Team 2468 VersaSwerve

Transcription

1 Team 2468 VersaSwerve Design Process Designed by Katie Anderson, Jonathan Bunt, Brett Etter, Michael Henderson, Nick Hummel, Sidhu Krishnamoorthi, Philip Liu, Grant Odama, Byron Seaberg Write up by Katie Anderson

2 Project Summary Our objective is to develop a swerve module built primarily from products manufactured and sold by VEX. This is because VEX has asked our robotics team to design such a swerve and has offered their support through supplying us with the parts we need, as well as answering any mechanical and swerve related questions we may have during this process. Our final products should include a reliable and easy to manufacture swerve module and a white paper detailing the design of the module including a basic chassis to house the modules and how to build it, both of which will be posted on the VEX website. Our ultimate goal for this project is to design a swerve module for teams with limited resources, including but not limited to: mentors, machinery, and funds. Project Goals: 1) The module must be simple to manufacture with as few custom parts that require machining as possible. We will assume that teams using our design will have access to a drill press, a mini lathe, and saws. We will also include a mill with the capabilities of the one in our shop as a usable tool, if it is needed. 2) Our design should be easy to follow and simple to assemble. Tools needed for assembly should be limited to basic hand tools. All lubrication should be applied via spraying. Finally, the entire module should be mounted as one unit, so it is either completely attached or completely detached at any one time. 3) The module must be reliable, with little room for failure or error. 4) The module should be able to travel at a speed of 8 12 feet per second. 5) Each module must be able to support about 50 pounds while still being able to spin freely. 6) The module should be capable of multi turn operation. VEX contact: Our contact with VEX throughout this project will be with Mr. Aren Hill, a mechanical engineer at VEX Robotics who has exceptional experience with swerve modules, having designed many himself. We will be consulting him throughout the design project for advice and tips. We will also get all of the VEX parts we need for our swerve module through him. 2

3 Tuesday, June 2, 2015 The Process The purpose of this meeting was to lay out the official groundwork for the project, as well as decide upon the basic design that we wished to use for a swerve module. Summary: We finished discussing our research of existing swerve modules. We then mapped out our exact goals for our module, as well as the limitations we would have to follow. We finished by discussing and selecting the basic design aspects that we wished to follow when designing our module and finally setting the goal to meet with Aren and discuss our design and project within the next week. Research: The first real step towards designing that we took with this project was examining and studying existing swerve modules. (The modules we studied can be found here.) We used these modules to learn about the different design possibilities there are for swerve modules and to compare the pros and cons of different designs. Further research will be conducted by team members as needed. Further project overview: Our purpose is to design a swerve module for teams with few resources and little funds. We are keeping our focus mainly on rookie or inexperienced teams and teams from less well off areas that lack the resources to use one of the swerve modules already in existence. Because of this, the module should be easy to manufacture, machine, and use. The final product of this project will include a functioning swerve module and a white paper detailing the design and construction of the module. Requirements: 1) The module must be primarily constructed out of VEXpro parts. 2) The project must be well documented, so that an easy to follow, detailed white paper discussing the design and how to build it can be written. This white paper will be posted on the VEX website. (See examples at VEXpro.) 3

4 Priorities: After discussing our purpose in relation to what we want to achieve, we have that the ease of machining and manufacturing and the reliability of the swerve module are our top priorities in our design. We will put most of our focus into creating a module that satisfies both of these priorities. We will also keep in mind cost and weight in our design and try to minimize those as much as possible, but not beyond the point where it interferes with the simplicity of manufacturing or the reliability of the swerve. Question for Aren: 1) Can bearing balls be used as a support? 2) How durable are the plastic bearing blocks and the clamp on gear boxes they have? 3) Are there any design priorities that we should add to our list? Design: Design limitations: The module must be capable of multi turn operation. This means that wires cannot connect the wheel module to the static module, as the wires will restrict the range of rotation that the wheel module has. Potentiometers should also be avoided for rotation sensing, as they often have a mechanical stop beyond which they will break, making them inappropriate for multi turn swerves. Potentiometers without this mechanical stop feature are still inappropriate as they have a wide dead band where their output is undefined. Finally, we will try to use an absolute encoder similar to those on the AndyMark swerve modules as our team is already familiar with these encoders. Our design will also have to contain the smallest number of custom parts or parts that the team must machine themselves to promote simplicity in the manufacturing and assembly of our swerve module. Components of the module: To simplify the design process, we have divided the swerve module into two smaller parts: The wheel module houses wheel and will spin. The static module houses the motors and will not move. Additional items: 4

