BME 4910 Final Report

Transcription

1 BME 4910 Final Report Novel Polysaccharide-derived Fixation Device for Anterior Cruciate Ligament (ACL) Reconstruction Team 13 Derek Holyoak Alexander Werne Benjamin Roberts Clients: Dr. Krystyna Gielo-Perczak (BME Program, University of Connecticut) Phone: Dr. Sangamesh Kumbar (UConn Health Center) Phone: Dr. Gloria Kolb (UConn Health Center) Dr. Cato Laurencin (UConn Health Center) Phone: University of Connecticut Storrs, CT UConn Health Center 263 Farmington Avenue Farmington, CT 06030

2 Table of Contents Abstract Introduction Background Purpose of Project Previous Work Done by Others Products Patent Search Results Map for the rest of the report Project Design Introduction Alternative Designs Alternative Design Alternative Design Alternative Design Choice of Optimal Design Optimal Design Objective Subunits Hamstring Tendon Casing Interference Screw Hamstring Tendon Graft Interference Screw Mold Optimal Design Analysis Hamstring Tendon Casing Interference Screw Prototype Hamstring Tendon Casing Prototype Interference Screw Prototype Prototype Operation Interference Screw Mold Prototype Prototype Testing with Client Realistic Constraints Safety Issues Impact of Engineering Solutions Life-Long Learning Budget Team Members Contribution to the Project Conclusion References Acknowledgements

3 12 Appendix Updated Specifications Other Information

4 Abstract Ineffectiveness of current ACL reconstruction techniques suggests that a novel method for this surgery may be necessary. Devices currently being used are either too rigid or too flexible, resulting either in a stiffened tendon or tendon displacement, which ultimately causes pain and discomfort for the patient. The goal of this project was to design a fixation device that can be used for ACL reconstruction and to integrate a novel biodegradable, biocompatible, polysaccharide-derived material. The device was designed taking into consideration specific constraints due to the geometry of the knee joint, the loading of the ACL, and ease of use for the surgeon. Using CAD and FEA, a fixation device with a novel geometry, called the tendon casing, was created for the femoral side of the tendon, and an interference screw was designed for the tibial side. Under 2000N of loading (maximum tension in a native ACL), the tendon casing experienced a maximum stress of approximately 9.3 MPa, while the tensile yield strength of the polysaccharide-derived material is, on average, 32.6MPa. These results show that the geometry of the tendon casing is sufficient for the ACl application. Thus, both the tendon casing and interference screw were successfully 3D printed out of PLA, which has similar mechanical properties to our biodegradable material. Future work includes fabricating a mold for the interference screw, so mass production can ensue, as well as physical testing of the device made from the novel biodegradable material. 1. Introduction Injuries to the anterior cruciate ligament (ACL) occur in approximately 200,000 people every year in the United States [1], and can be debilitating for individuals who lead a very active lifestyle. The ACL is vital to maintaining knee stability and coordinating the kinematic motion of the knee. The knee joint is a bone-to-bone junction that has ligaments and muscle attachment sites that surround the interface to prevent any abnormal motion [2]. As seen in Figure 1 below, there are four main ligaments that have a unique function in knee stabilization: anterior cruciate ligament (ACL) and posterior cruciate ligament (PCL) within the knee cavity, lateral collateral ligament (LCL) on the outside of the knee and medial collateral ligament (MCL) on the inside of the knee. ACL is a common injury due to its specialized function and insertion sites on the femur and tibia. It restrains about 85% anterior translation of the tibia with respect to the femur, with relatively lesser restraint against inward and outward rotation [2]. In vivo tests have shown that the ultimate failure load of the native ACL is approximately 2000 N [3]. 4

5 Figure 1: Rear view of a left knee joint [4] The ACL is most often torn or in severe cases ruptured when an athlete encounters a disruption while they are decelerating [5]. These quick motions are often commonplace in agility sports such as soccer and football. These disruptions can take many forms including player-toplayer contact or an uneven surface. While many ACL tears can recover through careful rehabilitation and re-stabilization of the knee, complete ruptures of the tendon are often surgically corrected. Due to the invasive nature of this procedure and potential for failure during rehabilitation, it is vital that the tendon placement itself is stable. While about 75-80% of the patients will regain stable knees post-surgery due to successful surgical techniques, the remaining patients have persistent pain, knee instability or knee stiffness that may require additional surgeries to correct these issues [6]. Overall, the surgery and the materials involved, while generally effective, have still not been perfected. In order to reconstruct an ACL, interference screws are generally used to secure a graft within the tibia and femur. Many of these screws are made from metals or polymers, which both cause problems after the surgery has taken place. Metal screws do not degrade, and if the metal remains in the body for too long of a period, the implant may both weaken the bone due to stress concentration and/or become loose due to constant friction on the bone-implant interface. Because of this, metal screws may need to be removed and may require sequential surgeries to correct the problem [7]. Bioabsorbable screws, on the other hand, are made from various polymers and composite materials, which ideally wouldn t have these compatibility issues. Unfortunately, the degradation byproducts of these polymers are acidic and have an adverse effect on the surrounding tissue at the surgical site. Therefore, interference screws are not a perfect solution for ACL reconstruction. Many other fixation devices, besides interference screws, have been used to reconstruct the ACL. These devices have been used with success, but not all surgeons are willing to work with them, because of difficulty to implement the device and they are not as available as 5

6 interference screws. Some of these devices are, however, an excellent alternative to screws, and provide a solid foundation for future designs for other fixation devices. There are, however, multiple problems associated with these devices, including metal staying in the body for too long of a period and loosening of the actual implant that causes the graft to be insecure. Hence, they propose similar problems to those of interference. However, with the correct bioabsorbable material and a combination of features from already existing products, as well as some innovative ideas, it is possible that a better fixation device can be designed for ACL reconstruction. The device should be both biocompatible and bioabsorbable. It should not degrade into any harmful byproducts, and it must provide the appropriate mechanical stability for the surgical application. 1.1 Background Our client has requested a new design for a fixation device using a novel biodegradable material that has recently been developed. The material is a polysaccharide derivative that has been used for other applications such as dialysis membranes and wound care systems, yet there is uncertainty about the applicability to the high stress environment in anterior cruciate ligament (ACL) reconstruction surgeries. Polysaccharide derivatives have both exceptional strength and biocompatibility, but also have the potential to be osteoinductive. The aim for our project is to design a mechanically stable, biocompatible, and bioactive fixation device made from this novel polysaccharide derivative. Mechanical simulations and finite element analysis will be required in order to understand the stresses placed on the device for ACL reconstruction surgery. Ideally, the device will maintain its integrity during placement into the bone and eventually degrade so that the patient s native bone can maintain stability of the tendon placement. 1.2 Purpose of the project We have two primary goals in mind for this project. The first goal involves recreating an interference screw on SolidWorks, followed by fabrication of an interference screw from the novel polysaccharide-derived material. Our client has expressed the need for a molding tool, so that the interference screw can be mass produced. Once this part of the project, is complete, we will be able to assess the performance of the novel material and compare it to the mechanical strengths of those materials that are currently being used for interference screws. The second goal of the project is to create our own, unique fixation device. We plan on incorporating all of the positive features from devices currently being used, while eliminating the negative features. This involves ensuring the device has enough mechanical integrity to withstand the tensile load caused by a native ACL, as well as allowing enough flexibility for the graft tendon to behave as a native tendon would. 6

