This morning we placed the nearly-$1700 order with VEX for all the swerve drive parts. The biggest cost was the VersaPlanetary gearboxes. We’re building six complete swerve modules.
This afternoon, we picked up a new refrigerated air dryer and a MakerBot digitizer from our generous sponsor Glendale Community College. These should be very welcome additions to our lab.
Today’s meeting began with a short ten minutes meeting where Mr. Hoard and Mr. Black which addressed snack-time etiquette, and the preparations for a visitor at 6:30.
The Team had the honor of having a guest, a previous mentor who had supported the team during it’s toddler years between 2001-2006, along with his great kids. The visitor, Mr. Daniel Limonadi, works at JPL where he works as a flight systems engineer for the Mars rovers. Mr. Limonadi worked with the team for five years, during this time a certain Mr. Black had jumped into the very same program and later returned to become the team’s lead mentor, and of those five years there were days where Mr. Limonadi would appear on the News in the morning and be mentoring students on the team in the afternoon.
Before Mr. Limonadi’s visit, the team was hard at work finishing the prototype and working on the robot design. Downstairs in room 5203, Aleksandr, the student programming lead, with the help of Daniel (programmer) and Sipan (CAD) worked to set up the VEX robot’s swerve drive. Of the team’s fourteen years in FIRST, the use of a swerve drive is a first (pun intended). The reason behind this decision has to do with the fact that Recycle Rush has practically no interaction with the opponent side of the field meaning no defense. Generally, the Team sticks with a west-coast-drive to stop defense in it’s tracks. In fact, last year, the team improved on the design of their gearboxes so that their were two different speeds: one for speeding around the field, the other for bouncing opposing robots off of us.
The CAD team is hard at work to finish the swerve-drive gearbox in the scheduled time frame. Currently, the team is working on the “Swerve Drive Ball Groove” as indicated in the “CrayolaCAD” sketch below.
On the Electrical side of things, the electrical team worked on making a pseudo mock-up of the electrical system on the robot for the programmers to test with.
With Mr.Hoard, today was the day for “You Don’t Know What You Don’t Know” where Mr. Hoard walked around asking students questions based on the manual. Such questions included: “What is the height limit for the transport configuration?,” “What are the dimensions of the totes,” “How many recycling cans are on the field at the beginning of the match?”.
On a final notes, much of the work left on the prototype is the elevator mechanism which we hope to have completed by tomorrow.
Today’s meeting was a vital point in this year’s overall season due to the final decisions made concerning the robot design and strategy. The day started with Mr.Black collecting the team together to articulate the importance of an engineering mindset: to be curious, to adapt, and to improve. He urged us to think about every detail, and to foresee what potential problems might spring from even the most minor of actions (a nut not tightened, a tiny piece of information forgotten from the schedule, leaving a door open).
To branch out from yesterday’s meeting, the team had concluded several strategies concerning robot design and human player actions that night. From those rough ideas, today the team’s main builders, thinkers, and leaders met together in the first hour of the meeting to conclude the thoughts of the last few days. It was decided that the team would begin prototyping, and began assigning people to each task as seen in the picture below.
During this meeting, students talked extensiveness on the mechanism behind holding onto a can when at the full six tote height and how one might keep a hold on a stack of totes while moving. There were many “rib” ideas, mechanisms that wrapped around the base of the can (because at this point you would only have about three inches of the can to hold onto), but there was also the “pulley” idea where two ropes would range from bottom to top of the robot to keep the totes moving vertically and not tip forward. This idea will be branched out into a prototype soon.
The first prototype of the day was a pushing mechanism based off the idea that if there ever was a need for a can or tote to be placed on a stack, that we would have the ability to do so. This was accomplished by two large pistons attached to a steel plate; this design was the initial idea, but the main problem was that the pistons were too heavy and the amount of air we needed to push a six tote stack was not that much. The idea was revised into a different system where two smaller pistons were zip-tied together and screwed into a wood plate. After testing this with several different tote sets, we found that this model was significantly more efficient than the first edition.
