The aim of this study was to develop an Artificial Neural Networks (RNA) able to model the Friction Stir Welding process (FSW) to provide an accurate prediction of the mechanical properties of AA 5754-H111 butt joints. The input data used for the development of neural models are taken from previous experimental investigations, which are developed according the Design of Experiments (DOE) techniques; all the joints were monitored on-line by means of infrared thermography techniques and they were characterized with destructive and non-destructive testing (visual and macro graphic analysis, tensile and Vickers microhardness tests), in order to highlight their mechanical characteristics. Finally, the significance of the FSW process parameters was evaluated by means of the Analysis of Variance (ANOVA). The use of Artificial Neural Networks to model the FSW process, has the aim of optimizing the technological parameters and to favor the development of a stable welding process, that is able to realize joints without defects and with high mechanical properties. For this purpose, two Artificial Neural Networks have been designed. They worked according the "cascade mode" in order to predict the mechanical behavior of the joints in terms of Ultimate Tensile Strength (UTS) and Vickers micro hardness of the Heat Affected Zone (HAZ). The back-propagation learning logarithm and the analysis of many network architectures has allowed us to formulate reliable predictions. This is shown by the results of the comparison between the model results and experimental data that led to the definition of a final model that can predict the quality of butt joints in aluminum alloy 5754 H111 FSW with good accuracy.

Giunti di testa in lega di alluminio 5754-H111 realizzati con il processo di Friction Stir Welding: Applicazione delle reti neurali artificiali per la previsione delle caratteristiche meccaniche dei giunti

DE FILIPPIS, Luigi Alberto Ciro;FACCHINI, Francesco;MUMMOLO, Giovanni;LUDOVICO, Antonio Domenico
2016-01-01

Abstract

The aim of this study was to develop an Artificial Neural Networks (RNA) able to model the Friction Stir Welding process (FSW) to provide an accurate prediction of the mechanical properties of AA 5754-H111 butt joints. The input data used for the development of neural models are taken from previous experimental investigations, which are developed according the Design of Experiments (DOE) techniques; all the joints were monitored on-line by means of infrared thermography techniques and they were characterized with destructive and non-destructive testing (visual and macro graphic analysis, tensile and Vickers microhardness tests), in order to highlight their mechanical characteristics. Finally, the significance of the FSW process parameters was evaluated by means of the Analysis of Variance (ANOVA). The use of Artificial Neural Networks to model the FSW process, has the aim of optimizing the technological parameters and to favor the development of a stable welding process, that is able to realize joints without defects and with high mechanical properties. For this purpose, two Artificial Neural Networks have been designed. They worked according the "cascade mode" in order to predict the mechanical behavior of the joints in terms of Ultimate Tensile Strength (UTS) and Vickers micro hardness of the Heat Affected Zone (HAZ). The back-propagation learning logarithm and the analysis of many network architectures has allowed us to formulate reliable predictions. This is shown by the results of the comparison between the model results and experimental data that led to the definition of a final model that can predict the quality of butt joints in aluminum alloy 5754 H111 FSW with good accuracy.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11586/439351
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact