This project involved designing and building a prototype automatic welding system that was capable of re-manufacturing the turbine blades using weld deposition. The system interfaced to a robotic welding arm to manipulate the turbine blades throughout the additive re-manufacturing process, thus proving that the system can operate in an environment containing the same hazards and obstructions that a human welder would encounter.
The prototype design allowed us to explore the specific issue relating to aerospace gas turbine blades, namely that after ~30,000 hours of operation, turbofan jet engines are entirely overhauled. The compressor blades are removed for inspection and where possible repaired using manual weld deposition. Investigations into the intelligent automation of the Gas Tungsten Arc Welding (GTAW) process for turbofan blade re-manufacturing has demonstrated that highly skilled welding engineers are required to carry out what is often thought to be a simple task, but is in reality a complex operation.
Existing standard practice for the re-manufacturing of high value compressor blades is the manual weld deposition approach, in which the highly skilled human welders work on-site under extreme environmental constraints (noise, heat and restricted spaces). However, high value components such as turbine blades require a high rate of re-manufacturing success, but this is not achievable with current manual processes that provide less than a 50% yield of re-useable blades. Experienced welding engineers use their knowledge and apply their skills almost automatically by subtly fine-tuning multiple welding control parameters to achieve the required results. Their dynamic inputs alter the weld deposition characteristics such as the size, shape, depth and microstructure of the weld. The variable welding parameters are collectively termed CLAMS (current, length, angle, manipulation and speed). These parameters comprise a procedure known as the “welding schedule”, which is used to provide numerical input for the manual operator to achieve the correct results with a high yield.
By using the Autonomous Aero Turbine Blade Re-Manufacturing System, the welding schedule becomes highly optimised, thus allowing successful blade repair yields of close to 100%. Other applications of this approach to robotic process automation include any scenario in which precise spatial information (including structural limitations) and locational data interconnect in real time, providing a fast route to process optimisation.
Read more about this project on the Aerospace Technology Institute website.