Robot Stability Attachment
Hi everyone and welcome back to the blog! As you may well know from all the pictures and gorgeous renders of the Elegoo Tumbler that I have been posting, the robot only has two wheels. And it doesn’t take a masters in engineering to figure out that the robot will need some help in staying upright because of this. Therefore this week we have been tasked with designing an attachment which will help the robot in balancing upright.
The physics behind this issue is the centre of gravity which can be assumed to be a point. I cannot locate this point exactly on the robot but I estimate it is on a vertical axis of the centroid of the bottom and top plate, a good guess would be just above the DC motors perhaps? The task we are trying to achieve is to keep this point inside the wheelbase of the robot at all times. If the point falls outside the wheelbase then the robot will overturn.
I approached this problem with a simple yet effective mechanical solution, namely a third wheel mounted under the centre front clear plastic of the robot. However I will be making this design as efficient and robot cooperative as possible. To do this the wheel must have the correct caster and have minimum weight. But why am I concentrating on these two aspects specifically you may ask? Well please keep reading to find out why they are so important.
Lets start off with the basics of caster, it is the axis that connects the upper and lower pivots points of the wheel. While positive caster is desirable for our road vehicles because of its increased tire contact and self-aligning torque, negative caster is more advantageous for our application. This is because negative caster will turn with a lot more ease compared to zero and positive caster. Also negative caster can dampen the oscillations and vibrations from the wheel to ground contact. By measuring existing caster wheels I found that the optimal caster is approximately -10 degrees, which is the exact caster I will implement in my design.
My design psyche over the course of this week was to create an extremely lightweight caster wheel with no compromise in reliability. The Elegoo Tumbler uses DC electric motors to move which are powered by the use of a battery. Any extra weight causes extra strain on these vital components of the robot. The robot weighs 2.48 pounds as stated by Amazon which is approximately 1.125 kg. My final design will utilise very low density materials with high stiffness.
This is the first design that I created. The blue wheel material was highly dependent on the available 3D printing plastics, in this case I assumed a standard ABS Plastic giving this component a mass of 4 grams. Next the mount and the swivel will be made out of aluminium. The mount and swivel consist only of simple geometry such as cylinders and square plates. The arms of the swivel that attach the wheel can easily be cut out of a plate of aluminium using conventional subtractive techniques such as a cutting blade or wheel. Finally these components will be welded together using a TIG welder. This combination of materials will leave a small impact on the extra energy needed from the battery and DC motors when it is attached to the robot.
The design outlined above had a mass of approximately 37 grams as measured by SolidWorks, which was definitely in line with keeping mass low. However the next iteration of my design was indisputably better in this category weighing in at a shocking 10 grams, 73% lower than the original. Furthermore using materials such as aluminium can be colloquially described as “overkill” for this application therefore I started building the whole assembly out of ABS plastic. At first I was eager to launch Solidworks again and create a new intricate design to 3D print! When suddenly I came to the realisation that I have unlimited amounts of ABS plastic at home, which has an extremely low density as it is mostly hollow on the inside. Without further ado that ABS plastic in my possession is Lego. I found that for a lightweight robot application such as the Tumbler this is the perfect part to add.
Firstly I detached the front clear plate off the robot. My intention was to use silicone glue to stick the first Lego piece onto this plate. However I didn’t want to make irreversible modifications to the plate. Therefore I then proceeded to wrap it in a layer of clear tape and then apply the glue to the tape. This was a highly successful idea as the tape will always be in compression pushing upwards against the plate and never in tension, so there is no issue of the tape separating. Next when the first Lego piece was firmly adhered, I attached a swivel piece to it. This piece will allow the whole wheel assembly to rotate as the robot is turning.
I then added blocks under the swivel piece making sure to adhere to the negative caster angle. On the very bottom I attached two wheels which can spin freely. Another design feature is the fact that the wheels will sit 2mm higher in relation to the robot wheels. This is caused by the robots high centre of gravity so the robot will tend to lean towards one side. This design cue makes it certain that the robot will always lean towards the side with the third wheel. The 2mm gap will have minimal effect on the overall upright angle of 90 degrees in relation to the ground, but will ensure the robot will be a lot less likely to tip over towards the side with no wheel meaning the loss of stability and balance. It was also important not to stray too far away from a perfect 90 degree angle from the ground as this is the optimal angle for the ultrasonic sensor. Finally this is only 0.8% of the current mass of the robot therefore it should have minimal effects on mobility. Now that we have the theory out of the way lets take a look at the final design!
You may compare the final design to the first prototype and think that I made a mistake going the Lego route. The first prototype looks cleaner and more advanced, however I am certain my final design is superior for this application and my only response is not all that glitters is gold. Finally I considered the contact area of the wheels with the ground in both designs which has an impact on stability and rolling resistance. The first prototype wheel has a width of 25mm whereas each wheel in my final design has a width of 10mm. Even though my second design has a smaller contact area, the outside of the wheels are made from rubber instead of plastic. Rubber has a lot more grip than plastic on surfaces such as wooden floors and tiles which the robot will be used on in a home environment. Furthermore rubber can dampen vibrations and bouncing better than plastic, therefore the final design was definitely preferable.
I hope you enjoyed this weeks blog post focusing on a stability attachment for our robot. Even though this seems like an small and nearly insignificant part compared to some other projects posted on this channel, I can assure you that this is the final piece of the puzzle to make the robot a complete machine. If you like the type of content posted on this blog please consider following my channel and I’ll see you next week with more new and exciting updates on this project!
