The Finished Rube Goldberg Machine
In the last blog post I shared some of my initial plans and the components of the machine that I had been working on. After a lot of trial and error as well as editing the design of some components, I eventually managed to come up with my final Rube Goldberg Machine and it works! As always I am going to talk you through the design process of the setup before showing off the final video, but feel free to skip ahead.
I am going to start by showing a clip of the final setup and then explain the various components involved. I didn’t build any other components since the last blog post but added in bits and pieces to connect everything together. I might not be completely clear from the video how every part of the machine works but I am going to talk through every element step my step.
One major challenge during this project was how to trigger and transition from the different components I had built for the machine, all within the restricted space of my dining room. I had to find creative solutions as well as design each element to be as accurate and ‘fail proof’ as possible. This turned out to be quite a tricky task.
The Ping Pong Launcher
I decided the very first part of the machine should be the the ping pong ball launcher. This is because obviously I can’t interfere with the machine after the initial trigger and the launcher requires an operator the way it is currently designed.
As discussed in my previous Rube Goldberg blog post I wanted to do some experimental analysis of the launcher in order to determine the optimal spring displacement for a set launch angle and horizontal distance. I was did this quite successfully and was able to mark a location on the launcher to make it a lot more consistent. This was extremely helpful as I didn’t want to have to worry about the launcher missing and knocking down finely balanced components farther down the line and setting of the chain reaction.
To avoid this problem every element had to be designed carefully to perform consistently and predictably.
The Buoyancy Device
The next element in the machine is the Buoyancy Device described in my last blog post. This component is triggered through the removal of a cork which lets out water held within the bottle. I decided the easiest and most effective way to do this would be to knock it out, side on, rather than pulling it out. Since the cork has to be placed in quite firmly, a significant force would need to be applied in order for it to be knocked out. This meant that simply redirecting the ping pong ball wouldn’t do the trick as it is too light (F=ma).
My next thought was to use a pendulum with a relatively large radius and starting height to build up enough momentum. This also needed to have a fair bit of mass on it so I used the steel head of a broken hammer. This can be seen in the image below.
As the head is quite heavy, a ping pong ball didn’t have enough force to knock it off of its perch on the edge of the pipe. This meant that I had to use a golf ball rather than a ping pong ball in the launcher which obviously required me to reassess the correct launch angle and spring displacement.
Since the last blog post a small change was made to the design of the Buoyancy device in that the size of the water outlet was made smaller to lower the plastic stopper less suddenly. This made the ping pong ball roll into place more consistently and less likely to fall off the side of the rails. It also slowed down the pretty rapid chain reaction quite nicely. I managed to do this by drilling a hole in the cork and fitting a wooden dowel inside. Blue food dye was also put into the water to make the process more clear on the final video.
Throughout the various attempts of the machine the pendulum and buoyancy device were extremely reliable and I never had any problems with them. This was likely due to the fact that I had positioned them very carefully so that the pendulum swung in the same plane where the dowel was outstretched. This can be seen quite clearly in the image labeled ‘Hammer Head Pendulum’ shown above.
The Mallet Projectile Launcher
This leads on to the next device I designed, again described in my previous post. Once the ball is released from the buoyancy device it travels down the rail it continuous over a thin plastic taken from an old sun blind. This can be seen in the bottom left of the image shown below.
The ping pong ball then knocks into the Jenga blocks which fall like dominos right onto a lever positioned under the mallet. This was designed to cause an anticlockwise moment on the mallet, tipping it onto metal strip used to launch the ping pong ball. These few steps may seem trivial but adapting them to perform correctly was everything but.
One aspect that was particularly challenging was setting up the metal strip to fire the ball to the next stage of the machine consistently. The metal strip is essentially a lever with the fulcrum over the cork, which acts as the pivot point. This means that the work output can be adapted through changing the length of the metal strip either side of the fulcrum. With the same amount of force being applied through the mallet on every swing, the output kinetic energy varies according to the length of the metal strip past the fulcrum (W=F x d).
The ideal mechanical advantage (IMA) given by this system is proportional to the the input distance over the output distance relative to the fulcrum. This can be visualised quite well through the image below taken from Control Automation.
Therefore through increasing the input distance, more energy is transferred to the ping pong ball. This, as for the original launcher, was tested experimentally to find suitable values for the input length.
The ball was fired into a plastic container tilted at an angle so that the ball would roll/fall where another Jenga block is carefully balanced. This is shown in the top left of the image below.
The Jenga block is positioned to fall on any contact and is attached to another block via a pink string tied to both. This pulls down the other Jenga block with it, setting off the next stage of the machine.
The Pulley System
Once the Jenga block is pulled away, a long line of ping pong balls are released and allowed to flow down another rail hanging from the ceiling. These fall straight into a plant pot suspended underneath the end of the rail. This is shown in the image below.