7 1.3 Previous Work Done by Others Products Currently there are many products on the market used by orthopedic surgeons for anterior cruciate ligament reconstruction. These devices are used to fix the graft tendon in place, ultimately restoring natural knee motion and support. These products range from metal or bioabsorbable interference screws all the way to button-like locks. The screw design incorporates a tapered profile along the entire length of the screw in order to fit securely and fill the tunnel. A drawback to interference screws in general, for both metallic and bioabsorbable, is their tendency to slip back through the tibial or femoral tunnel, significantly reducing the fixation strength of the system [8]. Even if the screws don t slip and keep the tendon locked in place, they can oftentimes provide too much stability and limit the range of motion of the knee. Biomet, a medical device company, currently markets a different femoral and tibial fixation device called the EZLoc. Basically, the EZLoc incorporates a loop which the bone block of the graft is attached. This loop is then attached to the EZLoc and the graft attaches to the titanium fixation device which is pulled through either the femoral or tibial tunnel. Once through the bone, a locking mechanism protrudes from the device preventing the graft from retreating back through the tunnel. Another device on the market called the Endobutton by Smith & Nephew functions similarly to the EZLoc. When using this device, surgeons thread the graft through a titanium loop and pull the graft through the tunnel using a guide-wire. Once through the tunnel, an auxiliary wire attached to the device is pulled and the device locks on the cortical surface of the femur. Both methods are advantageous for the following reasons: they do not cause wear or abrasion of graft as opposed to interference screws and they do not rely as heavily on the strength of local bone for mechanical stability, which is beneficial for patients suffering from osteoporosis [9]. However, there are noticeable disadvantages of these devices, which include: decreased stiffness in the overall graft fixation and the possibility of graft tunnel motion. There are two types of graft tunnel motion that can lead to enlargement of the tunnel and a decrease in bone formation, both around and inside the tunnel. The first type is called longitudinal graft motion, characterized by protraction and retraction of the graft through the tunnel. The second type is called sagittal graft motion in which the graft moves from one side of the tunnel to the other during normal knee motion. All of these problems that occur with the current products in the market suggest a strong need for an alternate fixation device. In addition, all the devices aside from the interference screws are made from synthetic materials that don t have the potential for biodegradation, which is a design flaw in the fixation method itself. 7

8 1.3.2 Patent Search Results Bioabsorbable interference screw for endosteal fixation of ligaments Patent number: Filing date: Apr 4, 2005 Issue date: Jan 29, 2008 This is a patent for a bioresorbable interference screw that provides superior fixation strength and promotes fast healing during reconstructive ligament surgery. The screw features a tapered tip for ease of placement as well as a hollow core in which the placement device fits into via radially extending slots. The screw has an ultimate failure load of 830 ± 168N and a stiffness of 60 N/mm [10]. Endobutton continuous loop for bone tendon with double loop knot Patent number: Filing date: Jul 20, 2004 Issue date: May 12, 2009 The Endobutton consists of a button-like fixation device and a tough string that secures the ACL graft tendon. Essentially, a small hole is made in the bone block that the string goes through. The string is then double-knotted around the button device, which then rests against the outer femur, holding the graft in place on the femoral side via tension. On the tibial side, an interference screw is generally used. This patent has been used because of its excellent flexibility properties, but sometimes, the button can become loose since it is not locked into place; it is simply resting against the femur. The ultimate load of this device is 572±105N and the stiffness is 61±11N/m [10]. Method and apparatus for performing ACL reconstruction Patent number: Filing date: Feb 2, 2006 Issue date: Jun 15, 2010 This patent outlines a procedure for ACL and PCL reconstructive surgery. It includes steps for the drilling of tibial and femoral graft tunnels and how to use guide-wires to place fixations devices such as interference screws. 8

9 1.4 Map for the rest of the report In the second section of the report we will discuss the overall project design with regards to the alternative designs and the optimal design. Then we will provide descriptions for each of the subunits of the optimal design. In the third section of the report we will underline the realistic constraints of our design including: engineering standards, manufacturability, economic and environmental to name a few. The fourth section will pertain to safety issues involved in the design, manufacturing, prototyping and testing of our device. The fifth section will include our evaluation of the impact of our device on medical industry as well as patient health care. In the sixth section we will discuss the areas in which we have gained knowledge throughout the duration of this project. We will provide a detailed budget for the project in the seventh section. The eighth section will list each group member s contribution to the overall project as well as a description of each of the individual contributions. In the ninth section we will include a conclusion that will provide any clarifications as well as an emphasis of the major points we hope the reader will take away from the report. Also we will include our source references, acknowledgments and our appendix in the tenth, eleventh and twelfth sections respectively. 2. Project Design 2.1 Introduction This section will describe the process by which we arrived at our optimal design. As shown in Section 1 of the report, there was a lot of prior research that needed to be completed in order to begin thinking of ideas for the design of our device. Much work was done to find the existing patents on fixation devices for ACL reconstruction surgery and to understand how many of the various types of surgeries actually work. Essentially, we needed to know the direction of the holes that were drilled in both the tibia and femur, where the appropriate location was to insert the fixation device and graft tendon, how previous devices locked into place, and which types of graft tendons were used for existing surgeries. Furthermore, much research was put into the actual forces and moments that are experienced by the ACL. This proved to be quite difficult, as many research articles use differing moments and loads. The goal for our device is to make it be able to withstand worst-case scenario moments and loads. Lastly, we needed to make sure that the novel polysaccharide-derived material that we would be using has sufficient mechanical properties to be used for ACL fixation device. Once all of this information was collected and verified, we were able to start thinking of some alternative designs that might prove to be sufficient for our task. 9