5 We will construct a basic chassis on which to mount our modules, outfit our module with proper electronics, and write sample code with which the module can be operated. Summary of current decisions: 1) We will design the first iteration of our swerve module to utilize a bearing block mount system in the static module. We plan to attach the bearing block so that the drive shaft points down and connects to the wheel module. This will simplify the mounting of the swerve module as well as helping to reduce the cost of our modules, in concurrence with our project goals. However, as this design has not yet been tested, it may not be strong enough for the weight the modules need to bear. We may have to change this design if this or some other problem occurs during testing. 2) We plan use a large washer to distribute weight of the robot, so that it does not crush the bearing block. 3) We plan to use C channel to create the outside of the wheel module. 4) Gearing is our preferred method for driving and steering. However, we may decide to use belts and chains, depending on how easy tensioning will be with our design. 5) We will use 3.5 wheels. Customization of design: Our module will also leave room for customization, including but not limited to: 1) A part compatible with the module to house an encoder, so an encoder can easily be installed. 2) The module will be compatible with a potentiometer, so that a team can easily mount and use one. Bearing Block or Plate design: Bearing Block Premade by vex (no machining needed) Plate Easy to mount, giving it more options Cheaper 5

6 More examples As the Bearing Block is made by VEX, it reduces the amount of custom parts that must be machined. This means that a Bearing Block design better aligns with our design priorities. Thus we chose to use a Bearing Block design. Gears or Sprockets: As gears are more compact and do not require the possibly complex tensioning that chains and belts call for, we would prefer to use a gear design in the interest of simplicity. However, regardless of our final design, teams should be able to modify our design to be operated with chains and bands. Wheel selection: 3.5 wheels Encoder notes: We will build this without encoders, as VEX does not sell them, but we will include a spot to install an encoder Steps from here: 1) Introduce ourselves to Aren and ask him the questions we have listed. 2) Create the first iteration of a CAD model of our design. 3) Discuss the design s limitations and where it can be improved with Aren. 4) Revise our CAD design. 5) Repeat 3 and 4 as needed. 6) Manufacture our first prototype. 7) Test the prototype. 8) Talk with Aren and revise the design as needed. 6

7 Friday, June 5, 2015 The purpose of this meeting was to tie up loose ends from last meeting, break the team into subteams, and begin to create the first CAD iteration of our design. Summary: During this meeting, we decided upon how we wished to break up the work before dividing ourselves into three subteams. As some of our team members do not have design experience and others will be gone for large chunks of the designing process, members of subteams may help with the work of another subteam, should it be needed. Finally, each subteam sketched out a basic design for their submodule, and we began to create the first CAD iteration of our swerve module. Subteams: Responsibilities: Each of the subteams have separate design requirements that they must keep in mind during the design process. The chassis subteam must design a chassis that can house four of our swerve modules. The chassis must also be able to have weight added to it as needed, so that we can guarantee that our modules are strong enough to support at least 50 pounds each. The static module subteam will be in charge of ensuring that the correct gear ratios are used to transfer a proper amount of power from the motors to the wheel module. They must also create a design that is easy to mount onto a chassis. Finally, the wheel module subteam needs to create a design capable of multi turn operation that is not hindered by wiring or design. They must also create a module that is easy to attach to the static module. Chassis: Requirements: The chassis must have places for all four modules to be mounted to it. The chassis must have specific places to hold a battery or roborio. The chassis must put at least 50 pounds of weight on each swerve module. However, instead of creating a heavy design, weight will be added to the chassis as needed to put the proper amount of weight on each swerve module. The chassis must contain extra slots for optional parts and add ons. The chassis will be a two stage construction to ease the housing of batteries and electronics as well as make mounting easier. 7

