Thesis defense of Jon Olaizola Alberdi

Thesis defense of Jon Olaizola Alberdi

Title of the thesis: “Soft Sensor-based Servo Press Monitoring”. Obtained the SOBRESALIENTE qualification and the CUM LAUDE and DOCTOR INTERNACIONAL mentions.

---

• Title of the thesis: “Soft Sensor-based Servo Press Monitoring”.
• PhD programme: DOCTORATE PROGRAMME IN MECHANICAL ENGINEERING AND ELECTRICAL ENERGY
• Thesis directors: Eneko Saenz de Argandoña Fernandez de Gorostiza, Aitzol Iturrospe Iregui.
  
• Court:
  • President: Gerardo Aranguren Aramendía (UPV/EHU)
    
  • Vocal: Andreas Sterzing (Fraunhofer IWU)
    
  • Vocal: Elena Silvestre Soriano (Fagor Arrasate)
    
  • Vocal: Mikel Mendicute Errasti (Mondragon Unibertsitatea)
    
  • Secretary: Lander Galdos Errasti (Mondragon Unibertsitatea)
    

Abstract

The force that a servo press exerts forming a workpiece is one the most important magnitudes in any metal forming operation. The process force, along with the characteristics of the die, is what shapes the workpiece. When the process force is greater than the maximum force for which the servo press was designed, the servo press integrity can be damaged. Therefore, the knowledge of the process force is of great interest for both, press manufacturers and users. As such, the metal forming sector is seeking systems that can monitor the process force and the operation of the servo press to analyse process’s performance and predict future deviations in the forming operation. Servo press users want to guarantee the quality of the formed parts and reduce facility downtimes due to malfunctions of the press.

This dissertation addressed the monitoring of the process force and the dynamic performance of a servo press based on a model based statistical signal processing algorithm known as the dual particle filter (dPF). Initially both, the developed model of a servo press and the proposed dPF, have been experimentally evaluated and validated in a reduced scale test bench. The test bench has been designed and manufactured based on a design methodology that allows to replicate the kinematic and dynamic behaviour of different servo press facilities in the same test bench. The experimental validation has been also carried out in an industrial servo press under three different metal forming processes. The estimation results have proved the ability of the dPF to track the process force throughout the evaluated processes, obtaining a deviation lower than 5% with respect to the measured force signals at the maximum force position. The dPF algorithm has been accelerated by means of a field programmable gate array (FPGA) to achieve a real time estimation.