10 2.2 Alternative Designs Alternative Design 1 This design aims to create a more stable and bioactive alternative to the Endobutton, where only one end of the patellar bone-ligament-bone graft is anchored to tibial tunnel with a biodegradable interference screw. The other end that passes through the femoral tunnel will be secured using this novel design. There are two opposing issues that can occur in graft placement depending on how the procedure is completed: laxity or stiffness. Oftentimes with interference screws and the Endobutton, you can develop these loosening or stiffness issues due to the variable surgical techniques employed with these devices. The following device satisfies the need for a stabilization device that does not allow slippage, but also provides a certain degree of flexibility to create the optimal amount of tension in the autograft. During ACL reconstruction surgery, there are a few main components that are constant across all techniques that were integrated into the design of this device. First, a hole is drilled entirely through the tibia and femur at an angle to simulate the attachment points of the native ligament. This is often 10mm in diameter, and the tunnel allows for smooth passage of the crimped bone graft which is often about 9mm in diameter [11]. With regard to the Endobutton, this device utilizes the idea of two different sized drills: a smaller drill to pass entirely through both bones and a larger drill (10mm diameter) that is only passed through the bottom half of the femur cavity and the entire tibia. This allows for the placement of the Endobutton. In our first alternative design, the standard 10mm drill and a larger 14mm drill will be used instead to create a large opening on the outer cortex surface of the femur. By creating a ledge or lip on the upper tunnel opening, this will serve as a table for our device to sit on and stabilize the graft below. Second, a hole is usually drilled in one bone block so that surgical thread can be passed through. This allows for the surgeon to pull the graft through the tunnel to the correct anatomical placement height. This design takes advantage of this concept to stabilize the graft to our device. In the surgical schematic shown in Figure 2, the thread simultaneously pulls the graft through the bone tunnel and correctly aligns the holes of the graft with the holes in the device. Once aligned, this thread is used as a guide wire to place a hollow cylinder made from the same biodegradable material into the graft-device combination. This will effectively fasten the graft into our device once the cylinder is locked into place. The thread can then be easily pulled out of the surgical site if biodegradable suture thread was not used. This setup with prevent any inferior translation through the bone tunnel, yet there is nothing preventing this device from superior translation. In order to solve this issue, a gear component on the superior surface of the device was created. When a screwdriver is placed in the gear s socket and turned to the right, this will move the two blades on either side of the gear outwards into the surrounding bone tunnel. Since the blades are double ended and can be extended in either direction, this allows further flexibility during surgery through alternative blade placement depending on the integrity and thickness of the surrounding bone. When fully 10

11 engaged, these blades will have passed into the bone, further securing the device. A view of the device can be seen in Figure 3 below from the top surface, showing this gear mechanism. Figure 2. On the left, a sagittal view with surgical schematic detailing the method of securing the bone-ligament graft within the biodegradable device. On the right, a threedimensional view shows all the components assembled. Figure 3. On the right, there is a three-dimensional view of the device. On the left, a transverse view of top of the device with schematic showing how the gear for this design would be engaged into the surrounding bone tunnel itself. 11

12 2.2.2 Alternative Design 2 Another main problem with interference screws on the market is that they fail to stay in place and tend to become loose after a period of time, resulting in constant friction between the screw-to-bone interface. This second alternative design improves the design of the interference screw, by adding a unique locking mechanism within the screw head, itself. The procedure to implant the screw will be similar to current techniques with only one additional step (securing the locking mechanism into the bone). From the outside, the device will look like a simple screw, made from the novel degradable polysaccharide, with one major exception. The difference lies on the screw head. Instead of the top of the interference screw looking like a normal screw head, there will be four tiny screws facing outward, each of them securely locked by an extension coming out of the head of the screw as seen in Figure 6. There will be a screw hole in each of these semi-circular attachments, to allow the passage of the four tiny screws. The four tiny screws will be preplaced into these screw holes prior to surgery, and then drilled into the bone post-implantation. The purpose of these tiny screws is to lock the interference screw in place, so that the screw threads are not the only aspect of the design that secures itself. The difficult part is to make these tiny screws come out of the screw head after the interference screw is in place in the femur or tibia. However, this has been considered, and a special tool will be made to insert these tiny screws into the bone. This tool consists of a simple gear mechanism, the will lock itself into the tiny screws, whose heads also are a gear. The gear from the tool will lock simultaneously to the gear-like heads of the four tiny screws as seen in Figure 5, and once twisting begins, the screws will gradually move outwards into the bone. The gear on the tool will be larger than those on the tiny screws, so that as rotation occurs, the gears will remain locked in place as the screws travel farther into the bone. The figures below show some different views of this second alternative design. To begin, Figure 4 shows how each of the tiny screws will be drilled into the bone once the interference screw has already been implanted. The tool will have a larger gear, which will lock into the gears on the tiny screws, and rotation can begin. As it rotates, all of the gears on the head of each tiny screw will be simultaneously drilled into the bone. Figure 5 shows a top view of the actual device. On top of the screw head, there will be the four tiny screws, which are held in place by extensions coming from the head of the interference screw. The tiny screws will move through these extensions into the bone by the unique gear mechanism. Figure 6 shows a side view of the head of the interference screw before the tiny screws have been set in place. The design of the extensions includes filleted edges, so that there are no jagged edges to causing any abrasions or other harmful side-effects. 12

13 Figure 4: Gear mechanism Figure 5: Top view of interference screw Figure 6: Side view the head of interference screw (without tiny screws) 13

14 2.2.3 Alternative Design 3 As with the other two designs, it is our desire that the patellar tendon will be the graft used for this design since we believe that the tendon graft with bone blocks attached will result in the optimal recovery of the joint and joint function. This design will prevent two flaws found in current fixation devices: excess graft stiffness and graft tunnel motion. Interference screws are most commonly used as fixation devices. These screws often result in the graft becoming too stiff. If the graft is too stiff it may develop micro tears which may result in a rupture of the graft. Graft tunnel motion occurs when the forces acting on the graft cause it to move longitudinally and sagittally which can be detrimental to the recovery process. This design will consist of two halves of a hollow cylinder divided longitudinally as seen in Figure 7 below. The outer surface of the cylinder will be threaded so as to be screwed into the graft tunnel. The two halves of the cylinder will be placed around the graft with the bone block sitting on top of the cylinder once the two halves are joined. In order to join the two halves of the cylinder, pegs on the joining surfaces of one half will fit into holes on the joining surface of the opposing half. The diameter of the bone block must be greater than that of the hole in the cylinder so as to prevent graft retraction through the tunnel when a force is applied. The bottom of the cylinder will contain slots for the placement device to fit into so that it can be screwed into place. Figure 7. Assembled device with graft and bone-block Once in place, the threads on the cylinder will be what hold the device in place. With threads along the length of the device, there should be no longitudinal movement in the graft 14