8 Design: The original drawing of the chassis had the chassis at dimensions of 26 by 40. However, due to spacing issues, those dimensions will be shortened by 1, making the current dimensions 25 by 39. The chassis will be constructed mainly from 1 by 2 stock, with channel mounted up and down for added stability in the design. The design sketches of the two stages with the original dimensions can be seen below: Static Module: Requirements: The static module will be designed with ease of manufacturing in mind. This will mean the submodule will likely be bulky, but it will keep with the team s goal of simplicity. The static module will be made to be as light and compact as possible, without sacrificing simplicity. The static module will contain all the gear reduction needed to transfer power to the wheels and the turning mechanism. Design: The static module will house both motors, as well as all the gear reductions. To do this, the motors are mounted to the static module. Gears then carry power from the motors down to the wheel module, reducing power as needed. The static module also has two plates mounted underneath it on either end to protect the gears and allow for it to be easily mounted to the chassis. 8

9 Wheel Module: Requirements: The wheel module must allow for multi spin operation. The wheel module must contain a mechanism that spins the module. The wheel module must house a wheel that is connected to the drive shaft via a system of gears. Design: The current design only includes the wheel housing. It does not yet have gearing or a turning mechanism; that mechanism will be designed at the next meeting. The wheel housing is constructed from a piece of 1 by 2 channel to allow for the drive shaft to connect to the wheel module from the static module. It also utilizes two pieces of 1 by 2 C channel that allows for the gearing and wheel to be given ample protection and support. Holes in both types of channels have bearings in them, which will allow the rods the gears will be on to spin and transfer energy to the wheel. Monday, June 8, 2015 The purpose of this meeting was to finish the first CAD iteration of our swerve module design and to create a complete list of questions to ask Aren when we meet with him (hopefully tomorrow). 9

10 Summary: We started the meeting by writing down all the questions we have thus far for Aren. Next, we split into our subteams and finished the first CAD iteration of each submodule. We also wrote down more questions for Aren as they arose in our designing process. Finally, we ed the CAD files of the submodules to Lewis Jones, who will put the files in a Google Drive folder. We will assemble the full swerve module as soon as all the files are placed in the dropbox. Questions for Aren: 1) How durable are plastic clamping bearing blocks, plastic clamping gearboxes, and metal clamping bearing blocks? Are any of these sturdy enough to be utilized in our design? 2) Are there any more design aspects (ie price, weight, etc) that we should make a priority in our design? 3) Is our first iteration design strong enough to withstand the forces that it needs to? If not, how can we make it stronger? 4) Will the modifications we are making to various parts affect the component in any way? 5) How should we assume sensor input will be used? 6) Currently, the wheel module requires very precisely distanced bearing holes to be drilled to house the gearing shafts. The bearing holes also cut halfway through some of the precut holes on the channel. Should we change our wheel module design to prevent this? If so, what modifications should we make? 7) Are there any modifications that should be made to our documentation of this process? Is there any important information that should be included in our documentation left out? Chassis: A belly pan was added to the chassis to house the battery. Beyond that, work on getting all the parts of the chassis assembled continued on. Static module: The design of the static module was cut down from 121 square inches to only 50 square inches to reduce cost and weight. It was also mounted into the chassis to see how it would fit. 10

11 Wheel module: Gears were added into the CAD model this meeting. However, due to the very precise distances that the gears needed to be spaced at, the design had to be modified to facilitate the ease of manufacturing properly. As we were having trouble figuring out how to do this, we decided to ask Aren about the spacing before proceeding further. Tuesday, June 9,

