measuring chip formation during orthogonal cutting procedure

Three types of cutting response were identified and categorized as 1) surface fragmentation; 2) continuous chip formation; 3) discontinuous chip formation depending on tool rake angle, depth of cut, foam density and cell size (Figure 8). Surface fragmentation was associated with cutting depth less than the PU-foam cell size. By cutting an order of magnitude of the cell size deeper, continuous chips were produced, which is a desirable feature whenever good surface finish after cutting is desired as in the case of bone harvester. A large rake angle (60°) was also inductive of continuous chip formation. Discontinuous chip formation was associated with 1) foam compaction followed by chevron shaped chip; 2) crack propagation in front of the tool. Compaction of the foam could be minimized by using a tool with a large rake angle and normal cut depth.

mechanisms. For a bone graft harvester, these parameters are the rake and point angles of

The influence of such parameters can be measured by characterizing chip types formed when cutting the bone. Smaller chips imply more fracturing per volume of bone material collected and due to the linkage between fracturing of the bone architecture and cell morbidity, and larger chips are believed to act as "life rafts" for the bone cells and will increase the rate of survival for embedded living cells. We have conducted two-part study where we identified various chip types during orthogonal cutting process(Malak and

In Part I (Malak and Anderson, 2005), we used polyurethane foams of various densities and cell sizes to investigate chip formation and surface finish. An optical arrangement made up of dynamometer, microscope and camera system (Figure 7) was used to visually record the cutting process, while horizontal and vertical cutting forces were measured. A total of 239 measurement were performed using rake angles of 23°, 45° and 60° with depths of cut from 0.1 to 3 mm (increments of 0.1 and 0.2 mm). Cutting events were observed on the video and then linked to simultaneous force events by merging both sets of data into a combined video

**Figure 7.** Experimental set-up for measuring chip formation during orthogonal cutting procedure

using a tool with a large rake angle and normal cut depth.

Three types of cutting response were identified and categorized as 1) surface fragmentation; 2) continuous chip formation; 3) discontinuous chip formation depending on tool rake angle, depth of cut, foam density and cell size (Figure 8). Surface fragmentation was associated with cutting depth less than the PU-foam cell size. By cutting an order of magnitude of the cell size deeper, continuous chips were produced, which is a desirable feature whenever good surface finish after cutting is desired as in the case of bone harvester. A large rake angle (60°) was also inductive of continuous chip formation. Discontinuous chip formation was associated with 1) foam compaction followed by chevron shaped chip; 2) crack propagation in front of the tool. Compaction of the foam could be minimized by

tool, the rotational speed and the feed rate (Figure 6 B).

Anderson, 2008, Malak and Anderson, 2005).

stream, generating force plot images.

**Figure 8.** Various chip types formed during orthogonal cutting of polyurethane foams of various densities

The same experiment was repeated with bovine fresh cancellous bone(Malak and Anderson, 2008) from the patella, the femur and the iliac crest. Similar orthogonal cutting experiments were conducted using the same device used in Part I to identify major parameters that influence the formation of chips after cutting. Three groups of experiments were done where the effect of the depth of the cut, rake angle and cutting speed. Similar chip types as the experiment with PU-foams in Part I were observed which were found to be dependent on rake angle and depth of cut (Figure 9).

**Figure 9.** Chip foramtion during orthogonal cut of cancellous bone.

When the results from cancellous bone were compared with those from PU-foams, both showed surface fragmentation, continuous and discontinuous chip formations. During polyurethane foam cutting, chip types were influenced by rake angle and depth of cut. Bone cutting showed similar trend, however, resulting chip type were mainly influenced by the tool rake angle. We also identified the depth of cut that marked the transition from surface fragmentation to continuous or discontinuous chip types. In PU-foams, such a transition is indicated by a change in cutting forces. A similar trend was observed in bone; a depth of cut of 0.5 mm ~ 0.8 mm led to continuous or discontinuous chips. These values approximately correspond to the trabecular separation values. This is in accordance with our findings during the cutting of polyurethane foam, where depths of cut had to reach values of the foam cell size diameter in order to be either continuous or discontinuous.

Use of Polyurethane Foam in Orthopaedic Biomechanical Experimentation and Simulation 183

techniques depending on their fracture types. The problem is how to validate the model. If cadaver bone is to be used, the issue of sample variability and the requirement of ethics approval, special storage and high cost need to be resolved first. Therefore the aim of this study is to develop a finite element model of the pelvis that can accurately predict the fracture load and locations of acetabular fractures and validate its performance with

Ten synthetic male pelves made with polyurethane foam cortical shell and cellular rigid cancellous bone (Full Male Pelvis 1301-1, Sawbones, Pacific Research Laboratories, INC, Washington, WA, USA) were used for fracture experiment. A similar set-up as (Shim et al., 2008) was used where the pelvis was placed upside down in a mounting pane filled with acrylic cement. Two different loading conditions – seating (or dashboard) fracture and fall from standing fracture – were tested. When the angle (α) between the vertical line and the line formed by joining the pubic tubercle and the anterior superior point of the sacrum (Figure 10) was 30 °, the whole set-up mimicked the standing position, hence simulating standing fracture. When the angle α was raised to 45°, the set-up mimicked the position of the pelvis when seated, hence simulating seating fracture. Force was exerted from the femoral head attached to the crosshead of the Instron machine (Instron 5800 series, Norwood, MA, USA). The femoral head was also from a matching Sawbone femur (Large Left femur 1130, Sawbones, Pacific Research Laboratories, INC, Washington, WA, USA) to the pelvis used. The femur was first chopped at the neck region and then attached to a custom made holding device connected to the crosshead of the Instron machine (Figure 10). The femoral head was dipped into liquid latex to ensure a complete and stable seating of the femoral head to the acetabulum. The force was applied from the femoral head to the acetabulum at a constant speed of 40N/s until failure. Total ten pelves were used for testing. Five were tested for standing fracture and the rest was tested for seating fracture. The fracture loads and patterns were recorded for comparison with finite element simulations

**Figure 10.** Photos of the experiment: The photo on the left shows the angle alpha that determined seating or standing positions; the center photo shows the close-up of the chopped femoral head attached to the holding device that goes into the crosshead of the Instron machine; the photo on the

right shows the overall set-up.

synthetic PU-foam based bones(Shim et al., 2010).

*4.2.1. Fracture experiment with PU-foam synthetic pelvic bones* 

Therefore polyurethane foam was successfully used in identifying optimal parameters for designing minimally invasive bone graft harvester. The next section will describe how PUfoam based synthetic bone was used in FE modeling of hip fracture.
