ME 250 DESIGN AND MANUFACTURING I

Fall 2013

Final Report

Team 4

ME 250 Section #8, Team #4

Team Members

Christina Gandolfo: gandolfo

Andrew Schneir: aschneir

Daniel Scott: scottdan

Stephen Trolio: troliost

Figure 1: “R.I.P. DAN” Robot .

Table of Contents

1  Abstract   3

2  Introduction   3

3  Vehicle Design    4

        3.1 Squad Strategy Selection    4

        3.2 Vehicle Objectives and Requirements    5

        3.3 Player Concept    6

        3.4 Detailed Design  7

                3.4.1 Analysis Model   7

                3.4.2 Solid Model   8

4  Product Manufacturing    10

        4.1 Manufacturing Resource Description    10

5  Bill of Materials    11  

6  Prototype Testing  11

        6.1 Preliminary Test Results    11

6.2 Redesign Based on Preliminary Tests   12

        6.3 Discussion of Competition Results   12

7  Discussion and Recommendations    13

        7.1 Project Summary   13

        7.2 Future Project Idea   14

8  References   14

9  Acknowledgments    14

Appendices  15

1 Abstract

This report is concerned with the analysis and review of Section 8 Team 4’s experience in this fall’s ME 250 course.  We will discuss the robot’s design, the manufacturing processes we used, the pre-competition testing we did, and the results from our competition.

After deciding on a squad strategy and our specific role in the game, we began an extensive design concept generation process through which we decided on our final design.  We had been given the job of an adaptable player that would adjust its plan of action (offensive or defensive) based on the strategy that our opponents chose to use against us.  We ended up selecting a design that had a plow in the front for offensive strategies and had distracting moving arms on the sides for more defensive strategies.

Throughout the course we were given the tools, literally and figuratively, to manufacture the parts we needed to build our design. We spent many hours in the shop machining parts and assembling the body and different components of our robot.  This document contains the manufacturing plans and details the machines and tools we used to make the parts we designed.

Once our robot was finally assembled we ran a few unsuccessful preliminary tests that required us to rebuild the double gearbox several times.  Once we finally got the robot to function properly we began to familiarize ourselves with the controls and prepared for the competition.

The competition went very well for our squad. If it hadn’t been for a malfunctioning robot and for one questionably route choice made early on in second round, we would have been in the championship game.  We are all very proud of our third place title though.

2  Introduction

M-ball, a game created by our professors is based on scoring different weighted balls into either a hole in the court or a raised basket.  All of the different scoring options allow for a wide range of strategies and design possibilities.  Our goal as a team was to provide the squad with either a reliable offensive or defensive player, which could adapt depending on the opponent’s strategy.  We were provided with an extensive list of rules and guidelines to be followed in the competition.  Some of these were design constraints that we had to maintain. These  included: size - less than 10” x 12” x 15”, weight - less than 15 lbs., and we had to have less than four motors on the robot.  

3  Vehicle Design

This portion discusses our robot’s role regarding the squad’s strategy.  We will elaborate on the robot’s design and the analysis we did to come to that design.  Also included in this section is our final solid model of the robot.

3.1 Squad Strategy Selection

Our initial squad strategy concepts were focused around having a balanced team, with two players working to score points and two playing defense to prevent the other squad from scoring. We believed that offense would be the more difficult task, as it required more finesse to get the balls and then put them in the basket or push them into the hole while the other squad’s defenders tried to stop them from scoring. Due to this, we contemplated the idea of using three players on offense and only one on defense. We concluded that having a sound defense and having our opponents score minimal points would prove a strong strategy. Having too many robots on offense would make it congested and difficult to maneuver.

We also considered the fact that in order to properly defend both the basket and the hole, we would have to devote two players to defense. Once we had determined that we would use two players on defense, one to defend the hole and one to defend the basket, we speculated on what our strategy should be on offense using the other two players. We realized we needed to make a player that would be able to get the squash balls off of the tower because they were worth the most points. We wanted that player to also be able to place balls into the basket because getting balls in the basket was worth more points than if they were dropped in the hole. The other offensive player should be able to score both squash and ping-pong balls in the hole.  All of the potential strategy sketches are provided in Appendix A.

