Ankle Positioning and Loading Device
December 4, 2013
MAE 435 – Project Design and Management II
Advisors: Dr. Stacie Ringleb and Dr. Sebastian Bawab
Students: Vasile Grigorita, Dominick Hudson, Jin Hyuk Kim, Bryan Mazac, Christopher Parrish, and Christopher Villaire
Table of Contents
List of Figures
List of Tables
Abstract
Introduction
Background
Methods
Results
Discussion
Conclusion
Appendix A – Layered Design
Appendix B – Inventory Lists
Appendix C – Algorithm and Flow Chart
References
List of Figures
Figure 1- Anatomy of the Ankle
Figure 2 - 1/3 Scale Plastic Model of Previous Team's Design
Figure 3- Before (left) and After (right) of the Transition Tiers
Figure 4 – Midterm Report Design
Figure 5 - Blueprint Example
Figure 6 - New Base Plate Design
Figure 7 - New Stabilizer Triangle Utilizing Fasteners
Figure 8 - Finalized Design
Figure 9 - Left: Ardunio Uno Right: Accelerometer with Gyro & Breadboard
Figure 10 – Setup of the Motion Control
Figure 11 - Algorithm for Servomotors to Move at Regular Intervals
Figure 12 - Control Interface
Figure 13 - Final Product
List of Tables
Table 1 - Input Values and Output Angles
Abstract
Even a simple anatomical analysis of the ankle region illustrates the complexity of replicating it with mechanical systems. Modern day ankle positioning devices are expensive, cumbersome and only analyze the ankle for instability while under no static loading. Since the natural human ankle frequently encounters instability due to loading from body weight the aim of the experiment emphasized replicating ankle motions in the latter case. By analyzing and improving upon the designs of previous teams, such as by utilizing fasteners and U-channels, a finalized normal scale device was designed and assembled to load a cadaver foot. Choice of materials and design was influenced by factors such as cost, cyclic loading fatigue, ease of use, practical applications, and machinability.By the conclusion of the semester, the device was assembled and able to move in five degrees of freedom while under a static load.
Introduction
An ankle sprain is an extremely common injury that can happen to individuals of all ages.Even just a simple task such as an individual walking down the stairs can cause a particularly serious ankle injury. In addition, some individuals can possess a bone disorder, which causes their ankle muscles to not function properly. Once the ankle is injured, it is critical for one to not aggravate the injury by placing unnecessary pressure on the bone and muscles. However, when a physician needs to examine the ankle it is important for him/her to be able to place a load on the ankle to see where exactly the ankle is injured. In addition, it may tell the physician if it is a bone fracture or ligament tear. Therefore, the purpose of this project was to design a device that can load and support a cadaver ankle while analyzing the forces in a closed kinetic chain.
Background
In order to design a device where a controlled load must be placed on the ankle and be confined to a particular range of motion, it is of extreme importance for one to understand the anatomy of the ankle.The ankle consists of three joints; the true ankle joint, the subtalar joint, and the transverse tarsal joint [1]. The main function of the true ankle joint is to absorb shock as well as to move the foot up and down [2]. In addition, the true ankle joint consists of the tibia, fibula, and talus [1]. Figure 1 displays the distinct parts of the ankle.
Figure 1- Anatomy of the Ankle
The tibia, known commonly as the shinbone, is the bone that extends from one’s knee down to the ankle. The tibia bone is one of the major weight bearing bones of the body [3]. The fibula lays adjacent to the tibia and is of comparable length to the tibia. The fibula too extends from the knee down to the ankle. However, unlike the tibia, it is not a major location for transmission of weight [3]. Lastly, the talus is located between the leg and foot where it acts as the bridge joining them [4]. The head of the talus connects to the tibia and fibula [4]. The tibia, fibula, talus, and ligaments all act in conjunction to stabilize and allow the true ankle joint to function. The region of the tibia and fibula near the ankle joint contains three distinct features. These three regions are the posterior malleolus, the back part of the tibia, the medial malleolus, the inside part of the tibia, and the lateral malleolus, the end of the fibula [1].
The subtalar joints main function is to allow pronation and supination of the ankle [1]. This joint consists of the talus, previously discussed, and the calcaneus, which is the heel bone. The transverse tarsal joints main function is to allow the foot to move from side-to-side [1]. This joint consists of the talus, calcaneus, navicular, and cuboid bones. The navicular bone and cuboid bone are rather minor bones in comparison to the tibia and fibula.