Other mechanisms were given the “OKAY“for prototyping like the Autonomous “Can-Hook” shown in the below video. This prototype is an autonomous strategy to grab the cans for latter use and to also drag them into the auto zone for points. This mechanism has two long arms, fairly light for their length, that hook under the rim of the can handles to be pulled into the alliance’s landfill/ auto area.
Another mechanism worked on today was the “Tote Rollers.” This prototype is based on tote alignment. The idea is two arms with two motors and wheels at the ends will spin the tote into alignment with the elevator. This is an important piece that might make or break a robots overall design: if we don’t have something to align the totes for us, we will probably never have the opportunity to pick up from the ground. This prototype was met with difficulties on all ends, in fact failure in some, due to electrical and mechanical errors. While the wheels were rolling by the end of the day, there was a major problem in the fact that the planetary motors attached at the ends of the arms were not strong enough to do anything to a tote. We learned from our failure and are continuing to work on fixing the overall design of this prototype.
The final prototype being worked on today is the elevator mechanism. The team had started work on this in the afternoon, and by the end of the day the prototype was well on it’s way. We’ll keep you updated on it’s progress and problems we faced.
In other parts of the team, Business had finally tackled an issue it had been dealing with for a while. The “books” of the overall balance of sales on Light bulbs, one of the fundraisers the team is doing, wasn’t weighed properly but after some investigation the books were settled! Now Business is working on the Chairman’s award and contacting new/old sponsors.
Throughout the day, people kept bumping into Mr. Black’s 1/72 scale Saturn V rocket model display case, causing the 1/72 scale German-made “Civilian Airport Personnel” figures to fall over. We are happy to report, they have all been returned to standing position, and the Saturn V is cleared for launch.
In Leadership, an important schedule was made by Jack, the president, on the due dates of manufacturing. This is an important step in planning out the rest of the season as we hope to have a few days of drive practice before the robot has to get wrapped up and tied.
Additionally, Jack took inventory of the Kit of Parts and separated out the important items such as Product Donation Vouchers.
The design of the VEX swerve drive was drawn out in ExpoCAD, and then in Autodesk Inventor CAD from start to finish. A picture of each is shown below:
On the CAD side of things, Mr.Black and Cynthia, the leader of the CAD team, worked to properly configure Autodesk Vault for the 2015 season. This proved to be a major undertaking, which needed several hours of attention, but in the end we set out to accomplish our goal of proper CAD data management. The difficulty was in setting up the Vault Project File and working folders, which we still don’t entirely understand the inner-workings of, but it works properly for now. Additionally, we activated a new feature on the Vault server (ADMS Console) to “Track File Status”. We initially had some concern about the Vault becoming “locked” from performing the “Update File Status” action, but experimentation on a test vault proved it to be a non-issue. The new “Track File Status” feature will warn “User A” who has an assembly open if “User B” has checked in updates to components that are children of the assembly.
On design of the swerve module, the following determinations were made:
The 50 tooth gear driving the wheel will hit the ramp in certain circumstances, producing an unfavorable condition. Reducing the size of this gear increases the primary reduction ratio needed for the desired speed. We attempted changing the 3×7/8″ colson wheel to a 3.25″ VersaWheel DT, but while the diameter fit, the width did not, by about 3/16″. We did not want to modify the wheels in any way, so we decided to reduce the size of the gear. While we discovered VexPro offers a 48T gear with the VersaKey pattern (they do not without the VersaKey), a 48T gear would still scrape the ramp when the robot is “sunk” into the carpet.
We changed out the final drive gear to a 44T gear, which brought its own set of problems in that the Colson wheel would interfere with the bevel gear directly above it. Fortunately, we were able to slide the Colson wheel over on it’s axle 1/8″ by removing a previously-designed-in spacer, providing adequate clearance between the wheel and all components in the region. So, the wheel’s center plane in the axial direction is located 1/8″ away from the rotation axis of the swerve module. We anticipate this will cause only a negligible effect on driving.
The portion we formerly referred to as the “hat” is now being named the “tube”. The “top hat” has yet to be designed.