15 tunnel. The diameter of the tunnel should be slightly smaller than that of the graft so that the graft fits snugly but not too tightly as to prevent any stretch. This will limit the longitudinal graft motion and entirely prevent sagittal graft movement. Since there will be limited longitudinal graft motion, the graft will maintain the optimal amount of stiffness. This will prevent damage to the graft, ensuring a better recovery. The bone block on top of the device will regrow in place and the novel polysaccharide material will degrade over time leaving only the graft. 2.3 Choice of Optimal Design The three alternative designs all had some advantages and disadvantages making the optimal design difficult to determine. Alternative Design 2 incorporated four small screws on top of the interference screw that would insert into the surrounding bone after the device is placed into the femur. We decided that this design was implausible as the four screws would be very small and difficult to manufacture. Also at the necessary size they might not provide the appropriate mechanical properties for this application, and may degrade too quickly to sufficiently secure the screw in place over long periods of time. Alternative Design 3 was also deemed implausible for this application. For this design, the graft would be placed between two adjoining halves of a cylinder with threads on the outside surface. The bone block would sit on top of the cylinder and the device would be screwed into the bone tunnel. We realized that this design was not ideal since the only thing securing the graft to the fixation device was the difference in diameter between the bone block and the hollow cylinder. With enough force, the bone block could be pulled through the hollow cylinder which would require further reconstructive surgery. We originally agreed upon Alternative Design 1 but through preliminary ANSYS testing and professional advice from our client, Dr. Kumbar, we recognized the need for a few modifications to the original design. Our stabilization rail design was implausible as the rails would be too small to withstand the forces during placement as well as during normal knee motion post-surgery. As you can see from the figure below, the material was also not strong enough to accommodate such elaborate features, and there wasn t a large enough surface area in contact with surrounding bone to provide stability. The stabilization rails and the housing tunnels experienced significant stress when a moment was applied, which exceeded the mechanical properties of our biodegradable material by nearly four times. 15

16 Figure 8: ANSYS simulation setup with applied forces and moments for alternative design 1, simulating the types of loading that occur in normal knee motion. Figure 9: Stress results of simulation in Figure 8 for alternative design 1. Another major design change was the elimination of the locking pin mechanism which we found through ANSYS testing would not withstand the maximum loading of 2000N that an ACL can experience. Under loading (without moments) of 1000N, it was found that the locking pin was the weakest component in the design that would experience the majority of the stresses in the knee. While the failure stress experienced was similar to that of other devices, it also did not satisfy our goal of withstanding 2000N of loading with moments. 16

17 Figure 10: ANSYS simulation setup with only forces applied to device on left, with its stress results on the right. Figure 11: Deformation Results of simulation shown in Figure 10, where maximum deformation occurs at the center of the locking pin as expected. As a result, this is the weakest point in our alternative design where it would fail if manufactured. Since the locking pin did not provide sufficient stability to support the loading experienced by a natural tendon, we instead designed a u-shaped tunnel in the bottom of the casing which the graft tendon would be threaded through (which can be seen in the following Optimal Design section). Since the stabilization rails did not provide a large enough surface area 17

18 to disperse the stress and torque experienced by the tendon, we created a threaded exterior surface which will improve the degree of fixation and stability of the device after implantation. A large benefit of this new design is that we can now incorporate a hamstring tendon instead of a patellar bone-tendon-bone graft previously considered. It was found through our client and recent research that the patellar bone block tendon isn t preferable. It was found that in allograft bone block grafts (the most commonly used graft), the graft most often failure mid-graft at the intersection between bone and tendon [12]. This isn t ideal for our device, and as a result the hamstring tendon was selected due to its consistency and strength. Lastly, we designed a hexagonal drive in the top of the casing with which the surgeon will use to place the device into the graft tunnel. This device will provide both flexibility and mechanical integrity, so that the graft tendon will have time to heal during the recovery process. It combines multiple features from the three alternative designs, but also simplifies the design significantly. This simple yet innovative design will be stronger and easier to use. Section 2.6 of the report shows the mechanical advantages of the device through finite element analysis. With this design, we will have to consider some constraints. Size and mechanical strength are very important constraints, as the device must be able to fit into the bone graft tunnel in the femur, and it has to withstand the forces and moments produced by the ACL. Dealing with economics and manufacturability, the device has more volume than most existing devices and will therefore require more material. Therefore, we need to find a way to mass produce our material with low cost. Health and safety is always an issue with fixation devices, as we have to be mainly concerned that the patient does not suffer any serious side effects due to our device. These constraints will be discussed in more detail in Section 3 of the report. 2.4 Optimal Design Objective This fixation device provides the surgeon with a novel technique for performing ACL reconstruction surgery, while maintaining the main surgical methods to limit the amount of training a surgeon would have to complete. In order to make our design effective, devices were needed to fix the graft tendon on both the tibial and femoral sides of the knee. The novel geometry of the tendon casing makes it perfectly appropriate to fix the tendon on the femoral side of the knee. It will provide both flexibility and rigidity that the tendon needs in order to grow to become a fully functional tendon. In addition, it will be placed first during surgery, which mimics current techniques for the Endobutton. The interference screw, which is commonly used in this type of ACL reconstruction procedure, is a sufficient way to fixate the tendon on the tibial side of the knee. It had been found that the majority of reconstruction surgeries fail at the femoral site compared to the tibial site for autografts [13], so we focused on creating the novel design for the femur. Together, these two components will serve as a unique fixation device, made from a novel polysaccharide-derived material, to allow for a graft tendon to replace an ACL after it has been injured. 18