Once we had the basic strategy finalized, we wanted to refine it. Since we could never be sure of what the other squad would do in a game, we wanted to have a versatile player that could adapt to the opponent’s strategy. This player would start off on defense, blocking the opposing squad’s hole. If the opportunity presented itself, this player would attempt to push the balls into our squad’s hole and then quickly go back to defending. This player would be built robustly, allowing it to block the other players and push them.

Our second player was designed with the intent of getting the squash balls off the top of the tower and into the basket. This would allow us to score a lot of points using just a few of the balls. The next player was meant to play defense on the other squad’s basket. It would block off the basket so that they could not get any balls in. The last player was designated to retrieve that heavy squash balls under the basket. These balls were worth a lot of points and we wanted to be able to quickly get them and push them into our hole. The player could also work on pushing the ping-pong balls into the hole after all the black balls were gone.

While we were debating our squad’s strategy, we were very concerned with how opponents would counter us. Our final strategy ended up being very balanced and adaptable, which is difficult to counter. The only way we could be beat is if the opposing squad dedicated most of their players to defense and covered our hole and basket completely. However, if they did that, then we would be free to push the wolverine over their hole, which would win us the game, assuming no other balls were scored. The most likely ways we would lose were if one of our players malfunctioned or if we were simply out-played by the other squad. The key to our strategy was that the first player was meant to be versatile and adapt to game conditions, which was meant to prevent us from being easily countered.

3.2  Vehicle Objectives and Requirements

Certain characteristics of our robot were based on constraints given to us by the M-ball rules and certain characteristics were based on objectives that our robot needed to meet in order to fulfill our team’s role in the squad strategy. However, since many of the objectives were not quantifiable, it was difficult to necessarily maximize or minimize them completely.

The size of the player is a constraint because the player must be within 15”x12”x10”. The dimensions can be arranged in any way. Certain dimensions needed to be maximized in our robot in order for our robot to be the most effective. For example, we maximized the height of our arms and the width of the body of the robot so that when the arms of the robot were horizontal, our machine could take up as much lateral space as possible in order to be an effective defender.

The weight of the player is a constraint because the player must weigh less than 15 pounds. We neither maximized nor minimized the weight of our robot since a heavier robot would require more torque from the double gearbox of our robot, yet we also wanted our robot to have enough speed to be a counter-attacking threat. However, the robot still would need to be heavy enough to not get easily pushed out of position or lifted by an opposing robot.

The speed of the player is an objective that we are trying to maximize. Speed is crucial to our player as it determines how quickly we can get into a position to block the opposing attackers and how able we are to adapt to the opposing strategy. The faster our player is, the more adaptable it can be. Yet, while choosing the gear ratios for our double and planetary gearboxes, we still had to take into account the starting torque that would be required to overcome the static friction between the wheels of the robot and the arena.

Turning radius is an objective that we are trying to minimize. In order to minimize our turning radius, the torque provided by each side of the double gearbox to each of the back wheels would need to accelerate the robot, but at the same time, we needed to figure out what to do with our front wheels. If our front wheels were not able to turn and were at the same width as the back wheels, then the only way that our robot would have been able to turn would be with if the friction between the arena and the front wheels was too great, and the front wheels slipped. Therefore, our front wheels were able to pivot on ball bearings so that there would be minimal friction in the front wheels that the robot would need to overcome in order to turn. As an adaptable part of our squad strategy, we felt that this minimal friction and minimal turning radius would prove most helpful in being able to quickly respond to the opposing strategy.

As a team, we wanted to be able to maximize the offensive effectiveness of our robot in order to make it a counter-attacking, versatile robot. Therefore, with a strong ability to score from a defensive position, our robot would be able to help our team win by not only stopping the opposing team from scoring, but possibly stealing their balls and scoring them into our own hole. Therefore, our robot featured a plow that was bolted on the front of the robot with an open mouth so that we can maximize the amount of balls that the plow can collect.

