Background

While operating Forging Machines, it is best practice to clean the Forging Die and oil it after every stroke. This removes any excess material (scale) on the Die that may affect the quality of the next machine stroke.

Usually this is done by an operator and is prone to human error, leading to quality issues.

In this project, I explored automation of the process of cleaning and oiling the Forging die after every machine stroke.


Automated De-scaling

The Descaler was a trolley with compressed air, oil and water provisions. It automatically sprayed a mixture of pressurized water and oil at the Die, followed by a spray of compressed air to remove scale.

The spray sequence needed to be initiated after every machine stroke. To detect a stroke, we made provisions for a Serial as well as a Binary (Limit Switch) input on the Descaler. For complex Sensors like IDEs, the Descaler received a message with spray parameters over Serial after every stroke detected. In the case of Binary Sensors, the spray was initiated based on its state.

Components of the Descaler

  • Pump for compressed Oil
  • Pump for compressed Water
  • Pump for compressed Air
  • 3 Solenoid Valves for controlling the Spray
  • A flexible nozzle for directing the Spray
  • PCB to accept Sensor signal and control the spray

We deployed the descaler at the Forging Press in a nearby Factory. We observed that the parts hence produced were slightly more repeatable with fewer deformation defects.

Observations and Learnings:

We observed that the Descaler occasionally sprayed after the Die was closed for the next stroke. We overlooked the variability in some of the operator-controlled parameters (cycle time, time between strokes, idle time) that affected the functioning of the Descaler. I infered that when a human works alongside an autonomous robot for a particular task, the robot must consider some inevitable human error and deviation from the set process. We compared our Device to it's pedal-operated counterparts and concluded that such a completely autonomous Device would require richer inputs and processing (Image Processing and Machine Learning) to be a 100% replacement for a human Spray Operator.

I worked with the production team at the Factory, a CAD designer, and a Graduate Mechatronics Intern. I designed the functionality and ergonomics of the Descaler. I guided the Intern with the design of the Device and getting all the parts of the Device to work together. I also worked closely with the CAD designer and the production team to get the Device desiged, fabricated and prototyped.



Links:

Synopsis by our interning team member.

Implementation with a semi-automated de-scaler.





Get In Touch

  • Pune (MH), India
  • ranjitbhinge@gmail.com