12 After a brief series of s with Aren, we set up our first team meeting with him. The purpose of this meeting was to introduce ourselves to Aren, discuss basic project questions, and begin to discuss our first CAD iteration of the swerve module with him. Meeting results: 1) As the plastic bearing blocks are too weak as we suspected, we will need to use metal bearing blocks instead. However, the plastic gear boxes are strong enough for our purposes, as they do not need to support the weight of the robot, and therefore can be used in our design. 2) As we decided earlier in our design process, we should use absolute encoders to track the orientation of the wheels in the swerve module. We will also continue with our plan to 3D print an encoder mount for our module. 3) We should stick to the list of priorities that we had decided on earlier, as they will provide a good goal while not being unachievable. 4) Although we had originally wanted to keep the wheels directly beneath the drive shaft connecting the wheel module to the static module, Aren has said that it will not adversely affect the steering of the robot if we were to shift the wheels up to.25 off center. Because of this, we will shift the wheels off center to avoid hitting the bevel gears, allowing the gear spacing to be reduced from 100 to 84. This will also simplify the precision needed in manufacturing the wheel module, as VEXpro sells a part with holes pre drilled to achieve the spacing needed for this gear spacing. 5) Originally, we had been mounting the static module at an angle to the chassis in an attempts to reach the wheel as close to the edge of the chassis as possible. However, since this did not seem to push the wheel out far enough to be beneficial and was taking up an unnecessary amount of space on the chassis, at Aren s advice, we decided to mount the static module parallel to the chassis frame. We will also consider building part of the static module straight into the chassis to help reduce the cost and the weight of the swerve module. 6) We will need to reinforce our design so that the module will be able to properly support the weight of a robot without bending or breaking. Because the middle axle will support most of the weight on the module, we will concentrate on removing force from this axle. 7) We will need to flip the flanges on the bearings in the wheel module so that they are facing the inside of the submodule. 8) As bearing balls require a lot of very precise machining to create, we will omit the use of these from our design to keep with our goals of ease of machining and cost. 9) Instead of having the c channel in the wheel module be located outside either end of the 1 by 2 channel, we will move the c channel to be located underneath either end of the channel to allow for added stability to the module. 10) If there is enough 1 x 2 stock left after building the four modules to replace the c channel on the wheel module with 1 x 2 stock in order to cut down on the cost of the module. However, if there is not enough of the stock left, we will continue to use c channel to reduce weight. 11) To increase the strength of the swerve module, we will try to use bolts wherever possible in place of rivets, as rivets may not provide the strength and stability that is required. 12

13 12) We will need to somehow prevent the lever arm from bending. To do this, we can either reinforce it or shorten the module to, in turn, shorten the lever arm. Saturday, June 13, 2015 This meeting was meant to begin to implement the changes to our design that Aren recommended at our last meeting and continue to work on our swerve module. Summary: We began this meeting by introducing new development team members to the project. These members were then assigned to subteams, and we began to implement the changes that Aren recommended we make to our design. We also began to reinforce our design to make it better able to withstand the forces that will be put on it during use. Chassis: Work continued to get both stages of the chassis created in SolidWorks. Static Module: Since Aren confirmed our suspicion that the plastic bearing blocks we had been using in our design would be too weak for our purposes, we replaced these bearing blocks with their metal counterparts. However, due to a miscommunication on the subteam, the blocks were switched on the old iteration of the design, not the current, smaller iteration. They will be replaced on the current iteration at the next meeting. A thrust bearing was also added to the design to help strengthen the module. Wheel Module: The c channel was moved to be placed underneath the top 1 x 2 bar of the module. The rest of the module was reassembled in this new set up, and the wheel was added to be shifted slightly off center to allow for the wheel to avoid the bevel gear while keeping the module as short as possible. There will also have to be a gear reduction that takes place within the module to account for the fact that Vex does not sell 42 tooth gears with ⅜ hex bores. 13

14 Friday, June 19, 2015 This meeting was meant to continue to work on implementing the changes begun in the last meeting. Summary: Due to schedule conflicts amongst team members, not a lot was accomplished in this meeting due to absences. The members who attended this meeting spent the entire time working on the module s design. Static module: An encoder mount with a 1 to 1 ratio was designed and mounted onto the static module assembly in SolidWorks. This ratio caused the mount to be bulky, and we will consider changing the ratio to reduce the size and weight of this piece. Wheel module: A new iteration of the design was made so that a 3.25 Versa wheel is used instead of a 2.5 Coulson wheel. This was done so that a gear could be attached directly to the wheel, allowing for the submodule to be made narrower, shrinking the size of the axles. This also allowed for the submodule to be made shorter. These two things will help strengthen the submodule. Sunday, July 5, 2015 The purpose of this meeting was to continue work on the submodules, update our bill of materials to properly reflect our current design, and begin to assemble the second iteration of the swerve module. Summary: We began by updating all the files in the team s Google Drive folder. After this was done, the submodules were assembled to let us see any disconnects between the designs of the submodules 14