A final objective that our team focused on maximizing was simplicity. We wanted our robot to be functional with an easy to manufacture design, and we wanted the controls to be simple. Two major ways our robot would fail during the competition, and thus leave our squad at a severe disadvantage, is if we were not able to manufacture all of the parts for our robot within the given tolerances, or if we got confused by adding too many moving parts, thus too many motors. Our team was aware that the simplest design is often the most effective and easiest to perfect.

3.3 Player Concept Design

All of our concepts included different combinations of components that had defensive and offensive uses.  Concept 1 included moving arms on the side and a plow in the front. Concept 2 included rotating flaps on the sides and a forklift in the front. Concept 3 included an arm in the front that could move up and down as well as swing in both directions. Detailed sketches of all 3 concepts are provided in Appendix B.  The ratings that each concept was given for each characteristic is based on a 1-10 scale.

Table 1: Concept generation matrix.

Although all of the concepts produced fairly similar scores, we decided to go with concept 1 because it had the highest score and it had the most opportunity to score on its own.

3.4 Detailed Design

This section includes a detailed image of our final design and the analysis of it.

3.4.1 Analysis Model

We chose to use the double gearbox motor because it was the simplest option, provided turning abilities, and was the most reliable for keeping our axles in line.  Another characteristic of the double gearbox that we found desirable was the number of options for gear ratios. To choose the gear ratio for our double gearbox, we did the following analysis.

As the versatile player of the squad, we didn’t have a set plan of action. We decided that our default first move should be to go straight to the opponent’s hole.  This was 5 feet away, and we figured we would need to get there in about 3 seconds or less.  Using this we calculated our average acceleration using a kinematics equation:

                                                       (1)

Given our mass of the robot was 3.464 kg. and theoretical acceleration above, we calculated the force needed to move the robot:

                                            (2)

We divided this number by the two wheels to get the amount of force each wheel needed to apply to the court: 0.586 N. Using this number we found the torque required by each motor:

                                    (3)

At this point, we went back to the specifications of the given motor and we calculated an appropriate gear ratio configuration.

                                     (4)

We found that we needed a gear ratio of 40.27:1 to accomplish our required torque, thus we went with the 38.2:1 ratio for the double gearbox.  After we made this decision, we realized we had not taken friction into consideration. This error increased our necessary torque and therefore our calculated gear ratio. We were satisfied with a 38.2:1 gear ratio that provided slightly more torque than 40.27:1.

The planetary motor that we used to raise and lower the arms on the side was configured to have a 25:1 gear ratio.  This decision was based on the same analysis, did previously for the double gearbox. We can also justify this gear ratio because speed was not as important, and we just needed enough torque to ensure the functionality of the arms. 

3.4.2 Solid Model

Our final design for the robot included lowering and raising arms on each side and a plow in the front.  The arms were attached to a planetary motor by kevlar thread, and pivoted about an aluminum dowel that was attached to the body of the robot.  The plow in front was bolted to the front of the body.

Figure 2: Final Solid Model.

The gearbox, Figure 2, is not visible in the previous image, but was a pivotal component of the robot. It was attached to the underside of the top body piece.

Figure 3: Gearbox and housing assembly.

We found that as we were manufacturing the gearbox housing above, the side walls were unnecessary, because the strength of the axle was great enough to withstand the stress applied by the weight of the robot.  The wheels that we ordered from McMaster were far heavier than we had anticipated.  We decided to counter this increase in weight by milling out five equidistant holes, as shown in Figure 2.

Another change we made was the shape of the plow.  Initially we had a much more angular design, that would have been heavier and less adaptable (e.g. we wouldn’t be able to “hook” onto or push another robot with it).  We ended up going with a more classic plow shape.

We also added hanging wooden dowels to the the arms using a ⅜” dowel and string.  We decided on this addition because it would stop balls from rolling past us into the enemy’s side of the court.

4  Product Manufacturing

This section includes the manufacturing plans and details the different machines we used to make the parts we needed to build our robot.

4.1 Manufacturing Resource Description

We used the drill press with the #29 drill bit and a #3 center drill to make all the holes for the metal plating and the L-brackets.  We used ¼” and  ⅜” drill bits for the arms.  Also, the holes that we added to the wheels were created using a 1” drill bit.  We used the shear cutter to cut all the different plates that we used to construct the body. The lathe was used to create our rear wheel axles.  For this it was necessary to use a live center in order to minimize deflection. We also used the lathe to decrease the radius of the rear wheels.  To create the L-brackets we used to attach everything, we used the bandsaw to cut the 1/16” 90-degree aluminum stock to size.  We also used the bandsaw to cut the arms to size.

