The Story of Our Robot. Vexmen Downingtown PA

Transcription

1 The Story of Our Robot Vexmen Downingtown PA

2 Table of Contents About Team 81 M... 2 The Base... 3 The Problem... 4 Relevant Rules and Parameters... 4 Brainstorming and Concept Models... 7 Changing the Base... 9 DoubLe reverse 6-bar The Problem Relevant Rules and Parameters Brainstorming Designing Building the Lift Failure and Physics The PantaGraph The Problem Rules and Parameters Design and Inspiration Build Montage Failures and Struggles Investigating Material Properties The Skyrise Claw The Problem Relevant Parameters Brainstorming Building the Skyrise Sweeper The New Scissor Lift Building the New Scissor

3 About Team 81 M Team 81M is a high school Vex Robotics Team. The team is part of the Vexmen Robotics Club from Downingtown Pennsylvania. This team is dedicated to learning advanced concepts connected to Science, Technology Engineering and Math by Designing, building and programming a competitive Robot. This year we focused on using experimental data and virtual modeling to influence our design and using Agile Project management to organize our build efforts. We modeled many of our ideas into fully built, detailed Autodesk Inventor Drawings to fully realize our design before building. Mystique also used experiments to design our robot. We experimentally tuned our PID control. We also devised an experiment to test the differences between the material properties of different Vex parts. We preformed tensile strength tests and analyzed stress strain curves. We also learned about project management. We learned about Agile project management and created sprints, burn down charts, epics and stories throughout the season. This helped us keep our hectic season organized and on track. Our team attended four competitions this year. We were very successful and won at least one award at every competition. We attended the Downingtown Classic on January 10, 2015 and won the build award. On January 24, 2015 we attended the Dockbots competition. There we won our season s first design award. The next event we attended was the DC Knights Qualifier on January 31, There we won the judges award. Lastly, we attended the Pennsylvania State Championships. There we won both the design award and the title of Pennsylvania s Programming Skills Championships Our team is composed of four members: Rob, Miranda, Sarah and Ian. Rob is a junior at Downingtown West High school. He has been participating in Vex Robotics for six years. He is an amazing builder and Designer. He has previously been on teams 81, 80G, and 81G. Miranda is a Junior at the Downingtown STEM Academy. She has participated in Vex for five years. She is the team s programmer and notebooker. She has previously been on teams 80 and 80N. This is her 3 rd year on team 81M. Sarah is also a junior at the Downingtown STEM Academy. She is our project manager. She manages our workload and ensures all our projects are on track. This is her second year in Vex robotics. Last but not least, 81M s rookie member is Ian. It is his first year in Vex robotics. He is a Freshman at Downingtown West High School. Ian is our best coach. Now that you know about the team its time to learn about the Robot! 2

4 The Base OUR DESIGN PROCESS Team 81M started off the season by designing our base. To design this subsystem we defined the problem, researched relevant rules, brainstormed ideas, created concept models, chose and built a design and then modified that design later 3

5 The Problem This year s Vex Robotics Competition Game is Skyrise. A picture of the game field is to the right. The game field is 12ft by 12ft big. The field has one main pile of cubes in the middle and cubes scattered around the outside edge. The cubes on the floor may be an obstacle when trying to maneuver. There is also the Skyrise bases in the corners. To reach those bases we do not need to move far because they are near the starting tiles. Our task is to create a robot base to navigate around the field and play the game. Relevant Rules and Parameters Size Requirement Specification: The robot base must be no larger than 18in by 18in by 18in large Testing Method: The sizing tool will be used to test the final base and any concept models in Autodesk Inventor will be measured before fabrication. Why? Robot Rule R4 constrains the dimensions. Since the base is the first system we build, we have the full 18 in by 18in by 18 in dimensions available. 4

6 Materials Restriction Specification: The robot base must be made only out of Vex parts. Testing Method: Comparing the parts list to legal Vex parts Why? Rules R5 and R 6 limit the materials to only Vex parts. We cannot use any non-vex parts sold from non-vex resellers. There are some parts that do not follow this rule. Any Identical parts can be used as well. These exceptions are defined in rule R7 Any IDENTICAL Vex part can be used. There is also the option to use polycarbonate to build. Identical Electrical Non-Vex parts are not allowed. 5