15 that had to be corrected. During this time, the bill of materials was also updated to correctly reflect the parts used in the current iteration of the swerve module. Finally, work on the submodules continued, and any changes that had to be made were dealt with. We ended with the plan to have the second iteration of our swerve module completed by the end of the next meeting. Current assembly: Chassis: Note: As the chassis subteam has been unrepresented at the past two meetings, the subteam spent much of the time examining the changes made to the module and updating the CAD files saved to their computers, allowing them to edit the current design. The light rail originally used for the chassis was changed to heavy rail in order to reinforce the chassis. Static module: Work continued to enact all the changes we had planned upon for this submodule. The plastic bearing block was replaced with a VersaBlock, removing a possible point of failure from the submodule. Further, the clamping gearbox was flipped so that the motor will be located under the chassis. This change allowed the pulley system originally utilized to transfer power to be replaced with a series of gears. The required gear ratios needed to obtain a speed of 12 feet per second was also recalculated, this time accounting for the gear reduction that takes place within the wheel module. The VersaPlanetary was also moved so that it is centered on the 2 by 1 15

16 channel that the static module is built off of. Finally, a bronze thrust bearing was added between the steering gear and the VersaBlock to help distribute the weight applied to the steering gear. Wheel module: The corners of the wheel module were trimmed by ⅕ to accommodate for the motor s new position in the module assembly. Further, changes were made to the bottom axle of the submodule to allow for the use of a dead axle in hopes that this would help strengthen the submodule. To facilitate this, the ⅜ hex bore gear was replaced with a bearing bore gear with VersaKeys. Bearings were placed in this gear and in the wheel before the two were mounted to the axle. To create the dead axle itself, the lower bearings in the c channel were removed, and the holes they used to occupy were shrunk to ½ holes. Finally, The ends of the bottom shaft were rounded, and shaft collars were mounted to hold the shaft in place. Wednesday, July 22, 2015 The purpose of this meeting was to enact final changes to the current CAD model to prepare it for another meeting with Aren Hill. Summary: We began the meeting by reviewing the changes we planned to implement after our last meeting with Aren and making a list of all the changes yet to be made. We then divided the changes that still needed to be changed amongst present members. While these changes were being implemented, a new list of questions for Aren was created. Finally, the newest CAD iteration of the swerve module was assembled. General The Bill of Materials was updated so that it no longer utilized the full price and weight of a whole length of stock in the calculations for the price and weight of each subassembly. Instead, the weights and prices are adjusted based upon the length of stock actually utilized. Additional parts used in the wheel module that were not yet included in the Bill of Materials were also added. However, the Bill of Materials for the static module are still not up to date, as the newest iteration of the submodule is not yet completed. Further, all of our CAD files were also transferred from Google Drives to GrabCAD to streamline the process of file sharing. Finally, a naming system was implemented so that every member of our team will be able to know what each file is. Chassis: 16

17 The chassis subteam began to finalize the chassis design. This involved moving the steering motor on the static module to put the wheel into the farthest corner of the drive base possible. This will create a larger drive base, allowing for better strength and stability. Static Module: A design for the steering encoder mount was started on. The current design has the encoder mounted directly to the channel of the static module. A 3D printed gear will be connected to the encoder and will be driven by the gear on the BAG motor. This design does not have a one to one gear ratio, thus the programing will be more difficult. However, as any other design would require more machining and parts, we will try to utilize this design. Finally, cams were added to the static module to help control the location of the clamping gearbox and to ensure that the gear teeth mesh properly. Wheel Module: Bolts were added to the wheel module. However, to allow for the bolts to provide enough hold, six extra holes had to be added to the c channel. The shaft collars were swapped for a washer that will be bolted onto the end of the shaft. This will still serve to hold the shafts in place while allowing for the module to be lighter and more compact. Research into possible designs for a drive encoder was conducted. Possible design ideas include using a slip ring and placing the encoder on the wheel module or the static module. Finally, potentiometer research was also conducted. To use the US Digital MA3 like we originally planned, we would need a to use a bearing ball version to use a 1 to 1 gear ratio and still stay within the maximum listed RPM. Tuesday, August 4, 2015 As our final changes to our design were not completed at the last meeting, this meeting was dedicating to tying up all of the loose ends. Summary: Right away, we jumped into finishing the changes we began at the last meeting. As the meeting progressed, we were able to complete all the final parts and create a list of questions for Aren about our design. Static Module: 17