5  Bill of Materials

Table 2: A graphical organizer of our bill of materials

6  Prototype Testing

The following section includes the discussion of our preliminary testing, how we had to adjust our design after those tests, and the results of the competition.

6.1 Preliminary Test Results

On Tuesday, December 3rd, we thought we had finally finished our machine. All of our manufacturing had been completed and our prototype was ready to be tested on the arena. However, the biggest problem that our team encountered was with neither the design nor the manufacturing of our robot. The reason we were having trouble with our robot was the extreme ineffectiveness of the double gearbox. The set-screw that was supposed to lock the motor shaft in the double gearbox in place continuously came loose despite us using maximum strength lock-tight to tighten it, so the gears would slip and the entire axle would come out of the robot since the motor shaft was press fit inside the axle. This clearly posed a major problem, because our robot could not functionally move, even though our robot was manufactured well. This was an unforeseen problem because our team overlooked the actual quality of the gearboxes.

On Wednesday, December 4th, when we tested our machine for our GSI Yihao Zheng, we attempted to minimize the slippage of the motor shaft and the gears by continuously tightening the set-screws and applying extra lock-tight. Our prototype then performed more as expected, with the exception of the battery not being fully charged. Our robot was able to navigate the entire table, and prove itself as a quick, versatile player, exactly the role designated to our robot in the overall squad strategy.

When considering how the robot would perform in the M-ball competition, we had to take into account the possibility of gear slippage and how charged the battery was. If a gear slipped slightly, the robot would have less torque and speed, so our robot would be significantly less effective. If the battery was not fully charged when we received it, our robot would be significantly slower and be able to produce less torque. Also, we had to take into account that as each round of the competition progressed, the battery would become more and more depleted of energy, so our robot would act more like in the tests on Wednesday. Thus, when developing our team strategy, we had to make sure our set-screws were as tight as possible, and our robot would not use an unnecessary amount of energy.

6.2 Redesign Based on Preliminary Tests

Our biggest problem was the issue of the motor shaft coming loose and the entire axle being able to slip through the ball bearing and come completely detached from our robot. After taking apart the double gearbox and reassembling it, it was clear that the set-screw that was supposed to lock the shaft in place kept coming loose enough for the shaft to slip through the gears. With the application of lock-tight, we were able to rectify the issue, and keep the set-screw tight enough to ensure that the shaft would not slip out of the gearbox.

Another small design change was that prior to testing, we realized that the front wheels set the front of the robot too high off the ground. As a result, we had to readjust the length of the strip of sheet metal that the front wheels were attached to. We, then, drilled new holes into the side of our robot to set the wheels in such a position that the front of the robot would be more level with the back. Because of this adjustment, the weight distribution was more level and took unnecessary stress off of the back wheels.

6.3 Discussion of Competition Results

After a long afternoon on Thursday, December 5th, Squad 8 won third place in the M-ball competition. Our overall squad strategy was sound, as our defense was solid, and we knew that with eight robots on the arena during a match, the arena would get very clustered and offense would become a mess. However, our strategy was slightly undermined when the robot that was supposed to reach the squash balls on the tower could not fulfill its designated function. Therefore, our squad had to improvise and this robot’s new role was simply to drive into the opponent’s offensive hole, so that the malfunctioning robot could still have a strong role as a defender and help our squad win. Our squad could have avoided losing in the semifinals of the tournament if our squad had been more careful while scoring balls and defending the opposite hole, since several of the robots fell into the holes and were not able to move afterwards. Overall, our squad strategy was good, and because of that, we were able to win two games and secure third place in the competition.

7  Discussions and Recommendations

The following section includes a summary of the project and course as a whole, as well as a few suggestions and ideas for future competitions.