7 Our base must also be light, easy to build and fairly maneuverable. Using these rules and our goals we brainstormed many different base designs including Tank Drive bases and Holonomic bases. We narrowed our brainstorms down to Holonomic versus Tank Drive. 6

8 Brainstorming and Concept Models Holonomic versus Tank Drive We brainstormed two possible designs, made sketches and rendered them in Autodesk Inventor. After a lively debate, we concluded the tank base design is far superior to the holonomic base for a variety of reasons. Even though the holonomic base is more maneuverable, Holonomic bases are generally harder to design and build. The holonomic base also is less powerful. This is due to the force vectors being split over an angle. The max potential motor exsertion is the motor power divided by the square root of two (see journal for more info) The tank base may be less maneuverable but they maximize the motor force and are easy to build. Because of these factors we chose to build a tank Drive base. 7

9 The Base was built almost exactly like the Autodesk Inventor Model. The only difference between the virtual and real models is the real model has quad encoders. The Quad encoders mesure wheel rotation and will be helpful when programming later. 8

10 Changing the Base In late December, we built a scissor lift (the lift is detailed later in the book). The lift required a narrower and shorter base due to the internal lift system and a planned grasping mechanism. We first sketched out the change in base dimensions in our engineering design Journal and we then prototyped the change in Autodesk Inventor to ensure it will not conflict with any existing mechanisms. The drawing above is the existing base. The drawing to the right is the improved base. The base was shortened and it was made narrower. The New base is 14 in wide and the old base is The base now can accommodate the news lift and the grasping Mechanism. After this iteration, we also eventually switched our motors to strength instead of the original speed. 9

11 THE CAM SHOOTER DoubLe reverse 6-bar. 10

12 The Problem Now that we have a fully functioning base, we need a lifting mechanism to raise the cubes off of the ground and score them. The goal heights are 47, 40, and 24. The mechanism must be both tall and stable because the higher the robot the greater the risk of tipping. We need to keep this in mind when designing. Problem Statement: To play Skyrise we must create a mechanism to lift game elements up into the air and score them on top of the various sized posts scattered around the field. We must build a stable mechanism to perform this task while also fitting it on the existing base. 11

13 Relevant Rules and Parameters Size Requirement Specification: The robot must be no larger than 18in by 18in by 18in large The arm can be no larger than 14 by 18 by 18. Testing Method: The sizing tool will be used to test the final arm design and any concept models in Autodesk Inventor will be measured before fabrication. Why? Robot Rule R4 constrains the dimensions. The current base also constrains the dimensions of the arm. The height of base is four inches tall. To comply with rule R4 we must have a lift no larger than 14 inches. Materials Restriction Specification: We must construct the arm solely out of Vex parts or Vex legal parts Testing Method: Comparing the parts list to legal Vex parts Why? Rules R5 and R 6 limit the materials to only Vex parts. We cannot use any non-vex parts sold from non-vex resellers. There are some parts that do not follow this rule. Any Identical parts can be used as well. These exceptions are defined in rule R7. Any IDENTICAL Vex part can be used. There is also the option to use polycarbonate to build. Identical Electrical Non-Vex parts are not allowed. We are allowed to use screws like cap head screws, and identical parts like Teflon washers. 12

14 Height requirement Specification: The lift must be able to reach the goals in the game and score cubes on top of them Testing Method: A tape measure will be used to measure the final design and all concept models will be measured. Why? Since the highest goal is 47 inches tall and the cubes are 8 in tall so the lift must h a v e a r e a c h o f 5 5 i n c h e s 13

15 Brainstorming The Double Reverse 6-Bar We looked at other robots in our club for inspiration. 80Y s double reverse four bar intrigued us. We brainstormed a new similar design. We analyzed the difference between a six bar and a four bar and found that there was a significant height difference. This difference made us wonder if a double reverse six bar would have the same advantage over a double reverse four bar. We sketched both double reverse Six Bars and double reverse four bars to analyze the difference. The height difference was very evident when two sketches were drawn side by side. The double reverse six bar has a large height advantage. Because of this we decided to build a double reverse 6-bar. 14