19 Now, to clearly understand the objective of this optimal design, it is essential to know more about how the surgery will be performed. First, part of the hamstring tendon is harvested from the patient. This will serve as the graft tendon, and it is known as an autograft. Next, a graft tunnel must be drilled, and it will be over-drilled from the femoral cortex. The diameter of the Tendon Casing (minus the thread size) will share the diameter of the over-drilled tunnel. This, as well as the threads, will prevent longitudinal translation back through the graft tunnel during normal knee movement. The overall shape of this component will be cylindrical and it will contain a hollowed out semi-circular tunnel, called the Tendon Tunnel. This provides a cavity for the graft tendon to be threaded through. Once the tendon is threaded through the Tendon Tunnel, the surgeon will pull the graft tendon through the graft tunnel, followed by using the hexagonal drive to insert the Tendon Casing on the femoral side of the knee. Lastly, the interference screw will be used to lock the graft tendon into place on the tibial side of the knee. A unique characteristic of both the Tendon Casing and the interference screw is that they will be composed of the novel biodegradable material. Because of this, it is expected that the patient will undergo a shorter period of rehabilitation in comparison to other fixation devices currently on the market. The diagrams seen in Figure 12 and 13 below show the entire process. Incisions in knee made for arthrocopic equipment and drills Using curved needle, thread graft through u-shaped tunnel in "tendon casing" device (#1) Pulling both ends of threads through femoral tunnel and tibial tunnel, screw in "tendon casing" device with thread into larger femoral tunnel with hex drive (#2) Hamstring graft is harvested from patient, approxiamately 2mm in diameter Tie 5/8 surgeon needle and thread to one end of tendon graft, and tie only surgical thread to other end While holding graft in tibial tunnel under tension, use guide needle to place interference screw, securing tendon in tibial tunnel (#3) Tendon is doubled upon itself, and free ends are sutured together Larger tunnel that is 16mm in diameter will be drilled 18mm deep through femoral cortex, centered at 10mm drill tunnel just created In exposed bone cavities, clean area and fill with bone cement to restore original bone geometry and reduce chance of infection. (#4) Knee cavity prepared by clearing debris, assesing damage, and smoothing ACL attachment surface Tunnel is drilled through tibia and femur at anatomical angles with 10mm drill bit Surgical sites are close and patient undergoes rehabililtation to strengthen knee joint and promote bone ingrowth Figure 12. Block diagram of implementation of our fixation device. 19

20 Figure 13. Visual schematic of flow diagram shown in Figure 12, demonstrating the final placement of the device during surgery. 20

21 2.5 Subunits The surgical system created for this design project consists of three components. The first component is a cylindrical unit called the Tendon Casing. The second is an interference screw based off of previous designs that are typically used in ACL reconstruction surgery. The third is the graft tendon, which will be harvested from the hamstring during surgery. These are the three main subunits that will be used to reconstruct the ACL. However, an additional request made by the client required the designing of a molding tool for the interference screw. This molding tool consisted of four parts including two main cavities, as well as two separate parts to create the hexagonal drive and the cannula. Further discussion of these components will occur in the following sections. After discussing each component individually, a diagram of how all of the parts fit together after the surgical procedure is shown Hamstring Tendon Casing Our novel fixation device for the femoral attachment site, the hamstring tendon casing, will serve as a more effective alternative to other methods like the Endobutton and femoral interference screws. As previously touched upon, there exists a discrepancy in operation success rates for reconstruction surgeries. While many of the surgeries are successful for patients, there are still the 10%-15% of patients who have continued discomfort due to over fixation or underfixation. Our device allows for a repeatable operation, where the tendon itself strikes a balance between flexibility and stability. This is created through our threaded suspension system that allows for small amounts of movement in the graft during loading, while restricting movement of the device itself entirely. The device itself is 17mm tall and 15mm in diameter, where the threads extend 1mm on each side (total of 17mm diameter thread tip to thread tip). The schematic detailing the dimensions of the device can be seen below: 21

22 Figure 14. SolidWorks drawing of device showing detailed dimensions and different views. There are three main features in the tendon casing design that were carefully selected throughout the year in order to optimize different fixation properties. The first feature is the threaded exterior of the casing. Similar to the screw, the threaded exterior allows for simple application as well as increased stability. Since we want the graft to remain secure in the drilled tunnel and not slip during physical activity, the device needed to guarantee limited translational motion along that tunnel. Using the threaded design will create a large surface area upon which the loading can act so that it does not twist or slide along the tunnel, entirely fixing it in place. Using a 16mm wide tunnel with 17mm threads will guarantee contact between the device and bone, which is important for osteoinduction and device viability. 22

23 Figure 15. Three-dimensional SolidWorks image showing the thread design on the tendon casing. The second feature of the tendon casing is the u-shaped hamstring tunnel on the bottom of the device. As detailed earlier in the surgical procedure schematic, the hamstring tendon will be harvested and doubled upon itself, creating a 4mm thick tendon that can be threaded through this u-shaped tunnel. This will be done with a 5/8 circular surgeon needle that fits the specific tendon casing size used. Once the tendon is threaded through and the two ends of the graft are equal length, the two free ends extending out of the device will be sutured together, creating a complete system. The benefit of suspending the tendon in this device is that it simulates the beneficial design feature of the Endobutton with the stability of the interference screw. While the device itself is locked into place through the threaded design, this suspension feature allows for small movements necessary during physical motion in the knee, otherwise the patient may complain of unwanted knee stiffness. The figure below shows the u-shaped tendon tunnel on the bottom of the device. 23

24 Figure 16. On the inside of the tendon casing device, a hollow tunnel for the hamstring was created to simulate the preferred suspension orientation of the hamstring tendon. The final feature of the tendon casing is the hexagonal drive on the top of the device. This was added for easy insertion by the surgeon. First, the surgeon already uses similar tools when inserting the interference screws, so the learning curve to accommodate this device will be limited. Second, the hex socket is a standard ¼ inch size, which can be replicated by any biomedical device company when creating a hand tool. The hex design can be seen in the follow figure: Figure 17. Hexagonal drive for easy insertion on the top of the device. 24

25 2.5.2 Interference Screw The second component of our design was the creation of an interference screw that could be used for testing and was easy to manufacture in large quantities. Interference screws, both plastic and metal, are very common forms of fixation for hamstring tendons and usually only cause issues when used on the femoral side. Our tibial interference screw will be similar dimensions to other screws, but will have a different thread design. This is a result of the manufacturing process that our client wanted to use when creating these screws. In order to create these screws, he needs them to be compression molded with heat. This requires not only an effective screw design, but also a design for a mold (seen in section below). The material used for the screw will be the novel polysaccharide-derived material developed at the University of Connecticut Health Center, which ideally has similar mechanical properties and increased osteoinductivity that current materials used in ACL reconstruction. Figure 18. SolidWorks drawing of the screw and its dimensions for tibial application. Eventually this product will be manufactured through injection molding for mass production for the market. 25

