**4.2. Development and validation of finite element fracture predictions with PUfoam based synthetic bones**

Acetabular fractures are one of the big challenges that trauma surgeons face today. Despite the great stride made in treating this fracture in the past few decades, one medical text book states that "fractures of the acetabulum remains an enigma to the orthopaedic surgeon(Tile et al., 2003)." The main reason for this difficulty lies on the complexity of acetabular fractures. Acetabular fractures are usually a result of indirect trauma where the major impact is transmitted via the femur after a blow to the greater trochanter, to the flexed knee or to the foot with the knee extended (Ruedi et al., 2007). Moreover acetabular fractures are dependent on a number of variables such as the type of force that caused the fracture, the direction of displacement, the damage to the articular surface as well as the anatomical types of the fracture (i.e. the shapes of the fragments). The relative rareness of acetabular fractures makes matters worse since general orthopaedic surgeons may not gain wide experience with them.

The past researches on acetabular fracture can be broadly divided into two categories. The first is experimental studies where the stability of different acetabular fracture fixation techniques was investigated with in-vitro mechanical experiments (Goulet et al., 1994, Konrath et al., 1998a, Konrath et al., 1998b, Olson et al., 2007). There are also clinical studies that examined the effectiveness and longer-term results of different fracture fixation techniques (Borrelli et al., 2005, Cole and Bolhofner, 1994, Giannoudis et al., 2005). Finite element (FE) models can enhance greatly the body of knowledge obtained from such experimental and clinical studies. FE models can overcome the limitations of in-vitro experimental studies because they can be used to simulate the behaviour of the fractured acetabulum under physiological loading conditions that include muscle forces. FE models can also elevate the results from clinical studies into a new dimension as they can be used to predict the outcome of particular fixation techniques after the surgery. If the model performance in fracture prediction is validated, it can be used to evaluate various fixation 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 synthetic PU-foam based bones(Shim et al., 2010).

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

182 Polyurethane

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

Therefore polyurethane foam was successfully used in identifying optimal parameters for designing minimally invasive bone graft harvester. The next section will describe how PU-

**4.2. Development and validation of finite element fracture predictions with PU-**

Acetabular fractures are one of the big challenges that trauma surgeons face today. Despite the great stride made in treating this fracture in the past few decades, one medical text book states that "fractures of the acetabulum remains an enigma to the orthopaedic surgeon(Tile et al., 2003)." The main reason for this difficulty lies on the complexity of acetabular fractures. Acetabular fractures are usually a result of indirect trauma where the major impact is transmitted via the femur after a blow to the greater trochanter, to the flexed knee or to the foot with the knee extended (Ruedi et al., 2007). Moreover acetabular fractures are dependent on a number of variables such as the type of force that caused the fracture, the direction of displacement, the damage to the articular surface as well as the anatomical types of the fracture (i.e. the shapes of the fragments). The relative rareness of acetabular fractures makes matters worse since general orthopaedic surgeons may not gain wide

The past researches on acetabular fracture can be broadly divided into two categories. The first is experimental studies where the stability of different acetabular fracture fixation techniques was investigated with in-vitro mechanical experiments (Goulet et al., 1994, Konrath et al., 1998a, Konrath et al., 1998b, Olson et al., 2007). There are also clinical studies that examined the effectiveness and longer-term results of different fracture fixation techniques (Borrelli et al., 2005, Cole and Bolhofner, 1994, Giannoudis et al., 2005). Finite element (FE) models can enhance greatly the body of knowledge obtained from such experimental and clinical studies. FE models can overcome the limitations of in-vitro experimental studies because they can be used to simulate the behaviour of the fractured acetabulum under physiological loading conditions that include muscle forces. FE models can also elevate the results from clinical studies into a new dimension as they can be used to predict the outcome of particular fixation techniques after the surgery. If the model performance in fracture prediction is validated, it can be used to evaluate various fixation

foam cell size diameter in order to be either continuous or discontinuous.

foam based synthetic bone was used in FE modeling of hip fracture.

**foam based synthetic bones** 

experience with them.

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.