7.1 Project summary

The project began with learning the basics of Solidworks and Computer-Aided Design tools. Amidst learning how we would eventually be able to build a solid model of our robot, we went through all of the training in the machine shop, where we learned first and foremost about safety. We also learned how to use machinery such as the mill, the lathe, the drill-press, the bandsaw, and how to adjust these machines depending on the material that we use.

Our squad then decided on a strategy that we would use in the M-ball competition. Our strategy included a robot that would be able to reach the squash balls on the tower, a robot that would focus on the squash balls under the basket, a robot to block the basket, and our assignment, a robot that would be versatile enough to defend the hole, yet score ping-pong balls and squash balls that fall to the table. Knowing our squad’s strategy allowed us to begin designing our robot. We felt that a short, boxy robot would best fulfill our team’s role in the squad. We also added arms that we decided could be raised and lowered with a planetary motor in order to allow our robot to extend its reach beyond the size requirement that we had to meet. Additionally, we added a plow on the front face of the robot so that we can easily capture and score loose balls.

With drawings of our robot and knowledge of how to use Solidworks, we were able to create preliminary solid models of our robot. After deliberation on certain details, such as whether to use spider couplings with the axles (which we decided not to use), we had finalized our solid model and were ready to create engineering drawings of our parts and begin manufacturing the robot.

After a few weeks of manufacturing, and buying supplies such as a 90-degree, L-shaped aluminum bar from Home Depot for our manufactured L-brackets, our robot was finally finished. We soldered the motors and connected the motors to the control module and the battery pack and were able to commence testing. Our team’s design of our robot was very strong. The robot was able to do everything that it was supposed to be able to do, plus the design of the body worked nearly flawlessly. However, our team did not put enough time into considering which gearbox we would use to drive and steer our robot. We ended up using the double gearbox for simplicity reasons, which was not the smartest idea. In hindsight, we should have used two metal gearmotors, one for each of the back wheels. The double gearbox was weaker and cheaper than we anticipated, so most of our robot’s problems stemmed from the double gearbox. The metal gearmotors would have been much more reliable and stronger motors to give our robot not only more torque, but more speed as well. The double gearbox was clearly the weak point of our design, but after a few modifications to the gears and set-screws inside of the gearbox, we were able to improve the efficiency and decrease the gear slippage of the gearbox. The strong point of our design was the planetary gearbox that lifted and lowered the arms to block opponents from scoring. They were easy to manufacture, provided a convenient way to organize the wires from the double gearbox, and were very easy to operate.

7.2 Future project idea

This course is based around University of Michigan’s original game M-ball, which is fun and requires a lot of creativity and strategizing, but it has some downfalls.  The game requires two teams of 4 robots to be playing at the same time, which creates a lot of congestion and confusion.  Although the court is very big, consisting of two adjacent ping-pong tables, the game is very one sided and one whole side of the court is completely unused. Most robots do not get a chance to show what they can really do purely because there is not enough space for them to move.  This could be avoided by lowering the number of robots on the court to 6, or by increasing the size of the court.  Another issue we encountered with M-ball was that many strategies involved specialized robots, and if they didn’t function properly, were useless to the squad.  All of these issues are addressed in our proposed new game.

If the entire idea of M-ball was rejected, a new competition to consider might be one based off of a favorite childhood game: “Capture the Flag.” This would bring new aspects of competition to the court: speed and adaptability.  We imagine the game would go as follows:

Each team has a designated side of the court on which they get to place the “flag” (a squash ball) that their opponents are trying to capture and place in their own hole. This would still allow teams to decide to use an offensive or a defensive strategy.  It also still allows for a lot of creativity in designs.  Because the location of the flag is unknown, this prevents teams from specializing their robots for specific tasks, and therefore there will be more success within each team.  When everyone’s robot is working properly the squad’s morale is a lot higher and the competition is more intense and more fun for all parties involved.

8 References

We occasionally used the The Machinery’s Handbook, 26th Ed. provided by the machine shop staff for specific numbers (reamer size, drill and lathe speeds, etc.). Our GSI, Yihao Zheng, also helped greatly by providing insight on how to alter our design in order to make the manufacturing of the robot much more manageable.