26 For both compression molding and injection molding, the thread design needs to be altered so that the screws can easily be ejected from the two part mold requested by our client. Figure 19. Solidworks model of screw showing thread design. As you can see in the following images, screw is also cannulated which allows for a surgical thread to be passed through and act as a guide wire during surgeries. This is a staple feature of all interference screws and was therefore important to incorporate into our design. It was decided to make the cannula with a 1.8mm diameter, since this wide of a cavity could accommodate most surgical sutures [14]. Also seen is a similar hexagonal drive on the top of the screw, which will be used with a hex drive handtool to insert. Figure 20. Blue regions in SolidWorks model show the hexagonal drive used for placement. 26

27 In addition, the threads have notches along the separating center line where the mold would separate (shown in blue in the figure below), which eliminates undercuts and allows for easy removal. If these undercuts were not corrected for, the threads would be stripped as they are removed from the mold, affecting the reproducibility and validity of any mechanical testing done to the screws. Most tibial interference screws are at most 30mm and at least 8mm in diameter, which we selected for our device dimensions. The reason for selecting the worst case scenario (weakest design scenario therefore) was to help our client produce mechanical data that can serve as the baseline. Depending on the patient, the drilled tunnel and other anatomical factors, different size interference screws are used to accommodate each specific surgical case. As a result, it will be useful to understand how the weakest combination of dimensions would react to the stresses in the knee. Figure 21. Blue regions in SolidWorks model show notches made in threads that correct the issue of undercuts for injection and compression molding design Hamstring Tendon Graft The final component of the design used in the operating room is the hamstring tendon itself. As mentioned beforehand, the tendon harvested will be about 2mm in diameter, which is relatively constant among all ACL reconstruction procedures. Ideally, an autograft will be used because of the increased biocompatibility and integration into the surrounding bone, therefore reducing risk of future operations. When the graft is fully harvested, it will be doubled upon itself to create a ~12mm long and 4mm thick double-stranded tendon that will be threaded through the tendon casing. Once threaded through, it will again be sutured, creating a quadruplestranded tendon graft that is a similar diameter to the native ACL. By taking a longer and 27

28 narrower section of the hamstring, this will reduce injury and limit the time needed to heal by this large muscle group [15]. Figure 22. Diagram showing the procedure for creating the tendon strand [23]. Figure 23. Diagram showing how tendon is harvested (Source: Immersion Media IMSports) 28

29 2.5.4 Interference Screw Mold As mentioned, our client required the creation of a mold that he could use to manufacture these screws from his polysaccharide-derived material that he synthesizes in his lab. Not only is the material expensive, but it is also difficult to make in large quantities at this moment in time. By creating this mold prototype design, we were able to provide him with a model that he could use to mass produce screws when the material is available in his lab. Unfortunately, the mold design is a lot more complicated than current molds he has created on his own, so it was necessary to think of an alternative method for compression molding these screws. The difficulty arose from the cannulation of the screws as well as the unique thread design (which couldn t be created through a threading machine in the machine shop). There will be four components to the device which can be easily assemble and disassembled. As you can see in the following figure, we were able to create an effective design that could be continuously reused in our client s laboratory: Figure 24. Exploded SolidWorks model of the four mold components assembled. The main components are the top and bottom half of the mold which define the outside thread shape and the head for the screw. When assembling these two parts, it was necessary to take into consideration any undercuts in the design. Since they were present in this screw design, 29

30 it was required to add these notches into the mold so that the screw could be easily displaced from the mold. Figure 25. SolidWorks diagram depicting the two main halves of the mold assembly. In order to create the cannula, we need to have a multi-piece mold with components in the center of the screw. Since we need to have a uniform cylinder through the center of the screw, we need to place a metal rod along the center of the mold to displace the material during the melting and molding process. Once the two halves of the mold are assembled as seen above (via the spokes extruding from the inner mold surface), the blue shaded component in Figure 25 is inserted. Once this has been done, the powdered biodegradable material will be inserted into the open end of the mold as seen by the red arrow. 30

31 Figure 26. SolidWorks diagram showing the second component being assembled into the mold (blue component). Once this has been completed, the powdered material will be packed into the screw cavity through the opening indicated by the red arrow. Once these three components are assembled and the material has filled the cavity entirely, the four component with hexagonal screw drive will be inserted from the open end, compacting the powdered material and locking with the cannula mold component from the other side. The final set up is the completed mold that will be used in the heat compression molding process. Figure 27. Blue component with hexagonal tip shown in final assembly. 31

32 2.6 Optimal Design Analysis Hamstring Tendon Casing To analyze the performance of our device, finite element analysis with ANSYS was performed to better understand the device performance under loading in the knee. The first test was to ensure that the device would be able to withstand the forces and moments experienced during knee motion (at the point of failure). In the following figure, you can see the typical motions of the knee, where internal tibial rotation (which causes a combination of large forces and moments that act on the ACL) is the source of ACL failure. Figure 28: Two-dimensional drawing of the knee joint experienced moment [16]. As mentioned, the undamaged ACL can withstand approximately 2000N of loading during knee motion, including the additional moments added during knee rotation. As you can see in the following figures, we applied 2000N to the simulated graft in the tendon casing in order to understand the response of the casing itself. Figure 29 and 30 show the setup of the simulation and the stress results, respectively. The threads on the device (shown in purple) were the portion of the device that was fixed, and the red tendon and arrows show the inferior loading that was applied to each face of the tendon. The reasoning behind fixing the threads is that their surface area will provide resistance to inferior motion, and the device will essentially be fixed during loading due to the threads integration into the bone. 32

33 Figure 29. Experimental setup in Solidworks for the tendon casing model, showing the total of 2000N and 10Nm of torque applied to the tendon graft interface. Figure 30. Equivalent Stress results of tendon casing test in ANSYS. The maximum stress experienced by the casing/tendon system was on the tendon itself, where the maximum stress on the device itself was MPa, which is below the material failure strength. 33