Besides bones fracturing, ligaments in the ankle can tear if the ankle is severely rolled. Ligaments are bands of tissue that connects one bone to another bone, and these ligaments are made of collagen, which make the ligaments extremely strong [5]. The main ligament that is injured when the ankle is rolled is called the lateral ligament [5]. According to the Cambridge Foot and Ankle Clinic, “The ligament on the outside of the ankle (lateral ligament) is made up of three separate bands: one at the front (anterior talo-fibular ligament), one in the middle (calcaneo-fibular ligament) and one at the back (posterior talo-fibular ligament) [5].”
A team, consisting of four members, worked on this project during previous semesters. This team started by researching and understanding the anatomy of the ankle. Their conclusion from this research was that the device needed to be incorporate several critical factors. These factors were, “user-friendly in a clinical setting, design for both isolated and combined motions, design a device which would allow for five DOF and create as simple a design as possible while incorporating all of the aforementioned goals[6].” With the above-mentioned goals in mind, the group then proceeded to design four different prototypes. A design rubric was then used to select the best design of the four. The winning design was the layered design, which can be found in Appendix A.
The group then proceeded to modify the design. The modifications included extra support for the tibia, a new location for the center of rotation and additional degrees of movement. In addition, the improved design allowed adequate space to place motors, bearings, and other critical components. Partswere then chosen that would meet the needs of the device. They settled on servomotors, which could support 210 pounds, which yielded a factor of safety of 1.22, and aluminum plates for the majority of the device. The final designis shown in Figure2.
Figure 2 - 1/3 Scale Plastic Model of Previous Team's Design
A preliminary parts list was then constructed and several of the parts were ordered. They then demonstrated the degree of freedom involving plantarflexion and dorsiflexion by displaying movement of the servomotors in conjunction with user controls. It is the goal this semesterto improve the design and to have a finished product.
Methods
The current team wanted to improve upon the previous design by eliminating possible sources of error when machining and assembling the device. An analysis was conducted on a partially completed tier and a three dimensional 1/3 scale plastic model of the device. One particular area of concern was the abundance of welding that the device required. Welding done on the tier plates was particularly problematic due to issues arising from the height of the weld. The welds could potentially obstruct the pathway of critical components, such as fasteners and bearings. In addition, the welds seemed unsteady. Another area of concern was the spacing as certain locations of the device, which were crucial for assembly and operation, were problematic to reach or obstructed as well. This was particularly evident in the partially assembled tier. For example, the middle tier plate was restricted from its desired range of motion due to a screw from a bearing obstructing its path.
Two methods were investigated to help alleviate spacing issues and welding issues. The first method was to space things out more evenly across the base plate. However, this would have required more material and thus more costs. Since the device was already approximately 3’ x 3’ x 3’ this option was not practical. The second method, which was chosen, was to convert all available components that were constructed of three plates in a U-shape into U-channels, which contain no welds. This option alleviated some of the potential error arising from welds and allowed the costs to remain virtually the same.The transitions tiers were converted to U-channels with dimensions of 12” base x 2” leg length and 8” base x 1.5” leg length. These two changes eliminated four total welds from the assembly. Figure 3 displays a before and after comparison of the device.
Figure 3- Before (left) and After (right) of the Transition Tiers
To help alleviate some of the stability issues, alternate methods were investigated to support the weight plate. One method that was investigated was the use of hexagonal support shafts instead of square shafts. However, this method was determined to be too challenging to machine due to tolerance issues. Instead,all of the stabilizing triangles, located at the base,were madethe same size with dimensions of 4” x 8”. Previously, two of the triangles per each shaft (four total), were smaller in an attempt to save on costs.
Appropriate bearings and shafts to use also had to be determined. One challenge in choosing the bearings and shafts were the material properties of each respectively. Since these two members were going to be in direct contact with one another, indentations could potentially become problematic after extended usage. A decision was reached to first select an appropriate bearing, as they were more limited, and then select a shaft. For selection of the bearings, sleeve bearings were decided upon in lieu of ball bearings. The rationale for selection of sleeve bearings was that these bearings would have limited use, as they were only for vertical translation, and would be operating at a normal temperature range[7]. This is in contrast to the research conducted for ball bearings, which showed that they were more practical for high temperature application, lasted for more loads, and were more expensive. Self-lubricating pillow-block linear sleeve bearings were chosen. These bearings were rated for use with a shaft that had a hardness of at least Rockwell B25 and an 8-16 rms micron finish. A4’ shaft made of 1566 steel was selected, which has a hardness of Rockwell C60 and a 9 rms micron finish. The shaft will be split to accommodate each side of the device.