After discussing the possibility of the “tube” piece (upper bearing support) and the lower bearing blocks being made as one part that extends through the gear, it was determined that the bearing blocks should stay separate pieces, for simplicity in the machining process.
The “tube” will be extended upward to near the height of the top steering miter gear support bearings, to reduce the height of the top-hat and reduce complexity of machining of the top-hat.
The tube piece will actually be made from thick-walled aluminum tube (3.5″ OD, 2.5″ ID) rather than from solid bar stock. This reduces machining time. McMaster is the only vendor we’ve found of aluminum tube in this size.
We discussed and debated how to constrain the module from dropping out the bottom of the static-mounted plate. Mr. Widholm made a proposal to add another plate with a circular-profile groove underneath the gear, and another circular profile groove in the bottom side of the large 84T steering gear. We determined that to maximize efficiency in our manufacturing production schedule, we’d rather place four radial bearings on shoulder bolts on the top side of the static-mounted plate.
We want to place material back into the center of the previously-designed lightening pockets in the tube, to support the radial bearings on the shoulder bolts discussed above. The other two radial bearings will be supported by the 30T gear axle, with the shoulder bolts threading into 10-32 tapped holes in that axle.
We determined that we should build 6 complete swerve modules. Four for the robot, and two spares / demo units. Due to the cost involved, we will not purchase a VersaPlanetary gearbox and Mini-CIM for one of the spare units.
We determined the electronics mounting area in between the swerve modules may not be as spacious as previously anticipated due to the lay-down nature of the VersaPlanetary gearbox.
As we were closing for the night, and discussing the chassis layout, we realized that the two static-mounted plates for each swerve module would need two different versions. The full assemblies on opposite corners would be identical. The geometry of the plate in the horizontal plane is the same, but vertical axis features will likely be different between each pair of modules, requiring different CAM programming.
A final picture of our design whiteboard before closing time, featuring our “ExpoCAD” of the robot for the day is shown below:
Going off of yesterday’s goals, the team had high expectations of what we needed to do today. To recap, our goals included an “expo” CAD model, a clear and well thought out strategy of how to play Recyle Rush, and the mechanisms to do so.
We started in the lab at 2pm and began work on strategy almost immediately. After yesterday’s reveal, everyone had gone home to study up on the game in preparation for Mr. Hoard’s, the business mentor, game “You Don’t Know What You Don’t Know” but we’ll talk about that later.
Each Subdivision had important jobs right off the bat for this season. Today, Business worked on the Chairman’s award along with the writing for the Kickstarter which will be up soon.
The Prototyping Team, along with the help of the CAM (Computer Aided Manufacturing) Team and a few CAD (Computer Aided Design, if you didn’t know what these acronyms mean) members, helped continue set up for the game field.
On the Programming side, members began research on encoders for a swerve-drive mechanism ,and continued work on a FRC simulator which would simulate robots that would run the code the programming team is working on.
The 3D printing members were not idle either, they were printing prototype parts for possible mechanisms and also worked on a few side projects that occur on and off season. (Thanks for the title, Sipan ).
Cynthia, the Leader of the CAD team, worked on the swerve drive (as Mr. Black has shown in his previous post about his midnight CAD Club), and props to her for the amazing mechanics going into her design!
Animation was working on metallic surfaces for this years robot animation along with rendering past season animations.
Now, to the team we’ve all been waiting for: Strategy! Less of a team and more of a huge glob of people, the strategy team worked hard on game dynamics and mechanism. Today, hours were spent on the overall strategy that the team wanted to accomplish this season. Recycle Rush proposes a challenge in that it requires several different mechanisms to do certain tasks. “How many yellow totes should we aim for?“, “What mechanism should we use for the green recycling bins, and can we use that same mechanism for totes?“, all of which were important questions discussed thoroughly today. The team regularly migrated from the computer lab, to the shop floor, and to a white board.
The Team’s goals today were to make an “expo” CAD model (half credit), and a well thought out strategy to go hand-in-hand with smart mechanisms (three-quarter credit). While we reached our main goals and mech for the autonomous period, the team still needs to think about the strategy for the teleop period which is a completely different ballgame.