34 As you can see in the figures above, it was found that our device was able to withstand a similar loading that is experienced by the natural ACL itself. In order to determine the strength and reliability of our device, we applied a total of 2000N and 10Nm of torque to the attached tendon to simulate the loading in the knee. We found that the maximal loading experienced by the tendon casing itself is 4.447e7 Pa which is right in the range of our material s mechanical properties (average failure loading between 1.9e7Pa and 5.1e7 Pa). As a result, we are very satisfied with both the material s mechanical strength and the design of our device, since the combination of both are able to withstand similar loading as a natural ACL in the knee. The issue with current devices is that they don t usually exceed a mechanical strength of 600N postsurgery, so our device far exceeds the performance of current devices on the market. When other companies tested their devices, they only included the loading forces without an added moment component. In our study, we included a moment of 10Nm since we believe there will also be rotation of the ligament during normal knee motion, further strengthening our results. Had this moment not been included, we could have applied an even larger isolated loading force as the other companies had done. Lastly, the strength of the hexagonal drive on the top of the device was tested. When a moment was applied, it was verified that the device was able to withstand the torque applied by the surgeon during placement. During many surgeries that use interference screw, a major design flaw is that the threads are either stripped or the screw breaks upon placement. In the following figure, the results of the torsional placement test can be seen. In conclusion, it was found that the torsional stress applied to the device will not negatively affect the mechanical properties of the material and is well below the threshold of failure. 34

35 Figure 31. Experimental setup in SolidWorks for placement torque on tendon casing, where 6Nm moment was applied to hexagonal drive and the bottom side of the threads were fixed. Figure 32. Equivalent stress results of tendon casing under 6Nm of torque, which was found to be the failure loading for many interference screws during placement. Our device had isolated regions with a maximum stress of 1.8e7Pa, where the majority of the device hexagonal drive was between 1.45e7Pa and 1.66e7Pa. These numbers are below the material s yield strength of 1.9e7Pa 5.1e7Pa. 35

36 2.6.2 Interference Screw Testing the screw was more difficult due to the complex thread design, so a simulated bone tunnel was created with a pre-threaded interior in order to test the screw in SolidWorks. When the device was tested with individual parts fixed (ie. the threads or tip of the screw), it did not provide logical results. When a screw is being placed during surgery, it will be rotate as the screw is placed with an applied torque. In order to better simulate the function of the screw in vivo, we applied a 6Nm moment to the top of the screw (at the hexagonal drive), and created a frictional surface between the screw and the simulated outside bone. As you can see in the following figure, the screw s thread shape matches with the threaded interior of the simulated bone tunnel. Figure 33. Solidworks models of screw in threaded bone tunnel, simulating frictional interface between screw and bone as it is inserted into the knee. 36

37 Figure 34. Experimental setup in SolidWorks of screw during insertion. Figure 35. Stress results of the screw insertion test. The majority of the screw head experienced a stress around 7.9e7Pa, which is on the upper end of our material s yield strength. 37

38 Since the 6Nm applied was higher than the forces experienced by screws tested by Smith and Nephew, we believe that this worst case scenario test shows the robustness of the screw. Intuitively it will break near the top where the moment is applied, since the entire bottom half of the already inserted screw is held in place by the frictional forces with the surrounding bone. 2.7 Prototype Since our surgical procedure incorporated multiple parts, three different prototypes were created: the tendon casing, the interference screw and the interference screw mold. Later in development, the tendon casing and interference screw will be made from the novel polysaccharide-derived biomaterial developed at the University of Connecticut Health Center. Due to financial constraints and limited quantity of material from the Health Center, the prototypes were created from plastics (specifically poly-lactic acid or PLA) instead. PLA is a very common material used in plastic three-dimensional printing, and has similar mechanical properties to the material we are using [22]. Not only is it biodegradable and commonly used for bone fixation, but it also has a similar tensile strength as our material. When PLA is used in three-dimensional printing, it has been found that the resulting prototypes have a tensile strength of 17 MPa [17] compared to our client s material s range of 19 MPa 51 MPa. As a result, it was an ideal, affordable material to create the first prototype for testing and demonstrations Tendon Casing Prototype After going through series of revisions and tests with our surgical model, our final tendon casing prototype can be seen in the image below in Figure 36. The device was three-dimensional printed from PLA at the UConn Health Center. The device below has maximal dimensions that are 17mm high and 15mm in diameter, with a tendon cavity that is 4mm wide. All the dimensions can be seen in the subunits section in the SolidWorks drawings in Figure

39 Figure 36. The left image shows the saggital view of the device, where the hexagonal shaped cavity on the top is where the hand tool would be inserted. The right image shows the bottom of the device, with the two 4mm openings for the tendon graft. Figure 37. Image comparing the tendon casing to a quarter to provide a size reference As previously described in detail, this device will be incorporated into the femur in order to secure the superior end of the tendon. When the tendon is threaded through the cavities in the bottom of the device as shown in figure 36, the total diameter of the tendon graft will be approximately 8mm which is the typical diameter of an intact ACL [3]. It is important to note that all devices used in ACL surgery are manufactured in a wide range of sizes in order to accommodate specific patient s knee dimensions. Due to the fact that some patients may have 39

40 larger/smaller femoral condyles or wider/thinner tibias, our device will also be manufactured in different sizes depending on the diameter of the drill bit used. While the overall diameter/height can be decreased for smaller patients (since our prototype would be used for a fully grown male), the width of the u-shaped tendon cavity on the bottom of the device will remain virtually unaltered. The reasoning is that tendons harvested from the hamstring tend to remain equivalent in diameter, so just the drilled tunnel size in the femur and tibia themselves will change to accommodate patient size Interference Screw Prototype The interference screw was designed similarly so that multiple sizes could be created. After researching common dimensions for tibial interference screws, it was found that the diameter is between 8-10mm and the length is around 30mm. The difference in diameter also accommodates variations in procedures and patients, just as it did with our tendon casing. In this case, the interference screw was designed with the smallest diameter of 8mm and a length of 30mm to simulate the worst case scenario for testing. Since our client will need to invest time and money to test these screws with his novel material, he requested we use these worst case scenario dimensions. By understanding the weakest combination of screw dimensions, this will strengthen his medical device proposal and make marketing the screw easier. As you can see in the following images, the interference screw was also constructed from the same PLA material as the tendon casing device. Due to the limited amount of novel material developed by our client, this PLA model will provide a good baseline for his future marketing. While the strength of PLA is similar to his material, it is more brittle and can degrade over time if not stored properly. Luckily, it is simple and inexpensive to three-dimensional print large quantities of these screws, and the facility is located at the Health Center as well. 40

41 Figure 38. Image of PLA prototype of interference screw and rounded thread shape. Figure 39. On top, an image comparing the screw to a quarter as a reference. On bottom, an image showing the cannula and hexagonal driver in the screw head. 41