16 Designing Designing the Double Reverse Six Bar When designing the double reverse Six Bar, we created a layered drawing. This drawing broke the design up into four distinct layers this helps us build and test whether the components will interfere with each other. The details also help us streamline the building process and quickly assemble the arm. The layers are all depicted in the pictures on the right. The design will use both high strength motors and high strength gears to ensure the lift is durable. The entire lift is made from Aluminum to ensure it is light enough to lift. We designed one tower and planned on building two. The second tower was a perfect mirror image of the first one. The two towers would be connected with c-channel support bars and a sync shaft After we were done designing, we gathered materials and prepared to build. 15

17 Building the Lift 16

18 Failure and Physics After constructing the double reverse 6-bar, we tested it. We ran the motors and the robot began to tip over before it reached maximum height. It tipped so often that we named it Sir Tipsalot. It became a serious problem and we decided to investigate into what makes the robot tip and for that we needed to look at the forces causing the robot to tip over. To do this, we started by drawing an Free body Diagram of the robot s center of gravity. That diagram is below. F g is the Force of gravity of the robot. F g = mg (m is robot s mass, g is acceleration due to gravity). F x is the force in the x direction. F x = ma x (m is mass, a x is acceleration in x direction). S is the distance between the wheel s touch point and the center of gravity. Θ is the angle between the Force of gravity and the line between the center of gravity and the wheel. The dotted lines are normal to line Center of Gravity F connecting the center of x gravity to the touch points. For the robot not 90 - θ to tip, the red component must be θ bigger than the blue θ S one. F g Wheel The red line s magnitude is defined as: F g sin θ = redline. The blue part s magnitude is defined as F x cos θ = blueline. Since the redline s magnitude in for must be greater than the blue line s to avoid tipping, F g sin θ > F x cos θ. After substituting ma x F x, substituting mg in for F g, and solving for acceleration in the x direction, we found our final equation. g tan θ > a x The acceleration due to gravity multiplied by the tangent of the angle from the touch point to the center of gravity must be greater than the acceleration in the X direction for the robot not to tip over. Due to the double reverse 6-bar s weight distribution, the center of gravity constantly moved. The center gravity often hung over the back of the wheels so any acceleration in the x direction would be greater than the tangent of the angle times the acceleration due to gravity. Basically, both the design s geometry and weight distribution caused the tipping. The physics revealed that major changes would be needed to prevent tipping After a long debate, we decided to redesign a new mechanism because we thought the problems were too severe to fix without scrapping the whole thing. 17

19 The PantaGraph THE PANTAGRAPH 18

20 The Problem Since our double reverse six bar was a complete failure, we needed to redesign a lift system. We reanalyzed the game and redefined our problem. Rob and Miranda noticed that this year s game Skyrise closely resembled an older game, Gateway. There was a similar scoring method, de-scoring rules, post owning bonus points and goal heights. The fields also had a large pile of game elements in a pyramid shape with other elements scattered around the field. There are a few differences between the two games but the methods of scoring points are fundamentally the same. Since the two games are so similar, examining previous successful gateway robots could help us with our design for Skyrise. Our team had not been to a Skyrise competition at this point in the season but Rob and Miranda were very experienced when it came to Gateway. Keeping the game similarity in mind, we redefined the problem. Updated Problem Statement: We must create a lifting mechanism inspired by Gateway to lift game elements onto the various sized posts without tipping over. 19

21 Rules and Parameters. Size Requirement Specification: The robot must be no larger than 18in by 18in by 18in large The arm can be no larger than 14 by 18 by 18. Testing Method: The sizing tool will be used to test the final arm design and any concept models in Autodesk Inventor will be measured before fabrication. Why? Robot Rule R4 constrains the dimensions. The current base also constrains the dimensions of the arm. The height of base is four inches tall. To comply with rule R4 we must have a lift no larger than 14 inches. Height requirement Specification: The lift must be able to reach all posts and must be high enough to successfully score on all the goals WITHOUT tipping over. Testing Method: A tape measure will be used to measure the final design and all concept models will be measured. The final design will be placed on a field and will attempt to score goals. Why? Since the highest goal is 47 inches tall and the cubes are 8 in tall so the lift must have a reach of 55 inches because to lift the cube on top of the goal, the bottom of the cube must be higher than the top of the highest post. 20

