Time period:
Start Date: September, 2022End Date: Ongoing
Background
For my fourth year project, I decided to build a robotic arm. I was initially going to knock it out of the park by buying and assembling a pre-made kit and adding some custom features for originality, but my faculty advisor had other plans. He suggested we build a robotic arm for replacing humans in the casting process, and he was quite adamant that we try. He even offered to fund the project by himself and so, we were funneled down a project that I didn’t even think was possible in the first place.
The Problem Statement
During casting, humans use a scoop to pick up and pour molten metal from a furnace to a cast. This is dangerous for obvious reasons, and as many engineers do this for long hours each day the fumes have a harmful effect on their health as well. Our problem statement was to attempt to replace this human, who is doing a pretty repetitive job of scooping and pouring, with a robot. The robot would be better at the task than the human, and also safer in the environment. It could also work longer hours without getting tired, and potentially would require lesser pay.
Important Hurdles
Most robotic arms are built using aluminum, its strong and light and easy to machine. But aluminum melts at temperatures that the robot was going to work with. So initially I was horrified thinking we would have to use cast iron for the robot chassis. This would essentially render the project impossible, since cast iron is around 4 times heavier than aluminum, and the robotic arm motors would not be able to handle such weight.
We tried building many iterations of models, did multiple calculations and everything pointed to what is written above. This project seemed like it could just not be done. Cast Iron was too heavy. Skinning and making every part hollow did not solve the problem. We could not find motors with higher torque. We were working with no factor of safety. Accuracy of the robot motion would also fall this close to the breaking point.
Solutions
I looked up robotic arm designs online and got a very interesting design idea. We could shift some motors down to the center of the robot and have the power transmitted through belts and gears. This would complicate the design a little bit but would result in all the motors having much lesser torque to handle.
The second revelation we had was that we could actually use aluminum, we just had to not use it in the areas that would be dealing with the insanely high temperatures of casting. The entire bot, hence did not have to be made of cast iron, just specific parts of it had to.
Given these 2 rays of hope, we got to work, did the calculations, made the report and immediately submitted it so that we could be funded and get started.
Conclusion
The project took a lot of work. After designing the entire model, we took it for fabrication to a CNC machine shop only to realise that our design was not practical to fabricate. We had to redesign the whole thing several times over, keeping in mind how each and every component would be fabricated, and the method of fabrication that would be used. Some fabrication methods were not suitable for some metals, and so that added a whole new layer of complexity to the problem.
After several redesigns and fabrication, we finally got to assembly. The whole concept behind designing and planning as an engineer is so that you wont have to redo your work, and hence I was glad when my bot was assembled on the first try. It was proof that I had succeeded as an engineer. We were pretty close to our submission deadline at this point, so I immediately assembled the robotic arm, wired up the Arduino and the stepper motors and got it ready for a demonstration in a single night. Needless to say, we passed our project module with flying colors.
Resources
Links
This is a link to the Google Drive folder where I have uploaded the SOLIDWORKS designs, and the reports of this project, which include the materials, planned weights of manufactured parts, the electronics required, the motors along with their torques and links to buy, the bearings along with their links, and any other information we have gathered to help us plan this to the fullest.
Gallery