After making all of the above design decisions, blueprints needed to be fabricated. To do this AutoCAD Inventor (Autodesk, San Rafael, California)was used to adjust the constraints and parameters from the previous groups Autodesk files. This task was more difficult than expected due to an organized, albeit abundance of files from the previous group. Once the correct group of files was located, the appropriate changes were made reflecting the expected parts inventory. The completed inventory list can be found in Appendix B.The midterm report design can be seen in Figure 4.
Figure 4– Midterm Report Design
With the midterm reportdesign completed, blueprints were now ready to be completed. To do this the structure was organized into tiers, as done by the previous team, to help keep the structural blueprints organized. Each tier is one level of the total structure. This was a critical step, as many of the plates look identical, but have slightly different dimensions. By keeping the tier structures, this may eliminate possible errors during the machining phase resulting from poor communication. One example of the many blueprints can be found in Figure 5.
Figure 5 - Blueprint Example
Following the midterm report, changes were made to the design to simplify device and to reflect time constraints. The two tibia clamps, one was on the weight plate and the other one the footplate were removed due to inadequate time to properly configure them. In addition, the length of the weight plate between the initial pressure transducer holes and the rear of the plate was extended in an attempt to balance out the moment that was being caused due to the weight of the plate and foot. This extended length will allow a weight to hang from the rear.
The structure also underwent changes based on the recommendations of the machine shop in developing the best way to align all of the components. They suggested securing all base components with fasteners instead of welds. The end result of their advice is that the device will now utilize ¼” – 20 fasteners to secure the components to the base, secure stabilizer triangles to the shaft, and to assemble the weight plate. These changes can be seen in Figure 6 and Figure 7.
Figure 6 - New Base Plate Design
Figure 7 - New Stabilizer Triangle Utilizing Fasteners
This came about because the machine shop could not guarantee the quality of the machining if imperfections, such as warping from the heat, were present due to welding. By securing the parts with fasteners it will give the capability to add or remove parts that are not aligned correctly and correct them, making troubleshooting significantly easier and faster. To reflect these changes, three new AutoCAD parts needed to be constructed and many more needed to be modified. The finalized design can be seen in Figure 8.
Figure 8 - Finalized Design
The blueprints were then made and submitted to the machine shop.The machine shop took approximately five weeks, due to the high volume of tasks they had received, to complete fabrications of all needed components. This unfortunately caused some scheduling issues, but did allowanother minor modification to be made. Originally, the device was utilizing button head screws for the rail carriages; however, it was realized that this would interfere with the movement of the rollers. To avoid this interference, a switch was made to 10-32 x 1” flat head cap screw with lock washers. This change has allowed for fluid movement of the rollers.
The design for movement of the device, which needed to imitate the ankle’s motion, was created using a servomotor, accelerometer with gyro inertial measurement unit (IMU), Arduino Uno (Ardunio, Italy), and Arduino Software IDE (Arduino, Italy). The hand controller, Figure 9, was made up of the 5-axis IMU and breadboard.
Figure 9 - Left: Ardunio Uno Right: Accelerometer with Gyro & Breadboard
To create the control algorithm, conversion of the IMU's input value from a servomotor through Arduino was needed. Figure 7 explains the input values relative to the output angles.
Input Values / Output Angles0 to 89 / 0 to -180; counter clock wise
90 / 0; neutral position
91 to 180 / 0 to 180; clock wise
Table 1 - Input Values and Output Angles
An attempt was made to use the accelerometer to control servomotors through Arduino Uno and algorithm through the Arduino Software. The physical setup for the control system is shown in Figure 10.
Figure 10 – Setup of the Motion Control
In this case, the accelerometer was the input, and the servomotors were the outputs. First, X axis' values, from the accelerometer, were read through Arduino's analog pin A0. Then the map function was used to set the position of the servomotor. Next, the positions' values were used to control the servomotor. The same method was applied to Y-axis (Arduino's analog pin A1) and Z-axis (Arduino's analog pin A2). Due to the filter, this method produced an abundance of noise and the Z-axis failed to work properly. An attempt was made to remove noise by Kalman filtering the code. The Kalman filter allowed the servomotors to move quickly and smoothly; however, there was some noise still present and the servomotors positions were not accurate. This correction also fixed the Z-axis using the gyro.
Due to the errors encountered with the Arduino software, an attempt was made to code the algorithm using Labview(National Instruments, Truchard, TX) and the Arduino Software in synthesis.The Arduino software would be used to control the servomotors. As can be seen in Figure 11, thealgorithmwas designed to move the servomotors at regular intervals.