22 21

23 Design Constraint Specification: The robot must be inspired by a mechanism used in the game Gateway. Testing Method: We will examine pictures of gateway robots and improve upon the designs from those robots. Why? Our old mechanism was inspired by a toss up robot and toss up was a very different game with much lower goals. The highest gateway goal was 30in tall. Robots for gateway were designed to reach lofty heights. Any extremely successful design from that game should also potentially be successful in Skyrise. We will not copy the mechanism but we will take inspiration only from gateway designs. Stability requirement Specification: The lift must be stable and have a stable center of gravity. A lift where the center of gravity only moves up and down and does not shift left or right would be ideal. Any particularly back heavy designs like the double reverse 6-bar are not acceptable. Testing Method: To test this we will run the robot and see if it tips over and we will guestimate the center of gravity in any of our designs. Why? We want our lift to refrain from tipping over during game play and we do not want it to fail like Sir Tipsalot failed. We want to incorporate the stability into the overall design because it is a crucial aspect. 22

24 Design and Inspiration Our inspiration Three years ago Miranda designed and built a special scissor lift for Gateway. The final height of her design was an astonishing 5ft tall. Her original inspiration for the design was an episode of the show How its Made. On the show, collapsible room dividers were created using a special scissor lift called a pantograph. Miranda was inspired by this design and created the gigantic robot seen in the picture on the bottom right. Her robot was taller than she was. That robot competed in Gateway and attended the Vex Robotics World Championship in It was a surprisingly stable design. It was very effective at playing Gateway and won multiple tournaments in the Pennsylvania region. We decided an improved pantograph would be the best lift solution due to its stability and success in Gateway. 23

25 Improving the Old Design We kept the main structural elements from the old lift design but we improved the lifting mechanism by making it lighter and more compact. The top pictures shows the old robot lifting mechanism. It was spaced out and primarily made from steel. The sync shaft required many 12 tooth gears and was fairly inefficient. The new lift design is narrower and lighter. Unlike the old deign, the new one uses the new style of linear slides to propel the tracks up and down. The new style is lighter because only one of the two components is steel. The new lift is also more compact because instead of a sync shaft, it uses two 60- tooth gears to synchronize the motors. The final rack design also was longer and allowed a greater range of scissor lift motion. 24

26 Build Montage 25

27 26

28 Failures and Struggles Failure The design was tall, refused to tip over and was fairly light but it ultimately failed. The top levels of the scissor did not extend upwards at the same rate. This led to one side raising higher than the other. The top was tilted and unstable. Instead of a tipping problem we had a tilting problem. On top of the tilt, some of the bars in the lower sections began to bend. We tried multiple solutions to stop the bending and tilting. All failed. Eventually we removed over half of the scissor levels severely shortening the reach. We eliminated the bending but we limited our reach to the short goals Our lift was almost identical to the one Miranda built in Gateway. We wondered why our lift failed because the previous design worked perfectly. 27

29 Investigating Material Properties After the failure of our lift, we decided to examine the pictures of Miranda s old Scissor to see if there were any differences between the two. We observed that the metal Miranda s scissor was constructed from was slightly different than our current Aluminum. We closely examined the aluminum in our box and realized there were two distinct types of aluminum: old aluminum and new aluminum. Our scissor was made from the new aluminum but the Gateway scissor was made from the old Aluminum. We investigated the material properties of the New and old Aluminum by preforming tensile strength tests on a special material properties testing machine found in Miranda and Sarah s high school. They tested the tensile strength of old and new Aluminum. The stress strain curve is displayed below: 28

30 29

31 A Stress Strain curve shows the elongation when a part is subjected to a force. It We analyzed the stress strain curve for the yield strengths, the ultimate strengths, the strains at break and the strain at the yield. We compared those points on both the old and New Aluminum Alloys. These results are displayed in the table below. New Aluminum Old Aluminum % difference Yield Strength (ksi) % Ultimate Strength (ksi) % Strain at Break (in/in) % Strain at Yield (in/in) % It is plain to see that the only advantage New Aluminum has over old Aluminum is that it stretches 51% longer before it breaks. Otherwise, the old Aluminum preformed much better than the new. Could this be the source of our failure? We decided to investigate and discover the old aluminum s exact alloy. Based on the published alloy for the new Aluminum (5052 H32) we determined that the old aluminum is 5086 H32. A table displaying the differences between the two is below. Aluminum Alloy Price Amount Yield Stress (psi) % diff Ultimate Stress (psi) 5052 H32 Aluminum Sheet $ x1 sheet H32 Aluminum sheet $ x1 sheet % So there is a distinct difference between the old and new Aluminum. We found the old aluminum is superior to the new. Knowing this we found bars of old aluminum and replaced the top half of our scissor with the Old bars. Combining the new bars with The old bars allowed us to build more levels and achieve greater heights. Now our robot is five skyrises tall. It still was not tall enough to achieve all our goals, but it was a start. 30