18 A mount that will house the drive encoder was created. This mount will be 3D printed and will be bolted to the main channel of the static module. A 3D printed spacer was also added to hold the motor in place. Wheel Module: Spacers were added to the wheel axle to prevent the wheel and gears from sliding around on the axle. One of these spacers will have to be sanded down slightly due to the length of the axle. Saturday, August 29, 2015 The purpose of this meeting was to make the final changes that Aren recommended when he viewed our GrabCad files. Summary: We began by reviewing the comments that Aren made. These can be found below. We then divided into our subteams and began to make corrections to our design. Aren s comments: 1. We can replace the 2 by 1 channel on the chassis with 1 by 1 channel. 2. We can use the belly pan panel on the chassis to replace some of the multitude of gussets on it. 3. The snap ring on the long drive shaft of the wheel module is likely to snap. 4. We should replace our gears on the static module with regular gears instead of VersaKey gears to save money on our design. 5. The gears for the steering encoder should be converted into a 1 to 1 ratio to help with programming. Further, we should mount the encoder to the VersaPlanetary to allow for more accurate readings. 6. The bolts on the wheel module should be spread out to minimize alignment issues. 7. The current gussets on the wheel module are not long enough to provide proper support. We should create custom ones to fix this. We made the following changes to each module to address these issues: Chassis: 18

19 1. The 2 by 1 channel was replaced on the chassis, as recommended. This has served to cut both the weight and cost of the module. 2. The belly pan panel was used to replace some of the gussets, further reducing the weight and cost of the module. However, this will require more custom parts. Static Module: 3. The snap ring on the long drive shaft was replaced with a shaft collar. This design will help provide the module with more strength and reduce the amount of machining. 4. The VersaKey gears were replaced with normal gears where possible to help lower the cost of the module. 5. A 3D printable encoder mount for the steering encoder was created. This mount places the steering encoder so that it is in line with the VersaPlanetary, preventing the encoder values from becoming off set from the wheel module when the robot is turned off. The gearing ratio between the VersaPlanetary and the encoder was also changed to 1 to help simplify the programming and make readings more accurate. Wheel Module: 6. The bolts on the wheel module were spread out, as suggested. 7. Custom u gussets were created to replace the old gussets and help strengthen the wheel module. This will require more machining, but we will experiment with methods of machining to make it as simple as possible. September 9, 2015 The purpose of this meeting was to discuss a few more issues to fix before we begin to manufacture the prototype of the swerve module. Summary: We addressed the last few design issues we were having with the module. Chassis: We simplified the mounting of the swerve module to the chassis by moving the VersaPlanetary back onto the same piece of 2 x 1 tubing that we used for steering. Static Module: The static module currently uses an idler gear for power transfer. We have replaced this gear with a belt to cut out the expensive, unnecessary gears. As mentioned before, the VersaPlanetary was moved onto the 2 by 1 tubing used for steering to help simplify the process of mounting the swerve module to the chassis. 19

20 Wheel Module: The ⅜ hex axle that is currently supporting the entire wheel module does not seem sturdy enough to resist the forces that will be applied to it. As an alternative, we added a 1 tube to help distribute the force put on the wheel module. This will require more machining, but it should be more reliable. However, we will make a prototype of both designs to determine if we really need to add the tubing. September 20, 2015 This meeting was meant to take care of the final details of the design, as well as update documentation so we will know what parts we need to order for our module. Summary: Work was divided amongst the members of the team: Task Details Team Member Broaching Gearing Bill of Materials Steering Encoder Mount Drive Encoder Mount Tensioning System We needed to figure out if we could successfully broach a keyed hole in a long piece of ½ hex axle (up to 2 inches in length). The current gearing of the VersaSwerve has an adjusted drive speed of 17 ft/s. As this is quite fast, we will regear the module to have a speed of between 11 and 14 ft/s. The Bill of Materials is very out of date. As a result, we will update the Bill of Materials to make it accurate to our current iteration. A 3D printable encoder mount must be made to hold the steering encoder. The current drive encoder mount no longer lines up with our current swerve iteration. As a result, the mount must be redesigned to properly fit the current module. The new belt and pulley system in the static module requires a system to tension the belt and ensure proper energy transfer with minimal teeth skipping. Sidhu Byron Lake and Katie Byron Lewis Michael 20