The pot is attached to a plastic bottle partially filled with water over a single pulley. The weights of the pot and bottle were balanced so that the pot would remain suspended until filled with the ping pong balls.
Due to frictional forces within the pulley, the pot needed to be significantly heavier than the bottle in order to pull the bottle into the air. Without the ping pong balls I measured the mass of the pot and bottle to be 355g and 260g respectively. This measurement is only accurate to plus or minus 5g due to the sensitivity of the kitchen mass balance I used.
With around 15 ping pong balls added to the pot, its mass was approximately 390g. This was enough to cause the pot to start a slow decent. Even if a number of balls were to miss, I expected the force of the balls hitting the base of the pot would likely be enough to overcome the static friction of the pulley. This friction force is significantly higher than the dynamic friction meaning that once the pulley gets moving the weight of the pot would likely keep it descending.
As it gets close to the floor, it is lowered into the path of the Tumbller’s ultrasonic sensor, which pick up a signal.
The code I wrote for the Tumbller is quite similar to that used previously for the Obstacle course. Once the sensor detects an object within a predefined range (in this case 25cm) the motors are instructed to rotate in a specified direction, at a specified speed, for a specified period of time.
The purpose of this was for the Tumbller to reel in a sign suspended at the end of another pulley once a signal was received from the ultrasonic sensor. This was done by fitting plastic ‘guide rails’ on either edge of one of the Tumbller wheels and tying a fishing line around the centre. I had previously though to incorporate the Tumbller into a lot more of the machine but found that almost none of it was at the same height or on ground level, making it very difficult to move the robot between components. On the other hand I cannot think of a more ‘starring role’ than flag bearer for the Tumbller as stated on the project brief.
Originally I had the fishing line tied to the axle but quickly discovered that it would take an eternity for the sign to be pulled up due to the its very small diameter. Obviously the larger the diameter of the rotating cylinder, the more area is being revolved per unit of time. This clearly leads to more fishing line being reeled in the same amount of time, something I initially didn’t consider.
The code used is shown below. Just as for the obstacle course I wrote forward and stop functions to get the motor to either start or remain at rest. I increased the speed of the motors in the forward function to 150. It wasn’t at all necessary to do this for both motors as only one was being used to pull the line but I guess I just like things symmetrical and in sync.
The code for the ultrasonic sensor function is exactly as described previously when working on the obstacle course. You can find a link to this here.
Finally the functions were placed in the void loop so that the ‘STOP’ function runs if the distance measured by the sensor is over 25cm and the ‘forward’ function if under 25cm. The time set for the motors to be active was 4 seconds based on trial and error to see how high the Tumbller would lift the sign.
And that concludes the my Rube Goldberg Machine.
I know there was a lot of taking between the different stages of the machine so I’m just going to list the steps quickly before showing off the completed run.
- The Ping Pong Ball Launcher fires a golf ball in through a funnel and down a pipe, knocking into the hammer head pendulum.
- The pendulum swings across to the Buoyancy Device, hitting out the wooden dowel which holds in the water.
- As the water flows out from the device a thin bit of plastic attached to floating corks is lowered, releasing a ping pong ball.
- This ball runs down a set of rails towards a series of Jenga blocks.
- The blocks are knocked over, falling onto a lever, which in turn makes the mallet fall and launch another ping pong ball.
- This is fired into a container where another Jenga block is carefully balanced.
- The Jenga block falls, pulling down yet another Jenga block, which was holding back a series of ping pong balls.
- The balls flow down the rail and into a plant pot.
- This changes the balance of the pulley and causes the pot to be lowered down.
- The ultrasonic sensors of the Tumbller detect the obstruction of the pot and it starts its motors.
- As the motors turn the fishing line is wound up, pulling up a sign on the other side of another pulley.
- The sign is then displayed.
Thanks for you patience in reading through my explanation. This now, is the final Rube Goldberg Machine run!
Thank you for following!!
As everything moved so fast towards the end of the run, it was very difficult to move around and catch everything on video. Unfortunately I sort of missed the ping pong balls being released and falling into the bucket.
This part can be more clearly seen in another video I took, which was a failed attempt where the Tumbller didn’t pick up a signal on its ultrasonic sensor. You can see in the video that I had to stick my hand out in front of the robot to get it going.
Luckily I didn’t require all too many attempts before getting everything to work in one. I had originally expected to be stuck filling up the buoyancy device all day with thing going wrong farther down the line. I’m really pleased with the end result though and have to say it was a lot of fun putting it together. I would definitely recommend giving it a go if you find yourself with some spare time.
As in the sign, I wanted to thank everyone for following and keeping up with the different projects I discussed here and on my Instagram page. This is the end of the project assignments for my design module but I do hope to continue working on some projects of my own and discuss them here. I have a long summer ahead of me after my exams where I can definitely see myself ordering some engineering tech to play around with.
I hope you enjoyed following everything going on here and thanks again.