32 The Skyrise Claw THE SKYRISE CLAW 31

33 The Problem After we built a lift, we needed a mechanism to score. We decided to focus on building Skyrises because Skyrises seemed the best possible way to efficiently score the most points. Each Skyrise section is worth 4 points and each cube on the Skyrise is worth 4 points. Since the Skyrise is close to the starting tile, building Skyrises is an easy way to score points in the autonomous period. This is important to us because there is a 10 point autonomous point bonus and that bonus can sway the match significantly. Our current lift configuration is perfect for Skyrise building. The linear motion of the scissor lift is perfect for nesting the Skyrises and for removing them from the Skyrise holders. This is because the front of the scissor lift travels up and down within the same plane. This keeps any front mechanism from traveling forwards and backwards. In other lift designs like 8-bar linkages, the manipulator travels forwards when raised. The arm travels in an arc instead of a straight line. The scissor travels in a line making it perfect for delivering Skyrise sections to the base no matter what the height. The current base design does not move in any direction. This is a problem for Skyrise building because it would need to turn to move a Skyrise from the holder to the base. This problem can be avoided if our grasping mechanism can swing and reach both the Skyrise and the holder without moving the base PROBLEM STATEMENT: After analyzing the scoring opportunities we decided our grasper must score Skyrises while the base remains stationary. 32

34 Relevant Parameters Size requirement. Specification: The robot must be no larger than 18in by 18in by 18in large The manipulator can be no larger than 18 long by 5 wide by 12 tall Testing Method: The sizing tool will be used to test the final arm design and any concept models will be measured before fabrication. Why? Robot Rule R4 constrains the dimensions. The current base and arm also constrain the dimensions of the manipulator The width of base is 14 in wide. To comply with the rule the manipulator must stick out no farther than 4 inches past the edge of the base Reach requirement Specification: The lift must be able to reach all posts and must be high enough to successfully score on all the goals WITHOUT tipping over. Testing Method: A tape measure will be used to measure the final design and all concept models will be measured. The final design will be placed on a field and will attempt to score goals. Why? Since the highest goal is 47 inches tall and the cubes are 8 in tall so the lift must have a reach of 55 inches because to lift the cube on top of the goal, the bottom of the cube must be higher than the top of the highest post. Grasping requirement Specification: The claw must be at least 3.5 inches wide and long. Testing Method: A tape measure will be used to measure the dimensions. All to scale drawings and models must meet this specification. Why? The largest Skyrise piece section has a diameter of 3.13 in. We need our claw to be a bit larger than that to accommodate for any misalignment in gives us a bit of wiggle room when grasping the sections 33

35 Brainstorming Brainstorming the Claw We brainstormed a swinging skyrise arm design with a linear slide grasper and claw made from forty fie degree angle brackets. The whole claw will be mounted on the scissor. A motor with a 12 tooth gear will turn the 36 tooth gear connected to the end of the manipulator. This will sweep the sweep back and forth and allow us to pick up skyrises with ease. The liner slide grasping mechanism ensures the skyrise is grabbed and will not tilt in transit. It also allow one side of the grasper to remain open. This will make clamping onto the skyrises easier. The drawings to the right show each claw layer in detail. 34

36 Building the Skyrise Sweeper 35

37 The New Scissor Lift THE NEW SCISSOR 36

38 The New Scissor After successfully winning the design Award at the Pennslyvania State championships, we decided to redesign the scissor lift. We used the same criteria and problem statement as before only this time we tried to think outside the box. We decided to design a c- channel 1x25 hybrid scissor lift because we believed a hybrid lift would be more stable than any other design. We also used bering blocks and lock nuts at each connection point to stabilize the lift. We designed an screwheadless bearing block system to minimize interference while lifting. It involves wedging two Bering blocks, piece of plastic nut bar and a spacer between the inside walls of a c-channel. A standoff coupler is screwed through the entire bar. This creates a solid Bering block connection with no screw heads on the outside. We created this lift and built an entirely new (yet almost identical) robot around it. 37

39 Building the New Scissor 38

40 39