9 Acknowledgements

We would like to thank the following members of the ME 250 staff, for providing the resources and help that we needed to complete this project.  Our amazing graduate student instructor, Yihao Zheng, was instrumental to the success of our robot in the competition. We would also like to thank Mark Stock and Bob Coury, the experts of the machine shop, for their great advice about manufacturing the parts we made.  Professor Panos Papalambros and Lecturer Mike Umbriac provided us with the necessary knowledge to design, build and test a functioning robot that could compete in the M-ball tournament.  We owe all of our success to these people.

Appendices

A. Strategies

Strategy A.1 

Two attacking robots, one of which will focus on high balls, one will focus on squash balls under the basket. One defensive robot which will focus on defending the basket. One versatile defender which will focus on defending the hole and counterattacking and scoring loose balls.

Figure A.1. Strategy with two offensive players, defensive player, and versatile defender.

Strategy A.2

Three robots are very attack-oriented. One of the robots focuses on the high balls on the tower, while the other two attackers just focus on scoring ping-pong balls and other squash balls. The fourth robot defends both of the hole and the basket.

Figure A.2. Strategy with three attacking players and only one defensive player.

Strategy A.3

One robot focuses on pushing the tower over our hole so that we can score all of the ping pong balls and the high balls will already be on our side. One robot gets the high balls into the basket. The other two robots defend the hole and the basket respectively.

Figure A.3. Strategy with a player to push the tower.

B. Concept sketches

B.1. Concept 1

Concept 1 consisted of almost everything that emulated our final solid model and our robot. It featured a plow in front for collecting and scoring balls, arms that were able to be lowered and raised with a planetary motor, and a bumped attached to the back of the robot in order to push the wolverine.

Figure B.1. Concept sketch 1 featuring a plow and extendable arms

that can be lowered and raised.

B.2. Concept 2

Concept 2 had a similar body shape as Concept 1, where both of the robots are fairly boxy, yet this concept featured a forklift powered by a rack and pinion that would be able to lift the wolverine and move it over the opposing scoring hole, and would be able to lift robots on the opposing squad. The arms in this concept are extendable flaps that would be more complicated to supply power to.

Figure B.2. Concept sketch 2 featuring a forklift (with a rack and pinion)and extendable flaps for arms.

B.3. Concept 3

Concept 3 is once again a fairly boxy robot with a rack and pinion to raise and lower a forklift, but this robot’s forklift would be attached at a pin so that it could swivel somewhat freely in order to hit balls out of the path of the opposing squad and so that it could distract and annoy opponents.

Figure B.3. Concept sketch 3 featuring a swiveling arm in front of the

robot which doubles as a forklift with a rack and pinion.

C. Dimensioned Drawings and manufacturing plans

C.1. Dimensioned Drawing(s) of Individual Parts

Figure C.1: Front/Rear plate of player body

Figure C.2: Side-front plate of player body

Figure C.3: Side-rear plate of player body

Figure C.4: Top plate of player body

Figure C.5: Plow

Figure C.6: Top plate of plow

Figure C.7: Arms

Figure C.8: Axles

Figure C.9: Rear wheels

Figure C.10: Bottom Plate of gearbox housing

Figure C.11: Custom-made L-brackets

C.2. Manufacturing plans (4 points)

Part Number: ME250-001

Part Name: L brackets

Raw Material Stock: Aluminum 3/4" 90-degree bar, 1/16" thick

Part Number: ME250-002

Part Name: body plating, gearbox housing plating (except for top)

Raw Material Stock: 6061-T6 Aluminum plate, 1/16” thick

Refer to “me250-w13_MS6toend” (p10) on Ctools for a more complete manufacturing plan.

Part Number: ME 250-003

Part Name: top of body

Material: 6061-T6 Aluminum plate, 1/16” thick

Part Number: ME 250-004

Part Name: axles

Material: Aluminum rod, 3/8" diameter

Part Number: ME 250-004

Part Name: arms

Material: Aluminum Square Tube Stock - 3/4"x3/4", 1/8" Wall

Part Number: ME 250-006

Part Name: Rear wheels

Material: Rubber wheels

D. Purchased and Traded Items

D.1 Purchased items

D.2. Traded items

We used no traded parts in the manufacturing of our robot.