42 Once the tendon casing has been put in place and the tendon graft has been pulled through both tunnels via the attached surgical threads, the interference screw is then placed. As you will see in detail in the following section 2.7.3, the tendon is placed under constant tension as the screw is placed. Since the screw is small and difficult to manage for the surgeon, it is guided into the correct location via a guide pin. The hex screwdriver also follows this guide pin to easily compress the tendon between the screw and surrounding bone. The dulled threads provide a large surface area on which the tendon graft can attach to, while limiting the amount of damage done to the graft itself. While the cannulated center of the screw is mainly used for placement purposes, it will also increase the surface area on which natural bone can grow and degrade the screw Prototype Operation A large constraint for the design was to ensure that it integrated current surgical procedures into the use of the new device. Unless the device easily accommodates long standing practices for ACL reconstruction, it won t be as readily accepted by the current market and the surgeons that will use it. During a traditional ACL reconstruction surgery, tunnels would be drilled through the femur and tibia. First, the tendon would be pulled through the tibia then into the femor, and the femoral end would be secured with either an Endobutton or an interference screw. Next, the other end of the tendon is pulled back through the tibia putting it under tension, and a second interference screw is applied from the inferior opening in the tibia to secure both ends of the tendon. This novel device uses a similar procedure order, with only slight modifications that would accommodate our device. The first step in the procedure is to harvest the tendon. The patient is usually placed in A4 position and a small incision is made near the knee. The distal part of the hamstring tendon is identified by the surgeon, and the tendon is stripped proximally under tension with a blunted tendon stripper [18]. The next step after harvesting the graft tendon is to drill the appropriate tunnels in the tibia and femur. The starting position and technique for drilling can significantly affect the length and angle of the tunnel, and should be taken into consideration during surgery. For our device, we recommend and drilling the appropriate tunnels is to thread it through the unique fixation device s hollow tube. This can be accomplished with the use of a surgical needle and sutures, if need. This process can be seen in the Figure below. 42

43 Figure 40. Graft tendon (rubber band) threaded through fixation device. Next the fixation device with tendon should be placed in the hole drilled in the femoral cortex and screwed in using a hex drive screw driver, making sure that the tendon has passed into the thinner diameter of the tunnel so that it can be accessed from the top of the tibia in the knee joint. This can be seen in the figure below. Figure 41. Fixation device placed in the femoral tunnel with the tendon between the condyles of the tibia and femur. 43

44 Once this step has been completed the graft tendon must be threaded through the tibial graft tunnel. Then the knee should be flexed completely and the tendon pulled to the proper tension. The knee should be flexed to a 45 degree angle for tensioning, where the graft will be pulled under an initial tension of 22N [19]. Once at this proper tension, the interference screw can be placed in the tunnel pinching the graft in place as seen in the figure below. Figure 42. Interference screw fixing the graft tendon in place on the tibial side Interference Screw Mold Prototype Our client requested not only a design for the tendon casing and interference screw, but also a mold that he could use to mass produce the interference screws for testing in his lab. In order to carry out a longitudinal study of this screw s strength and reliability compared to other devices, he will need the mold that was created to manufacture numerous identical screws. Not only will this give him reproducible prototypes to test, but it will also reduce the cost by creating the 44

45 screws in his own lab (opposed to sending off the design to a company to manufacture them). The material that he has synthesized comes in a powdered and solid form, where the powdered form is the easiest to work with. As a result, heat and compression will be applied to the mold in order to shape the material into the necessary geometry. As you can see below, we were able to manufacture a mold through stereolithography (SLA) in order to provide a visual representation of our idea to the client. Ideally in the future it will be constructed out of steel, but due to the complex nature it would have to be shipped to a metal three-dimensional printing company and would be outside our price range. Figure 43. Four parts of the SLA mold. The two main components on the left define the thread shape, where the connected part on the right defines the cannula and hexagonal cavity in screw head. 45

46 Figure 44. Assembled mold, where the spokes sticking out simply provide a handle to disassemble the mold after heating and cooling Prototype Testing with Client The main testing was done with the mold itself by ourselves and our client, since this process of heat compression molding is critical to manufacturing identical screws trial after trial. Our client needed to have an understanding about how the mold would work and was procedure to follow when creating the screws on his own. As you can see in the previous section which described the mold subunit, we also went through these same steps in order to ensure reproducibility. First the two halves of the mold are assembled, and the smaller cannula component is inserted from the bottom as seen in Figure 26. Then the powdered material is loaded from the top of the mold opening, making sure that the mold is tapped throughout the process to eliminate air bubbles. Once it is filled to the top of the screw head, the final hexagonal mold component is pressed into place. This is then placed flat (with two larger components on top of one another) in the heat compression plate. There is both a temperature gauge and pressure gauge on the compression system. When applying pressure, we found that minimal pressure needed to be applied in order to hold the components together, where too much pressure deformed and damaged the mold s shape. Eventually this mold prototype will be constructed out of steel in order to maximize heat transfer, and will not be the polymeric material seen in the image below. We found that plastic material doesn t conduct heat well enough, and only the lower part of the screw closest to the would be melted, where the material farthest away would still be powder. Once the temperature was steadily raised to 500C, it was allowed to sit for a half hour in order to ensure maximal melting. Afterwards, the heating plate that lays underneath the mold was turned off and allowed 46

47 to cool. The pressure was kept on the device throughout the extensive cooling process in order to prevent mold component separation. Usually leaving the system overnight ensured that it cooled entirely so it could be safely removed from the plate the following day. Once complete, the mold is disassembled in reverse order, ensuring the two larger components of the mold are separated last so that the screw would not break during separation. If all the powder melted and the procedure was followed correctly, the screw would be intact and could be stored in the refrigerator to prevent degradation. The image below shows the heat compression molding system with one of the plastic mold prototypes. Figure 45. Demonstrating use of molding technique with powdered material and heat/compression plate setup. 3. Realistic Constraints The most important constraints in the design deal with size and mechanical issues. First of all, the ultimate goal of the device is to implant it into the knee and hold the ACL graft in place. Because arthroscopic surgery must be performed to place the implant and graft through the femur and tibia, the goal is to make the device as small as possible, so that the surgery involves drilling away the smallest amount of natural bone necessary. Therefore, our device is constrained to a maximal size of 17mm tall and 15 mm wide. Because of these size constraints, our device could not have elaborate features, or else their mechanical strength will not be sufficient. So, the ultimate goal for the geometry of our device is to have a simple and small structure that will not require a large amount of space in the bone, so that the procedure is minimally invasive and the mechanical integrity of the features are sufficient. Besides size constraints, the mechanical and biocompatibility constraints are also an issue to be considered. Properties such as elastic modulus, compressive modulus, compressive strength and maximum compressive load must be similar to those of human bone. In addition, 47