21 Chassis Belly Pan 1 Tube Support Iteration Shaft Collars and Spacers The new mounting hole pattern must be added to the chassis to allow for the proper mounting of the modules. Mounting patterns for the electronic components of the module must be added to the belly pan. The CAD for the 1 tube supported swerve needs to be complete, so we can manufacture this iteration. All missing shaft collars that are necessary for the design to function properly need to be inserted. Brett Ben Lewis Lewis September 21, 2015 The Swerve team was given the go ahead to manufacture both the swerve iteration with the 1 tube support and the iteration without the support, so we could determine if there were advantages to having the extra support. 21

22 September 28, 2015 A new 3D printed pot mount for the drive encoder was created. This will mount directly to the versablock, so it can move with the CAM tensioner. Since the change does complicate installation and maintenance, we have come up with alternate mount. This mount features a slot for adjustments of the CAM tensioner, therefore allowing the mount to remain securely attached to the 1x2 channel. Also, the drive gears on the wheel module were changed to a 34 and a 50 tooth gear. September 29, 2015 The belt tensioning system was attached to the static module in Solidworks. The belt is 400 teeth long and 9 mm wide with a 5mm pitch. The configuration of the swerve module that uses a 1 tube support was created in Solidworks as well. Finally, the 1.25 ID thrust bearing was replaced with a 1 ID thrust bearing with drilled out holes. September 30, 2015 The team discovered that a.5 inch hex shaft with an 8mm keyed hole was sold on Vex s website, making our struggle with broaching the shaft obsolete. A new design was made that used the pre broached shafts, simplifying the manufacturing process. It uses two of these keyed shafts to transfer the force from the CIM motor to the pulley in the static module. November 25, 2015 Preparations to begin manufacturing the first modules began. Two swerves will be manufactured: one with the tube support and one without. A small test rig will be designed and built to allow for testing to see if one iteration has an advantage over the other. A few roles were assigned to be sure this process went smoothly. They are as follows: Lead in Testing and Research : This person is charged with setting up experimental parameters, conducting experiments, and recording all data. He will report on the results of this testing and create proper documentation of the testing process, including pictures and video footage. Lead in Design Development Document : This person is in charge of finishing and refining the design process write up for this project. Lead in Manufacturing : This person must manage the manufacturing of the two modules. He will be in charge of ordering all the parts needed for manufacturing and organizing 22

23 the manufacturing team to ensure the swerves are created correctly and completed on time. Lead in CAD Module Refinement: This person is in charge of updating and polishing the CAD files. The bill of materials must also be updated to reflect the current design. December 9, 2015 A small test rig was created to allow for testing the two modules. The rig includes a small chassis for attaching the swerves and housing the electronics, as well as a castor wheel to keep the rig stable and drivable. The testing proved that the 1 tube support model of the swerve had the upper hand over the unsupported model. The tube support model reduces the amount of custom machining that has to be dune to the thrust bearings, and it is considerably stronger and more durable. Two more tube support swerves will be manufactured for full scale testing. December 12, 2015 Final changes were made to the CIM output shaft coupler. While the old design had a set screw that was screwed into the CIM output shaft, the new design has the set screw relocated higher so that it can be used as a hard stop to prevent the coupler from falling out. This design has also prevented the coupler from moving up on the CIM shaft, as it did in the old design. The assembly instructions have been changed to reflect this, and manufacturing has been minimally affected. December 18, 2015 The test drive has been completed and further testing has reinforced our conclusion that the tube support model is the stronger and better suited of the two models we designed.with the design completed and the drive base functional, the VEX documentation has been completed and sent to Aren Hill to be posted on the VEX website. 23